MatchUp

User’s Guide
Copyright
Information in this document is subject to change without notice. Companies, names, and data used in
examples herein are fictitious unless otherwise noted. No part of this document may be reproduced or
transmitted in any form or by any means, electronic or mechanical, for any purpose, without the
express written permission of Melissa Data Corporation. This document and the software it describes
are furnished under a license agreement, and may be used or copied only in accordance with the
terms of the license agreement.
Copyright © 2007 by Melissa Data Corporation. All rights reserved.
Information in this document is subject to change without notice. Melissa Data Corporation assumes
no responsibility or liability for any errors, omissions, or inaccuracies that may appear in this document.

Trademarks
MatchUp is a registered trademark of Melissa Data Corp. Windows is a registered trademark of
Microsoft Corp.
The following are registrations and trademarks of the United States Postal Service: CASS, CASS
Certified, DMM, DPV, DSF2, eLOT, First-Class Mail, LACSLink, NCOALink, PAVE, Planet Code, Post
Office, Postal Service, RDI, Standard Mail, U.S. Postal Service, United States Post Office, United
States Postal Service, USPS, ZIP, ZIP Code, and ZIP + 4.
NCOALink, LACSLink and DSF2 are provided by a nonexclusive licensee of the USPS. Melissa Data is
a nonexclusive licensee of the USPS for DPV processing and a Limited Service Provider Licensee of
the USPS for NCOALink. The prices for NCOALink and DPV services are not established, controlled, or
approved by the United States Postal Service.
All other brands and products are trademarks of their respective holder(s).
Document number: MU0507UG
Last Update: April 20, 2007
MELISSA DATA CORPORATION
22382 Avenida Empresa
Rancho Santa Margarita, CA 92688
Phone: 1-800-MELISSA (1-800-635-4772)
Fax: 949-589-5211
E-mail: [email protected]
Internet: www.MelissaData.com

For the latest copy of this reference guide, visit http://www.MelissaData.com/tech/

 Contents
1 Installation and Requirements
.................................................................................................................................... 5
1.1 Introduction
........................................................................................................................... 5
1.2 What's...........................................................................................................................
New 6
1.3 System
...........................................................................................................................
Requirements 9
1.4 Installation
........................................................................................................................... 10
1.5 CASS...........................................................................................................................
Updates 12
1.6 File Requirements
........................................................................................................................... 13
2 Concepts
.................................................................................................................................... 23
2.1 Matchcodes
........................................................................................................................... 25
2.2 Matchcode
...........................................................................................................................
Components 26
2.3 Component
...........................................................................................................................
Properties 30
2.4 Matchcode
...........................................................................................................................
Combinations 33
2.5 Blank...........................................................................................................................
Field Matching 36
2.6 Matchcode
...........................................................................................................................
Mapping 40
2.7 Optimizing
...........................................................................................................................
Matchcodes 44
2.8 Other...........................................................................................................................
Uses for Swap Matching 50
2.9 Gathering
...........................................................................................................................
and Scattering 53
2.10Dupe...........................................................................................................................
Group 57
2.11Canadian
...........................................................................................................................
Users 59
2.12United
...........................................................................................................................
Kingdom Users 60
2.13International
...........................................................................................................................
records 62
3 File....................................................................................................................................
Menu 63
3.1 New ...........................................................................................................................
Setup 63
3.2 Open...........................................................................................................................
Setup 64
3.3 Save...........................................................................................................................
Setup 65
3.4 Save...........................................................................................................................
Setup As 66
3.5 Exit ........................................................................................................................... 67
4 Setup
.................................................................................................................................... 68
 4.1 Merge
...........................................................................................................................
Setup 68
4.2 Merge/Purge
...........................................................................................................................
Setup 79
4.3 Purge
...........................................................................................................................
Setup 118
4.4 File ...........................................................................................................................
Update 136
4.5 CASS
........................................................................................................................... 152
5 Processing
.................................................................................................................................... 165
5.1 Merge
........................................................................................................................... 165
5.2 Merge/Purge
........................................................................................................................... 168
5.3 Purge
........................................................................................................................... 171
5.4 File ...........................................................................................................................
Update 174
5.5 CASS
........................................................................................................................... 177
6 Analyze and Report
.................................................................................................................................... 181
6.1 Edit...........................................................................................................................
Results 182
6.2 View...........................................................................................................................
Reports 198
7 Tools
.................................................................................................................................... 238
7.1 Browse
........................................................................................................................... 238
7.2 Modify
...........................................................................................................................
Structure 275
7.3 ASCII
...........................................................................................................................
Conversion 278
7.4 Matchcode
...........................................................................................................................
Editor 288
7.5 User...........................................................................................................................
Settings 310
8 Additional Tools
.................................................................................................................................... 336
8.1 View...........................................................................................................................
File 336
8.2 Field...........................................................................................................................
List 338
8.3 Expression
...........................................................................................................................
Builder 340
8.4 Check
...........................................................................................................................
Mapping 343
8.5 Check
...........................................................................................................................
Truncations 344
8.6 Global
...........................................................................................................................
Scatter/Gather 346
9 Reference
.................................................................................................................................... 350
9.1 Command
...........................................................................................................................
Line 350
9.2 Batch
...........................................................................................................................
Processing 353
 9.3 Errors
........................................................................................................................... 355
9.4 dBASE
........................................................................................................................... 358
9.5 About
...........................................................................................................................
Melissa Data Corporation 361
9.6 Copyright
...........................................................................................................................
& License Agreement 364
9.7 Limited
...........................................................................................................................
Warranty 367
....................................................................................................................................... 369
INDEX
1 Installation and Requirements
1.1 Introduction
MatchUp introduces a new concept in mail list processing. CASSMATETM allows you to do
CASS certification and merge/purge simultaneously - in one pass.

You can process CASS only or merge/purge only, but both processes are improved by
combining address standardization and deduplication. Users get more CASS hits,
deduplication is more accurate, processing is easier because there's only one setup, and it's
faster because the records are only passed once.

MatchUp's new features also include direct reading and writing of Excel, Access, dBASE,
FoxPro, ASCII, SQL Server*, Oracle*, and DB/2* tables (* optional); over a dozen additional
matching capabilities; and more reporting capabilities, including the ability to output and print
in Excel and Word.

Installation and Requirements 5

1.2 What's New

The following is a short list of new features. You will find many more as you read the manual.

New Setups:
· "CASS" - to CASS-certify as well as Delivery Point Validate (DPV) a list.
· "File Update" - to post fields from one table into another.

Main Window:
· Now has a toolbar for commonly-used commands.
· Customize the look (right-click) with 4 different toolbar settings and 3 background
settings.
· You can open a setup file by dragging and dropping it anywhere on this window (or
you can still open a setup the traditional way)

Matchcodes:
· Extended 8 simultaneous combinations to 16.
· Extended 3 swap match pairs to 8.
· Added new component types and matching strategies.
· "Start at word:" to start extraction after a specified number of words.
· "Stop at word:" to limit amount of words extracted.
· "Advanced" combinations - combination(s) can be selected to be used only for
Suppression comparisons but not Regular comparisons, etc. Restrictions can be
ANDed or ORed with specified dBASE expression(s).
· "Optimize" button to alter placement of components for faster processing.
· Right-click menu for specifying combination settings.
· CASS validated Zip code matchcode component for CASS users.

Setup: Input Files:

· Advanced Table Type - Base file table on the contents of a field.
· Rank by a specified Custom Expression.
· Advanced Ranking - User-provided list that ranks the priority of a field based on its
contents.

Setup: Output Field Mapping:

· Added new output data types.
· Improved Gather/Scatter options.
· Added Drag and Drop to make field mapping easier.

Setup: Output Files:

· Added "Non-Intersected Record File:".
· Added "Format" to output files to produce alternate output file types such as Access,
Excel, SQL, etc.
· New "Multi-Buyer Field" and "Multi-Buyer Format".

6 MatchUp
 · Added the ability to CASS the output table.
· Drag and Drop field mapping.

Setup: Advanced:
· Added "Non-Intersected Record File:"

Processor:
· Direct access to MS SQL Server 7/2000 (optional), MS Access, Excel, Oracle 8i/9i/10g
(optional), and DB/2 (optional).
· Multi-processor version (optional).

Analyzer:
· Added "Autosize Columns" and "Reset to default settings" option.
· Show/Hide: Added "Dupe groups having hits from (any/all) selected table(s)."
· Added <Source>, <Dupe Group>, and <Status> as available sortable components.
· Added a right-click popup menu for common tasks.

Reporting:
· Enhanced look of reports.
· Added ability to output reports to Excel and Word formats.

Installation and Requirements 7

8 MatchUp
1.3 System Requirements
MatchUp requires:
· Microsoft® Windows 98, Me, NT, 2000, 2003, XP or Vista.
· Intel Pentium III® or better.
· 64 MB of RAM minimum. 128 MB of RAM is recommended.
· 10 MB (+500 MB with CASSMATE, +500 MB with DPV) of available hard disk space.

Installation and Requirements 9

1.4 Installation
Please exit all programs before installing MatchUp.

To Install from a CD:

1. Load the installation CD into your CD-ROM drive.
2. Installation program should start automatically and the install program will appear.
3. Follow the on-screen instructions.

If Installation does not Start Automatically:

1. Click Start on the task bar.
2. Click Run...
3. Type d:\setup.exe (where d: is your CD-ROM's drive letter).
4. Follow the on-screen instructions.

To Install from an Activation E-Mail:

To receive MatchUp via an electronic download, Melissa Data Corporation must be provided
with a valid e-mail address. Electronic orders are followed by a ground shipment of the CD
and manual, which are your proof of ownership.

1. Download and save the MatchUp Demo from the "Downloads" page on our website:
www.MelissaData.com.
2. Run the downloaded file to install the demo version.
3. Once it is installed, run MatchUp. On the demo window, click OK.
4. Go to Help | Activate MatchUp.
5. Enter the activation code that was provided in the e-mail sent to you (use cut and
paste to save some tedious typing).

10 MatchUp
Installation and Requirements 11
1.5 CASS Updates
CASS Updates are mailed bi-monthly.

To Install from a CD:

1. Load the installation CD into your CD-ROM drive.
2. Installation program should start automatically and the install program will appear.
3. Follow the on-screen instructions.

If Installation does not Start Automatically:

1. Click Start on the task bar.
2. Click Run...
3. Type d:\setup.exe (where d: is your CD-ROM's drive letter).
4. Follow the on-screen instructions.

In addition to updated CASS and DPV tables, any program updates will be installed
automatically.

12 MatchUp
1.6 File Requirements
MatchUp processes tables in .dbf format (dBASE III+/IV, FoxPro), Access, Excel, and ASCII
fixed field, delimited, and flat formats. Add-on modules are available which allow direct
processing of tables from Oracle, SQL Server, and DB2.

If you haven't purchased these add-ons, you can still process these formats (as well as
others not listed here) through ODBC. Direct processing is usually recommended for large
runs, as it is considerably faster than ODBC.

Visual FoxPro databases can be processed directly by MatchUp. However, do not use
MatchUp to modify the structure of a Visual FoxPro database. The additional header
information (structure indexes, catalogs, etc.) will be stripped.

MatchUp can process dBASE as FoxPro tables containing memo fields. However, it cannot
process memo fields themselves. Additionally, you should never try to modify the structure of
a database containing memo fields. Also, don't use MatchUp's utilities to sort, copy, append
from or pack records in databases containing memo fields.

Databases and Tables:

Because of the varied DBMS formats that MatchUp supports, we have tried to use a
consistent naming convention for what is a 'Database' and what is a 'Table'.

SQL Server, Oracle, DB/2, and Access:

A database is a database and a table is a table. No mysteries.

Installation and Requirements 13

dBASE and ASCII file types:
The traditional dBASE standard doesn't have the database/table relationship that the 'larger'
DBMS's use. We've adopted the ODBC standard: the file's folder is referred to as "the
database", and the file name is the table. However, if you enter a full path (folder and file) in
an edit where a database or table is needed, MatchUp will split the name appropriately, so
you aren't continually forced into our mode of thinking.

Excel:
The spreadsheet (.xls file) is the database and the sheet (the active tab on the bottom left
when viewing the sheet in Excel) is the table.

ODBC:
The database is the ODBC connection string (which may contain a database name, user
names, passwords, and who-knows-what-else). The table may describe a table or it may
describe something else, it depends on what type of source you're connecting to.

Field Types:
Just as we had to develop a naming convention for databases and tables, we also had to
develop a naming convention for field types. MatchUp uses the following field types:

Fixed Character Any text, usually with a specific size

Variable Character Any text, usually with varying sizes.
Integer Numeric integer values, -2147483647 to 2147483647.
Float A floating point number.
Decimal Numeric values with a fixed decimal place.
Logical A true or false value.
DateTime Chronological data consisting of a date (and maybe time).
Some DBMS's don't support some field types. In these situations, one type may be
substituted for another. For a chart on how a particular DBMS's data type will be interpreted
by MatchUp please see the File Requirements section of MatchUp's Help File.

14 MatchUp
1.6.1 ODBC

Open Database Connectivity (ODBC) is a database-neutral interface. That means an

application like MatchUp can access a table through ODBC without any knowledge of the
table's method of data storage, location of data, or knowledge of how the data access is
being done. For developers, ODBC functions as a 'Rosetta stone' that translates a table's
contents into a universal format.

With ODBC, we can process tables that we would otherwise be unable to access. This ability
comes at a cost, however. Read operations (building keys) typically take 10% longer to
perform. Write operations (updating status fields, and scattering) can take between 20% and
50% longer. With such a decrease in performance, one may wonder why we use ODBC.
There are three reasons:
· The alternative, accessing the data directly, is often difficult or impossible, as the
internal workings of many databases are proprietary.
· Changes made to a database format by its vendor make maintaining custom database
drivers difficult.
· Using a vendor's ODBC driver is a (nearly) guaranteed way of ensuring that a
database's integrity is maintained (particularly with multi-user databases).

Installation and Requirements 15

How to Add or Open an ODBC Table:

At this point, you can select a table in either of three ways:

Method 1:
1. Select the Data Source (depending on the data source, additional prompts may
appear).
2. Answer any additional prompts appropriately.
3. Select the Table. The ODBC driver may ask some additional questions and there may
be a small delay while the list of tables is being retrieved.

Method 2:
1. Click Setup.
2. Select File Data Source or Machine Data Source.
3. Select the desired source from the list.
4. Click OK. The ODBC driver may ask some additional questions and there may be a
small delay while the list of tables is being retrieved.
5. Select the Table.

16 MatchUp
Method 3:
1. Instead of clicking Add, simply drag and drop the desired .dsn file into the list of
sources.
2. The ODBC driver may ask some additional questions and there may be a small delay
while the list of tables is being retrieved.
3. Select the Table.

Method 2 will often work if Method 1 has difficulties connecting to a data source.

MatchUp's ODBC Requirements:

Not all ODBC drivers are created equal. Some provide more functionality than others. When
you try to use an ODBC driver, MatchUp retrieves the driver's capabilities. The following
capabilities must exist to read from a table:

· Conformance Level 1 or above

· SQLGetData() support
· SQLFetchScroll() support
· SQLTables() support
· Read access

If the above conditions are not met, MatchUp will display an appropriate message when you
try to open a table with such a driver. The following conditions are required if MatchUp is to
write to a table:

· SQLSetPos() support
· Absolute Fetching
· Write access

If these last three conditions are not met, MatchUp will still allow you to open a table, but you
will not be allowed to use status fields or scatter to the file (basically any option that would
involve writing to the table). You can check these capabilities in Help | Test ODBC
Connection.

Be sure to use the most recent version of a driver (usually available on the vendor's website),
as we see occasions when an older version lacked capabilities or was much slower than a
newer version. Often the DBMS driver vendor provides a much more complete driver than
the stock Microsoft driver that comes with Windows.

Installation and Requirements 17

Passwords and ODBC:
Accessing ODBC sources that require password access can be a real pain. During several
points of a process, MatchUp may need to re-open ODBC data sources. Each time a data
source is opened, a password may be required. At the very least, this is bothersome. Even
worse, it makes batch (i.e., overnight) operations nearly impossible. There are two solutions:

Solution 1:
If the ODBC driver and database administrator allow it, use a trusted connection. In a
nutshell, a trusted connection is a way of associating a Windows user/password login with a
database user/password. Only Windows NT, 2000, and XP currently support these
connections. Check with your database administrator on the steps to set up a trusted
connection.

Solution 2:
You can supply the password to MatchUp in the setup file (.dt):
1. Open the setup file (.dt) in Notepad or some other text editor.
2. Locate the appropriate source file's section start (indicated by [Source#]).
3. You should see a SourceConnection= entry two or three lines below the section start.
For example:

[Source1]
SourceTable=DEMO
SourceConnection=DRIVER=SQL Server;UID=Marc;PWD=;WSID=MARC

4. Note the PWD=; section. Insert the user's password in this area. If the PWD entry is
not currently present, add it anywhere in the entry.

[Source1]
SourceTable=DEMO
SourceConnection=DRIVER=SQL Server;UID=Marc;PWD=jasmine;WSID=MARC

5. Save the file and test it out.

WARNING: If you use this method, you are compromising your password!!! Be sure
that this does not violate your company policy!!! By using all of these exclamation
points, you can tell we are serious, and will not be held responsible for any negative
consequences.

You may be wondering why MatchUp can't remember the password you enter. It is actually
the ODBC driver that asks for a password, MatchUp is unaware of this request and is not
given the information. The designers of ODBC do this for a good reason: from a security
point of view, it is the only way to ensure that an unscrupulous program couldn't use a
'harvested' password for evil instead of good.

18 MatchUp
1.6.2 SQL Server

SQL Server access is an optional add-on module that must be purchased separately.

How to Add or Open a SQL Server Table:

1. Click the SQL Server tab.

2. Check Use Trusted Connection if your SQL Server connection used this technology.
If unchecked, the standard login security is used (a User Name and Password must
be entered).
3. Select or type the name of the Server.
4. Select or type the name of the Database.
5. Select the Table or View that you want to use. There may be a short delay while this
list is being retrieved.

Installation and Requirements 19

1.6.3 Oracle

Oracle access is an optional add-on module that must be purchased separately.

How to Add or Open an Oracle Table:

1. Click the Oracle tab.

2. Enter the User Name and Password.

3. Enter the name of the Database
4. Select the Table or View that you want to use. There may be a short delay while this
list is being retrieved.

20 MatchUp
1.6.4 DB/2

Oracle access is an optional add-on module that must be purchased separately.

How to Add or Open a DB/2 Table:

1. Click the DB/2 tab.

2. Enter the User Name and Password.

3. Enter the name of the Database
4. Select the Table or View that you want to use. There may be a short delay while this
list is being retrieved.

Installation and Requirements 21

22 MatchUp
2 Concepts
How do you Dedupe?
Nearly every deduplication application uses a matchcode key to represent each record. A
matchcode key is a string of name and address information that best represents the unique
identity of a given record. A typical matchcode key might consist of the Zip/Postal Code, First
and/or Last Name, Company Name, and Street Address.

For speed and storage reasons, it is common to use only small segments of these fields. For
example, only the first five characters of a last name might be used instead of the entire last
name. Pre-processing is often performed on the incoming data to achieve consistency. This
processing may be as simple as upper-casing the data or as complex as performing a CASS
lookup on an address.

Exactly what goes into the construction of a matchcode key can be found in the Matchcodes
section of this manual.

Clustering:
Once a matchcode key is generated for a given record, it can be compared to the keys of
other records. Ideally, every record's key would be compared to every other record's key.
This, however, is not practical in all but very trivial applications because the number of
comparisons grows geometrically with the number of records processed. For example, a
record set of 100 records requires 4,950 comparisons (99+98+...). A larger set of 10,000
records requires 49,995,000 comparisons (9,999+9,998+...). Large record sets could take
several lifetimes to process - and few customers are willing to wait that long!

So, we developers have made a bold assumption that in order for two matchcode keys to be
considered matching, there must be something in the keys that must match exactly. In many
cases, this will be all or part of the Zip/Postal Code. So what we do is only compare records
that are (in this example) in the same Zip/Postal Code. On the average (in the US using 5-
digit Zip Codes), this will cut the average number of comparisons per record by a factor of
thousands.

This concept is known as "break grouping", "clustering", "partitioning", or "neighborhood

sorting". Chances are good that every deduping program you've ever used has used some
type of clustering method.

Here is an example set of matchcode keys using Zip/Postal Code (5 characters), Last
Name(4), First Name(2), Street Number(3), Street Name(5):

02346BERNMA49 GARD
02346BERNMA49 GARD
02357STARBR18 DAME
02357MILLLI123MAIN
03212STARMA18 DAME

Concepts 23
When the deduping engine encounters this set of matchcode keys, it compares all the keys
in "02346" (2 keys), then "02357" (2 keys), and finally "03212" (1 key). For this small set, 10
comparisons are turned into 2.

24 MatchUp
2.1 Matchcodes
As discussed in Concepts, a matchcode key is a string of name and address information that
best represents the unique identity of a given record using parts of several components (like
street number, street name, street suffix, etc).

These components are joined together into a single string. The resultant string is known as a
matchcode key. To compare two records, one needs to compare the records' keys. If they
match, the records are duplicates. Well, it's actually not that simple...

Combinations of Components:
Many deduplication applications use the matching technique described above. It's simple, it's
fast, and works well most of the time. Some of MatchUp's simpler matchcodes use it
themselves. But MatchUp's true power comes from an enhancement to this technique:
concurrent matching.

MatchUp allows you to specify combinations of components (up to sixteen simultaneously).

What does this do for you? Assume you were deduping with the following conditions (the
number in parenthesis is the size of each component):

Condition #1 Zip/PC (5) + Last Name (5) + Street # (4) + Street Name (4)
Condition #2 Zip/PC (5) + Last Name (5) + PO Box (10)
Here are three example records:

Name: Joe Smith Suzi Smith Billy Bob Smith

Address 1: 326 Main Street PO Box 11086 PO Box 11086
Address 2: 326 Main St.
City/State/PC: Pembroke, MA 02066 Pembroke, MA 02066 Pembroke MA 02066
Here are the matchcode keys that would be derived using the matchcodes listed above:

Condition #1 Condition #2
Joe's Keys: 02066SMITH326 MAIN 02066SMITH
Suzi's Keys: 02066SMITH326 MAIN 02066SMITH 11086
Billy Bob's Keys: 02066SMITH 02066SMITH 11086
Your garden-variety deduplicator would never match these three records - the matchcodes
are close, but not exact. The type of information available from each record is different.

With MatchUp, Joe and Suzi match using condition #1, Suzi and Billy Bob match using
condition #2. Billy Bob is considered a match to Joe through inferred matching (Billy Bob
matches Suzi who matches Joe). Inferred matching is an important part of MatchUp's
matching technology.

Concepts 25
2.2 Matchcode Components
The following are the matchcode components at your disposal:

Prefix A name's prefix.

First Name A name's first name.
Middle Name A name's middle name.
Last Name A name's last name.
Suffix A name's suffix.
Gender The sex of the name (Male/Female/Neutral).
First/Nickname A representative nickname for a First Name.
Middle/Nickname A representative nickname for a Middle Name.
Department/Title A title and/or department. (Note 1)
Company A company name. (Note 1)
Company Acronym A company's acronym. (Note 1, Note 2)
Street Number An address's street number. (Note 3)
Street Pre-Directional "South" in "3 South Main St". (Note 3)
Street Name An address's street name. (Note 3)
Street Suffix An address's suffix. (Note 3)
Street Post-Directional "North" in "3 Main St North". (Note 3)
PO Box PO Boxes also include Farm Routes, Rural Routes, etc. (Note 3)
Street Secondary Apartments, floors, rooms, etc. (Note 3)
Address A single unsplit address line. (Note 4)
City A city. Zip/Postal Code is usually more accurate.
State/Province A state or province.
Zip9 A full Zip+4 code (9 digits). (Note 5)
Zip5 The Zip code (5 digits). (Note 5)
Zip4 The +4 extension of a Zip+4 code (4 digits). (Note 5)
Validated Zip9 A full Zip+4 code (9 digits) that has been CASS-verified. (Note 9)
Validated Zip5 The CASS-verified Zip code (5 digits). (Note 9)
Validated Zip4 The CASS-verified +4 extension of a Zip+4 code (4 digits). (Note 9)
Postal Code (Canada) A Canadian Postal Code. (Note 5)
City (UK) A city in the United Kingdom.
County (UK) A county in the United Kingdom.
Postcode (UK) A United Kingdom Postcode.
Country A country. (Note 6)
Phone/Fax A phone number. (Note 7)
E-Mail Address An e-mail address. (Note 8)
Credit Card Number A credit card number.
General Any general information. ID, birthday, SSN, etc.
Custom General information to be processed with a user table. (Note 10)

26 MatchUp
Notes:
1. Company, Company Acronym, Department/Title Frequently these components don't
match exactly because of 'noise words' such as "the", "and", "agency", etc. MatchUp
strips these words from these components. The lookup tables that contain these lists are
the Company table and the Title/Department table and you can edit them to meet your
needs.
2. Company Acronym MatchUp converts any multi-word company name into an acronym
(for example, "International Business Machines" is squeezed into "IBM"). Single-word
company names are left as they are. This conversion is done after noise words are
removed.
3. Street Address Components he seven street address components (Street Number,
Street Pre-Directional, Street Name, Street Suffix, Street Post-Directional, PO Box,
Street Secondary) are obtained by splitting up to three address lines. Note that PO Box
and/or Street Secondary do not have to appear on their own line, or in a particular field.
MatchUp's proprietary "street smart" splitter does all the work for you.
4. Full Address When using the Full Address component, you are at the mercy of every
little deviation in data entry. Because MatchUp's street splitter is so powerful, you will
want to use street address components instead of the Full Address in nearly all cases.
The only exception may be when processing foreign addresses that don't conform very
well to US and Canadian (or optionally UK) addressing formats. This is discussed in
more detail in the Advanced Matchcodes section.
5. Zip9, Zip5, Zip4, Canadian Postal Code MatchUp removes dashes and spaces from
Zip codes. If you are processing a mix of Canadian Postal Codes and US Zip Codes, use
the Zip9 component.
6. Country Countries are verified and corrected with the Country lookup table.
7. Phone Number MatchUp removes non-numeric characters from phone numbers.
Leading "1-" and trailing extensions are stripped if present. Numbers lacking an area
code are right justified so that the local dialing code and number are aligned with
numbers having area codes. If your table often has missing or inaccurate area codes
(i.e., after a recent area code split), you will probably want to start at the 4th position of
the phone number component. Do not use the rightmost 7 positions, as badly formatted
extensions can sometimes cause the phone number to get coded improperly.
8. E-Mail Address MatchUp removes illegal characters from e-mail addresses. Incomplete,
changed, and commonly misspelled domain names are corrected using the E-Mail
lookup table.
9. Validated Zip9, Validated Zip5, Validated Zip4 Whenever the Validated Zip9, Validated
Zip5, or Validated Zip4 components are used in a matchcode, MatchUp will use CASS-
verified results for Street Number, Street Pre-Directional, Street Name, Street Suffix,
Street Post-Directional, PO Box, and Street Secondary. If an address does not CASS
verify, MatchUp falls back to using the record's original Zip code, and splits the street
components as it would for a non-validated matchcode. Although it is not common, it is
possible for a non-validated address to match a validated address; this procedure
ensures the best chance of catching such an event.
10. Custom Custom component allows you to use a search and replace table to make
substitutions as a matchcode is being built. For example, here in Boston, a local bank
has been bought and sold several times, and is currently Bank of America. If you were
matching banking executives, "Joe President" at "Fleet Bank" would not match "Joe

Concepts 27
President" at "Shawmut Bank", "Bank of Boston", or "BancBoston". A short custom
lookup table containing these names (and replacing them with "Bank of America") would
take care of this bank's identity crisis. When a custom component is specified in the
Matchcode Editor, you must also specify a Custom File, which will be used for this
purpose.

28 MatchUp
Concepts 29
2.3 Component Properties
Every component in a matchcode has its own set of properties.

Size:
The number of characters that will be used. Excess characters are trimmed from the right.
This trimming is done after all other processing. For example, if MatchUp is generating a size
4 Street Name component and encountered "123 Everest Street", it would first split the
address to produce a street name of "Everest", and then extract the first 4 characters.

Label:
A label that is attached to this component. MatchUp does not use this label, but it can be
helpful in remembering what a particular General component was for.

Maximum number of words:

The maximum number of words that will be used. Like the Size property, the removal of
excess words is done after all other processing (except Size).

Start:
Where character extraction should take place. Choices are:
· Left From the left end of the data (most common choice).
· Right From the right end of the data.
· Position From a particular character position in the data.
· Word From a particular word in the data.

Trim:
Whether leading and/or trailing white space should be stripped from the data. You nearly
always want this option turned on for both leading (left) and trailing (right).

Fuzzy:
Fuzzy settings allow for matching of non-exact components. The following settings are
available are mutually exclusive (i.e., you can only select one at a time):
· Phonetex (pronounced "Fo-NEH-tex") An auditory matching algorithm. It works best in
matching words that sound alike but are spelled differently. It is an improvement over
the Soundex algorithm described below.
· Soundex An auditory matching algorithm originally developed by the Department of
Immigration in 1917 and later adopted by the USPS. Although the Phonetex algorithm
is vastly superior, the Soundex algorithm is presented for users who need to create a
matchcode that emulates one from another application.
· Containment Match when one record's component is contained in another record. For
example, "no" is contained in "innovation".

30 MatchUp
 · Frequency Match the characters in one record's component to the characters in
another without any regard to the sequence. For example "abcdef" would match
"badcfe".
· Fast Near A typographical matching algorithm. It works best in matching words that
don't match because of a few typographical errors. Exactly how many errors is
specified on a scale from 1 to 4 (1 being the tightest). The Fast Near algorithm is a
speedy approximation of the Accurate Near algorithm described below. The tradeoff
for speed is accuracy; sometimes Fast Near will find false matches or miss true
matches.
· Accurate Near A typographical matching algorithm. The Accurate Near algorithm
produces better results than the Fast Near algorithm, but at the cost of speed.
· Frequency Near Like Frequency matching except that a slider lets you specify how
many characters may be different between the components.
· Vowels Only Only vowels will be compared. Consonants will be ignored.
· Consonants Only Only consonants will be compared. Vowels will be ignored.
· Alphas Only Only alphabetic characters will be compared.
· Numerics Only Only numeric characters will be compared.

If the mutually exclusive requirement is too restrictive for you to bear, check the Advanced
Matchcodes topic for a workaround. See Matching Strategy.

Short/Empty:
There are several options available for comparing incomplete data. These options can be
used together (i.e., not mutually exclusive).
· Initial Only Will match a full word to an initial (for example, "J" and "John").
· One Blank Field Will match a full word to no data (for example, "John" and "").
· Both Blank Fields Match this component if both fields are blank. This is a very
important concept in creating matchcodes. See Blank Field Matching for more
information.

Swap Match:
Swap matching is the ability to compare one component to another component. For
example, if you were to swap match a First Name component and a Last Name component,
you could match "John Smith" to "Smith John". Swap matching is always defined for a pair of
components. MatchUp allows you to specify up to 8 swap pairs (named "Pair A" through
"Pair H"). It is strongly recommended that both member components of the swap pair have
the same properties. See Swap Matching for more information.

Concepts 31
Uppercasing:
Components are always converted into uppercase. No exceptions. This upper casing is
internal to MatchUp - we don't change the casing of your data.

Foreign Character Translation:

Foreign characters are translated into English equivalents. For example, "Ç" is converted to
"C". All translations are based on the assumption that your data was entered with the 1252
(Windows Latin 1) code page. If this is not true, you can specify an alternate code page in
Translation Table. Again, this is internal and does not actually change your data.

32 MatchUp
2.4 Matchcode Combinations
As discussed in Matchcodes, MatchUp allows you to specify up to sixteen combinations of
components that define a match. In order for a match to be found, any one of the
combinations must be found to match. Programmer-type users will think of this as a boolean
OR condition. Remembering this fact will keep you out of a lot of trouble with matchcodes.

Specifying Matchcode Combinations:

You specify matchcode combinations in a columnar format. For example, say we wanted to
create a matchcode that matched on the following conditions (this is the same example as
found in the Matchcodes topic):

Condition #1 Zip/PC (5) + Last Name (5) + Street # (4) + Street Name (4)
Condition #2 Zip/PC (5) + Last Name (5) + PO Box (10)
Remember, satisfying condition #1 or condition #2 will constitute a match. We would create
the following matchcode:

Component Size 1 2
Zip/PC 5 X X
Last Name 5 X X
Street # 4 X
Street Name 4 X
PO Box 10 X
For clarity, columns 3 through 16 have been omitted for this example. The trick to
understanding this table is to look at the vertical columns of X's. For example, looking at
column 1, we see X's in Zip/PC, Last Name, Street #, and Street Name which indicates our
desired condition #1 exactly. Looking at column 2, we see X's in Zip/PC, Last Name, and PO
Box matching our desired condition #2.

Let's look at a more advanced example:

Component Size 1 2 3 4
Zip/PC 5 X X X X
Last Name 5 X X
Company 10 X X
Street # 5 X X
Street Name 5 X X
PO Box 10 X X
This matchcode defines four conditions:

Concepts 33
Condition #1 Zip/PC (5) + Last Name (5) + Street # (5) + Street Name (5)
Condition #2 Zip/PC (5) + Last Name (5) + PO Box (10)
Condition #3 Zip/PC (5) + Company (10) + Street # (5) + Street Name (5)
Condition #4 Zip/PC (5) + Company (10) + PO Box (10)
Combinations #1 and #2 are very much like the previous example. Combinations #3 and #4
allow for a different situation, where a Company Name matches instead of Last Name.

First Component Requirements:

MatchUp places some special restrictions on the first component in a given matchcode.
These restrictions are necessary so that MatchUp can properly cluster your record set (see
Concepts). The following must be true for a matchcode's first component:
· It must be used in every condition that you are using.
· It cannot use certain types of Fuzzy Matching: Containment, Frequency, Fast Near,
Frequency Near or Accurate Near (other types are okay though).
· It cannot use Initial Only matching.
· It cannot use One Blank Field matching.
· It cannot use Swap Matching.

When using the Matchcode Editor to create matchcodes, the editor will ensure that these
conditions are met. If you come across a situation that demands that you have no such
component, you can use the workarounds discussed in the Advanced Matchcodes section.

34 MatchUp
Concepts 35
2.5 Blank Field Matching
This needs a special discussion, as its importance is often overlooked. As discussed in the
Component Properties section if this property is on, then the absence of data in both records
would indicate a match. If this property is off, then two records with that missing piece of
data, but matching in every other way, will not match.

Upon reading this, one may wonder why on earth anyone would not desire this behavior.
Actually, you won't for certain situations. Take the following matchcode (keep a close eye on
the Blank column):

Component Size Blank 1 2

Zip/PC 5 Yes X X
Last Name 5 Yes X X
Street # 4 Yes X
Street Name 4 Yes X
PO Box 10 Yes X
Using what we've learned about Matchcode Combinations, we can derive the matchcode
combinations:

Condition #1 Zip/PC (5) + Last Name (5) + Street # (4) + Street Name (4)
Condition #2 Zip/PC (5) + Last Name (5) + PO Box (10)
Using the following records as an example:

Name: Joe Smith Suzi Smith

Address: 326 Main Street 405 Main St
City/State/PC: Pembroke, MA 02359 Pembroke, MA 02359
The following matchcode keys would be generated:

Zip/PC Last Name Street # Street Name PO Box

1 02359 SMITH 326 MAIN
2 02359 SMITH 405 MAIN
Looking at these matchcode keys, it is clear that this pair do not satisfy condition #1. But
because blank field matching is selected, they do satisfy condition #2. The Zip/PC, Last
Name, and PO Box are exactly the same. Therefore the two records do match.

36 MatchUp
Obviously, this is wrong. Lets just make one change to the matchcode:

Component Size Blank 1 2

Zip/PC 5 Yes X X
Last Name 5 Yes X X
Street # 4 Yes X
Street Name 4 Yes X
PO Box 10 No X
The same comparison is done for combination #2, but the match is disallowed this time
because we have indicated that missing (blank) information is not allowed to figure in the
matching condition. Looking at another example (using the same matchcode):

Name: Joe Smith Suzi Smith

Address: PO Box 123 PO Box 456
City/State/PC: Pembroke, MA 02359 Pembroke, MA 02359
We get the following matchcode keys:

Zip/PC Last Name Street # Street Name PO Box

1 02359 SMITH 123
2 02359 SMITH 456
We've got the same problem as before, but this time its combination #1 which is the culprit.
So an even better matchcode would be:

Component Size Blank 1 2

Zip/PC 5 Yes X X
Last Name 5 Yes X X
Street # 4 No X
Street Name 4 No X
PO Box 10 No X
Now what we have here is a matchcode that works pretty well. There's one more tweak that
you might want to do, however - turn on Both Blank Fields for Street #. Occasionally, you
may come across records such as:

Name: Joe Notarangello Suzi Notarangello

Address: Oceanfront Estates Oceanfront Est.
City/State/PC: Pembroke, MA 02359 Pembroke, MA 02359
Not to make any wild generalizations, but this seems to be a trend in your up-scale

Concepts 37
neighborhoods. Neither street address has a Street # component, though it is very likely
these records should match.

So our new improved matchcode will do the trick:

Component Size Blank 1 2

Zip/PC 5 Yes X X
Last Name 5 Yes X X
Street # 4 Yes X
Street Name 4 No X
PO Box 10 No X
Because Street Name still has Both Blank Fields off, we are protected from the accidental
combination #1 match. The above matchcode (with or without Last Name) is the basis for
nearly all of MatchUp's canned matchcodes.

38 MatchUp
Concepts 39
2.6 Matchcode Mapping
The components in a MatchUp matchcode represent specific types of data, and they aren't
directly linked to the fields in your databases. Mapping creates the link between your data
and the matchcode.

This may seem like an extra step, but in fact, it's a great time save. Because the matchcodes
are abstract, they can be used with many different files and field structures. This saves you
the time of recreating them for each new job.

Say you used this matchcode (Address, Last Name):

Component Size Fuzzy 1 2

Zip5 5 No X X
Last Name 5 No X X
Street # 5 No X
Street Name 10 No X
PO Box 10 No X
And you are working with a table which contains the following fields:

NAME Contains full names ("Mr. John Smith").

ADD1 Contains first (primary) address line ("50 Cole Parkway").
ADD2 Contains second (secondary) address line ("Suite 34").
CSZ Contains City/State/Zip ("Scituate, MA 02066").
You need to create the link between your table's fields (NAME, ADD1, ADD2 and CSZ) and
the matchcode components (Zip5, Last Name, Street #, Street Name, PO Box).

40 MatchUp
On the Matchcode Mapping tab, you would do the following (actually MatchUp will probably
default intelligently and do it for you):

This mapping tells MatchUp that the 5-digit zip code information is in a field named CSZ.
The Last Name can be found in a field called NAME.

The purple background indicates that some sort of conversion is to take place. If you
highlight a mapped field, the "Conversion" text (bottom right) tells you what conversion is
needed. In our example, a "CSZ to Zip 5" conversion for the CSZ to Zip 5 mapping, "Full to
LN" conversion for NAME to Last Name, and an "Address Split" for STR and STR2.

Matchcode mappings follow three rules:

1. For every Matchcode Component, you should specify a mapping. The only exception
is specified in Rule 2.
2. Actual Address components names (such as Street Number, Street Pre-Directional,
Street Name, Street Suffix, Street Post-Directional, PO Box, and Street Secondary)

Concepts 41
 are not listed for mapping purposes. Instead the names Address Line 1, Address Line
2, and Address Line 3 are used. In the example above, we've used three address
components in the matchcode (Street #, Street Name, PO Box), however we've only
used two (of possible three) address line.
3. If you use address components in your matchcode Address Lines 1-3 will require at
least one line to be mapped, but not all. If you only have one address field in your
database, you will only need to map that field to Address Line 1.

Note on Rule #1:

Sometimes people have a bit of a problem with Rule #1. Say for example, your database
doesn't have a name field. What does one map to the Name Component in that case? You
have no business using this matchcode. Don't use matchcodes that you haven't got the data
to support. Another matchcode should be created or selected that better suits your data. Of
course, you don't have to have exactly what the matchcode calls for (if MatchUp can convert
your data to what it needs, you're okay).

42 MatchUp
Concepts 43
2.7 Optimizing Matchcodes
Some matchcodes process much faster than others in spite of the fact that they detect the
same matches. This section will help you to create the most efficient matchcode. 99 times
out of 100, clicking the Optimize button in the Matchcode Editor will optimize your
matchcode sufficiently. But we've included this discussion so that you can better understand
why certain things were done while optimizing, as well as what you can do to make the
optimizer work better.

Optimizing can make a huge difference in processing speed. We've actually seen 58 hour
runs reduced to 4 hours simply by using matchcode optimizations.

It is important that you get your matchcode working the way you want before attempting any
of these optimizations. If your matchcode is not functioning properly, these optimizations will
not help, and could quite possibly make the situation worse.

Component Sequence:
As discussed in the Matchcode Combinations section, the first component of a matchcode
has certain restrictions:
· It must be used in every combination.
· It cannot use certain types of Fuzzy Matching: Containment, Frequency, Fast Near,
Frequency Near or Accurate Near (other types are okay, though).
· It cannot use Initial Only matching.
· It cannot use One Blank Field matching.
· It cannot use Swap matching.

If the matchcode's second component also follows these conditions, MatchUp will
incorporate it into its clustering scheme (clustering is discussed in Concepts). Furthermore,
third, fourth, etc. components will be used if they too satisfy these conditions. Incorporating a
component into a cluster greatly reduces the number of comparisons MatchUp has to
perform which, in turn, speeds up your processing.

44 MatchUp
Your existing matchcode(s) may be only a few easy steps away from this optimization. For
example:

Component Size Fuzzy Blank 1 2

Zip/PC 5 No Yes X X
Street # 5 No Yes X
Street Name 5 No No X
PO Box 10 No No X
Last Name 5 No Yes X X
As shown here, MatchUp will only cluster by Zip/PC. But note that the last component (Last
Name) satisfies all the conditions listed earlier.

If you move it to the second position, it too will be used by MatchUp in a cluster:

Component Size Fuzzy Blank 1 2

Zip/PC 5 No Yes X X
Last Name 5 No Yes X X
Street # 5 No Yes X
Street Name 5 No No X
PO Box 7 No No X
That's all there is to it; instant speed. The Matchcode Optimizer will always perform this
optimization.

Fuzzy Algorithms:
Fuzzy algorithms fall into two categories: early matching and late matching.

Early matching algorithms are algorithms where a string is transformed into a (usually
shorter) representation and comparisons are performed on this result. In MatchUp, these
transformations are performed during key generation, which means that early matching
algorithms pay a speed penalty once per record: as the record's key is built.

Late matching algorithms are actual comparison algorithms. Usually one string is shifted in
one direction or another, and often a matrix of some sort is used to derive a result. In
MatchUp, these operations are performed during key comparison. As a result, late matching
algorithms pay a speed penalty every time a record is compared to another record. This may
happen several hundred times per record.

Concepts 45
Obviously, late matching is much slower than early matching. If you find that a particular
matchcode is very slow, changing to a faster fuzzy matching algorithm may be the way to go.
Often, a faster algorithm will give nearly the same results; but, as always, you should test this
before processing live data.

The fuzzy algorithms, ranked from slowest to fastest:

Algorithm Late or Early Speed (10=fastest)

Accurate Near Late 1
Fast Near Late 3
Containment Late 4
Frequency Near Late 4
Frequency Late 6
Phonetex Early 7
Soundex Early 8
Vowels Only Early 9
Numerics Only Early 9
Consonants Only Early 9
Alphas Only Early 9
Exact N/A 10
The speed values are very rough estimates.

Another benefit of using a faster fuzzy algorithm is that you may be able to exploit the
Component Sequence optimization listed earlier. All of the early matching algorithms satisfy
the restrictions listed, so they are fair game.

The Matchcode Optimizer will not perform this optimization as it can have a drastic impact on
your matching results.

Unnecessary Components:
Components that aren't used in any combinations (i.e., have no X's in columns 1 through 16)
are a sign of bad matchcode design. If you find yourself with a matchcode having
components like this, you should be asking yourself why. There is one exception, an
advanced technique, explained in the Advanced Matchcodes section. Because of this one
exception, the Matchcode Optimizer will not automatically remove unnecessary components.

46 MatchUp
For example, in the matchcode:

Component Size Fuzzy Blank 1 2

Zip/PC 5 No Yes X X
Last Name 5 No Yes X X
First Name 5 No Yes
Street # 5 No Yes X
Street Name 5 No No X
PO Box 10 No No X
First name is not being used in any combination. Maybe it was used in an earlier iteration of
this matchcode, but not any more.

Unnecessary Combinations:
Take a look at the following matchcode:

Component Size Fuzzy Blank 1 2 3 4

Zip/PC 5 No Yes X X X X
Last Name 5 No Yes X X X X
First Name 5 No Yes X X
Street # 5 No Yes X X
Street Name 5 No No X X
PO Box 10 No No X X
When a user calls our tech support with such a matchcode, we know we're in for a long call.
This is usually a warning sign that the user doesn't quite understand the concept of
Matchcode Combinations. What's wrong?

Here are the four conditions for matching:

Condition #1 Zip/PC (5) + Last Name (5) + First Name (5) + Street # (5) + Street Name (5)
Condition #2 Zip/PC (5) + Last Name (5) + First Name (5) + PO Box (10)
Condition #3 Zip/PC (5) + Last Name (5) + Street # (5) + Street Name (5)
Condition #4 Zip/PC (5) + Last Name (5) + PO Box (10)
There is absolutely no match that will be detected by condition #1 that won't be detected by
condition #3. Similarly, matches found by condition #2 will always be found by condition #4.
In other words, condition 3 is a subset of condition 1, and condition 4 is a subset of condition
2. Subsets are generally bad.

Concepts 47
So either conditions 1 and 2 aren't needed or conditions 3 and 4 were a mistake. If you do
eliminate conditions 1 and 2, you probably will want to remove the First Name component, as
it won't be needed (discussed above).

There is one exception to this rule, discussed in the Advanced Matchcodes section. The
Matchcode Optimizer will perform this operation (unless Advanced Matchcode Settings are
used).

48 MatchUp
Concepts 49
2.8 Other Uses for Swap Matching

Swap matching is used to catch matches when two fields are flipped around. The most
common occasion is catching the "John Smith" and "Smith John" records. But there are other
uses:

· Comparing household records, where there's two or three first or full names. Although
the list provider will tell you differently, you know you can't always rely on the
convention used (i.e., they'll say "It's always husband, wife, then children", but the first
record will read wife, child, husband):

In the above example you would want to select Either component can match for
Swap Pairs A, B, and C. See Swap Match Pairs for more information.

· Comparing up to three address lines. Although the address splitter works well in the
US and Canada, some European countries can cause problems. A typical Euro-
Matchcode will not use street split components and look at three address lines instead:
The swap matching ensures that every address line is compared with every other
address line.

Like the previous example you would want to select Either component can match
for Swap Pairs A, B, and C. See Swap Match Pairs for more information.

50 MatchUp
· Don't discard the street split component matchcodes just because you're working on a
foreign database, however. Sometimes the street splitter will yield usable results.
Often, a combination will do the trick:

Concepts 51
52 MatchUp
2.9 Gathering and Scattering
2.9.1 Gathering

When duplicates are found, MatchUp's basic merge/purge operation can be described in
these steps:
1. Group the block of duplicate records into one "Dupe Group".
2. Using the ranking method you specified on the General tab, determine one of the
records to be the "Output Record".
3. Copy the output record to the output table.
4. Copy the remaining records in the dupe group to the dupe table (if a dupe table was
specified).

Note step 3:
Only the contents of output record are copied to the output table. Which means that the
quality of your output database is completely controlled by the specified ranking method.

In some cases, this is not enough. For example, you may have one database with up-to-date
addresses, and a second database with new phone numbers, but bad addresses. Obviously,
you will prioritize the first list over the second list. But what do you do about those phone
numbers?

This is where gathering comes in. Gathering allows you to consolidate data from records
other than the output record. To specify gathering for an output field, you need to do three
things:
1. On the Output Field Mapping tab, highlight the output field that will receive the
consolidated information. Check the "Gather" box in the "Output Field" box (bottom
left). This activates gathering for this particular output field.
2. Select the desired gathering method from the drop down list to the right of the "Gather"
box. The details about the various gathering methods are described below.
3. Highlight the input field that will contribute to the consolidated information. Check the
"Gather" box in the "Input Field" box (bottom right). This activates gathering
contribution from this particular input field. If you want several input fields (i.e., from
other tables) to also contribute, you must highlight each one and check the "Gather"
box each time.

Concepts 53
2.9.2 Gathering Methods

There are many different ways of consolidating information:

First Data:
MatchUp scans through the Dupe Group and takes the first field it comes across that
contains data. This may not necessarily be the output record. If you are using the First Data
method on several output fields, the consolidated data will not necessarily come from the
same record - each output field is considered individually. This is a very common method of
consolidation and is available for all output field types (character, date, logical, and numeric).

Join:
MatchUp joins together all of the data contributed into a single string. Each piece of data is
not separated by a space; leading and trailing spaces are stripped from the incoming data.
Note that the size of the output field should be adjusted to accommodate this joining of data.
This method of consolidation is sometimes used when generating a multi-buyer field. Join is
only available for character fields.

Join with Spaces:

Same as Join above, but each piece of data is separated by a space.

Longest:
MatchUp scans the group of duplicates and takes the longest data string that it comes
across. Leading and trailing spaces are not considered as part of the size. The Longest
method is rarely used, but it can be used as a simple "take the field with the most data"
selection. Longest is only available for character fields.

Shortest:
The opposite of Longest above.

Maximum Value:
MatchUp scans the group of duplicates and takes the field with the largest value. For
transactional databases, this is a good way to determine a maximum purchase value.
Maximum is available for character and numeric fields.

Minimum Value:
Same as Maximum, above, but takes the field with the smallest value.

Add Values:
The value in each contributing field is added together to obtain a total result. For
transactional databases, this is an excellent way to determine total sale values and the like.
Add is only available for numeric fields.

54 MatchUp
Earliest Date:
The field with the earliest date is selected for output.

Latest Date:
Just like Earliest, above, but takes the field with the latest date. For transactional databases,
this is a good way to obtain "last called" or "last updated" information. Latest is only available
for date fields.

AND Values:
Logically "ands" the fields together. This is a very rarely used method, only available for
logical fields.

OR Values:
Logically "ors" the fields together. This is a very rarely used method, only available for logical
fields.

Stack Group (A-H):

There are 8 stack groups (letters A through H). Each letter represents a series of output
fields that MatchUp will incrementally fill. For example, you may have output fields FIRST1,
FIRST2, FIRST3, and FIRST4 each indicated as Stack Group A, and LAST1, LAST2,
LAST3, and LAST4 each indicated as Stack Group B. MatchUp will take the first name
information from the output record and deposit it into FIRST1, the last name information into
LAST1. The next record's first and last name information will be deposited into FIRST2 and
LAST2, the next record into FIRST3 and LAST3, and so on. MatchUp will do this for as many
fields in a series as you can specify. When it runs out of fields in a series, the data is
discarded. You don't necessarily need to have the same number of output fields in each
series (though that is the most common configuration). Be sure to map each source field to
each output field (i.e., FIRST to FIRST1, FIRST2, FIRST3 and FIRST4). Also, remember to
check "Gather" for each mapped field (input and output).

Fields in a Stack Series don't necessarily have to be adjacent to one another. Stack Fields
are populated in the order they appear in "Output Field Mapping", even if they are
numerically out of order. That is, if you've specified FIRST3, FIRST2, and FIRST1 in that
order, FIRST3 will be populated before FIRST2. You should drag and drop the output fields
into a more suitable sequence.

Concepts 55
2.9.3 Scattering

Scattering is the process of taking the information in the output record and re-distributing it
back into your source file(s). Like gathering, you have complete control over which source
files and fields should receive the new information. Gathering and scattering are two
independent processes. You don't need to gather to scatter and vice versa, though in most
cases, you will find the two used in concert. Selecting an input field for scattering involves
one step:
1. Highlight the input field that will receive data from the output record. Check the
"Scatter" box in the "Input Field" box (bottom right). This activates scattering
redistribution to this particular input field. If you want several input fields (i.e., from
other files) to receive data, you must highlight each one and check the "Scatter" box.
In some cases, you may want to gather from one field, but scatter into another. For these
cases, you can specify that field in "Alternate Scatter Field". Data will be gathered from the
mapped field, but scattered to the specified Alternate Field.

56 MatchUp
2.10 Dupe Group
Every original record encountered by MatchUp gets a unique number. When a duplicate
record is found, it is assigned the same number as the original. This number is known as the
dupe group.

MatchUp reports this. Its uses include:

· A way for you to connect a 'parent' record with its 'child' records. - also creating a
group identifier
· A way to sort your data which will ensure that duplicate records are grouped together.
· An easy way to see if similar records fell into the same group when browsing duplicate
tables.

Concepts 57
58 MatchUp
2.11 Canadian Users
MatchUp recognizes Canadian provinces and postal codes. In fact, it will abbreviate province
names to their two letter abbreviation automatically. Here's a couple notes for the benefit of
our neighbors to the north…

MatchUp does handle the "QC" province abbreviation for Quebec, and "PQ" entries are
automatically changed to "QC".

In Canada, "5-20 Main Street" means "20 Main Street, Apt 5", but in the US, it means "5
Main Street, Apt 20". When deduping, MatchUp uses the contents of the zip/postal code as a
basis to determine a record's country of origin, and splits this type of address accordingly.

When creating matchcodes for use with Canadian Postal Codes, you should use the Postal
Code component. However, if your file is a mix of US and Canadian records, use Zip9 as the
component type. Don't worry. Zip9 will not adversely affect processing of Canadian records.
The goal here is to prevent the deduper from trying to extract a Zip +4 from a Canadian
Postal Code.

Concepts 59
2.12 United Kingdom Users
MatchUp can recognize United Kingdom Cities, Counties, and Postcodes. When creating
matchcodes for use with United Kingdom addresses, you should use the Postcode (UK)
component. Depending on your requirements, you also may want to use the City (UK) and
County (UK) components. The Postcode component is structured in the following format:
AADDIII, where AA is the Postcode Area (left justified), DD is the Postcode district (right
justified), and III is the Inward Code (left justified). Extra spaces and dashes are removed as
this structuring is done, so the size of this component is always 8.

Like any other matchcode component, you can compare a portion of the Postcode by
reducing it's size and/or starting at a specific position. For example, starting at position 5 for
a size of 3 will compare just the Inward code.

MatchUp's street splitter will not split United Kingdom street addresses as well as Canadian
and US addresses. Usually, a matchcode containing a mix of split address components and
full address components is a good way to get the benefit of the street splitter (which often
does perform well), along with a full-address match backup. We have included the United
Kingdom Address matchcode to be used as a starting point for you to build on.

60 MatchUp
Our United Kingdom users are welcome to contribute any comments, matchcodes and data
that would improve MatchUp's processing. Our exposure to United Kingdom addresses isn't
like it is to US and Canadian addresses, so any help from more experienced users is
appreciated.

Concepts 61
2.13 International records
MatchUp was designed to work with US and Canadian addresses, and it does a pretty good
job with addresses from other English speaking countries.

The big problem with international records is with the Street Splitter. Try doing a test run with
one of the "canned" matchcodes. If the street splits are not working, use the full address
when creating a matchcode instead of using the components (such as street number, street
name, etc.).

Often users have had great success when combining the full address and street splitter. For
example, here's an international version of one of our supplied matchcodes:

Component Size Start Fuzzy Short/Empty 1 2 3

General 10 Left No Both Empty X X X
Last Name 5 Left No Both Empty X X X
First Name 3 Left No Both Empty X X X
PO Box 7 Left No No X
Street # 4 Left No Both Empty X
Street Name 4 Left No No X
Full Address 20 Left No No X

62 MatchUp
3 File Menu
3.1 New Setup

New Setup lets you select which process you want to use on your tables.

File Menu 63
3.2 Open Setup

Once you open a setup, you can edit, process, print, or analyze the results.

When you create a setup, MatchUp saves the setup details in a file with a .dt extension.
Table locations, field names, matchcode details, etc. are all stored in the setup file.

64 MatchUp
3.3 Save Setup

Use File | Save Setup to save a setup that you have created or modified. If you have never
saved this setup before, you will be prompted for a file name. You can give a setup a new
name by using Save Setup As. The settings you save are available for future reuse through
Open Setup.

File Menu 65
3.4 Save Setup As

Use File | Save Setup As to save a setup with a new file name:

This allows you to slightly modify an existing setup without losing the old setup and without
having to completely re-enter a new setup.

66 MatchUp
3.5 Exit

Use File | Exit to exit MatchUp. If you've made changes to the setup, you will be prompted to
save the changes.

File Menu 67
4 Setup
4.1 Merge Setup

Select Merge if you want to take several tables and put them into a single table. A merge will
not detect nor eliminate duplicate records.

4.1.1 Input Tables

This screen displays the tables you have selected and each table's format type (dBASE,
Excel Access, ASCII Delimited, etc.).

You can change the placement of an input table by clicking and dragging it into the desired
position. Alternately, you can press CTRL+UP ARROW or CTRL+DOWN ARROW to move
the table. Note that the order of input tables has no impact on processing, but is purely for
your benefit.

68 MatchUp
Add... Allows you to add an additional table to the setup.

Remove Removes a table from the merge setup.

Change... Allows you to replace a table in the merge setup with another table.

View Although you cannot edit the data, you can view your table here.

Input Tables: Information about

Input Filter (optional) Enter the input filtering criteria for each source table. MatchUp will
allow you to use a different filter for each input table. The filter must be entered in dBASE
syntax. Instead of typing in a dBASE expression, or if you're unfamiliar with dBASE syntax,
you can use the Expression Builder.

Table is Read-Only This warning message will appear if MatchUp will not be able to write
any data back to the table.

ASCII File (Cannot be modified) This warning message will appear if the Input Table is
ASCII - MatchUp cannot write results back to ASCII tables.

Setup 69
4.1.2 Output Field Mapping

The initial Output table's structure is determined in Default Output Table. You can modify the
structure here, but it is easier to start with one that is close to what you want for your output
table.

This screen tells MatchUp how it should go about mapping input data into the desired output
structure. When you create a setup, MatchUp will try to determine the field mappings for you.
Usually it does a pretty good job. If it doesn't, simply double-click on the incorrectly mapped
field and select the correct field. Alternately, you can right-click on the field and select the
field from a drop-down menu. If your field naming conventions are different from ours, see
Field Naming.

70 MatchUp
In addition to mapping fields, MatchUp determines if a conversion is needed to get the
correct data type from the input field to the desired output field. For example, if an input field
contains full names ("Mr. John Smith"), but the desired output fields are First Name and Last
Name (as two separate fields), MatchUp will perform a name split. On this screen, this action
is indicated in the "Conversion" prompt on the bottom right. If the Conversion is incorrect, you
can click Change Data Type to select a more appropriate data type for the input field.
Alternately, you can right-click on the field.

Note that information sometimes has to be extracted from complex fields. For example,
under national.dbf above, First and Last will be extracted from Name.

Add Output Field Adds a new field(s) to the output table's structure.

Remove Output Field Selects a field(s) to remove from the output structure.

Check Mappings Ensures that you haven't forgotten any source table's fields.

Check Truncations MatchUp warns you about the possibility of truncation here.

View File Although you cannot edit the data, you can view your table here. This is a great
way to make sure you have the right table!

Field List Shows a list of each field in each table. You can use your mouse to drag and drop
fields on this list into the desired mappings on the Output Field Mapping Screen.

Setup 71
Output Field Mapping: Output Field

Describes the currently highlighted output field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly.

Change Output Field Allows you to change all the specifications of a field: Name, Field
Type, Size, Decimals, and Data Type.

72 MatchUp
Output Field Mapping: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Data Type and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion Indicates what type of conversion will be done to get from the highlighted field's
data type to the output field's data type. For example, an input field containing a full name will
have to be split if the output table has first name and last name fields.

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

Setup 73
Output Field Mapping: Add Output Field

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field. Some Field Types have fixed sizes.

Decimals If the field type is numeric there might be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR".

74 MatchUp
4.1.3 Output Table

Output Table Specify the table that will receive the merged records. You can use the
Browse button for the easiest way to select an output table.

Source Code Field: (Advanced) (Optional) Enter the field to be used for source tracking.
Source tracking gives you a way to track what table (and optionally record number) the data
in the output table came from. If you like, you can name a new field, and it will be added to
the table. Table Information and Record Information allow you to format the information in the
Source Code field.

Table Information Select how you would like to indicate which table a record came from:
· Table Name Uses up to 8 characters of the input table's name.
· Table Number Uses up to 2 digits to store the input table number.
· Table Letter Uses 1 letter to indicate the input table.

Setup 75
76 MatchUp
Record Information Select how you would like to indicate which record a record came from:
· Long numbers Uses up to 8 digits to store the input record number.
· Medium numbers Uses up to 6 digits to store the input record number.
· Short numbers Uses up to 4 digits to store the input record number.
· "0" Fill Fills the unused record information spaces with "0"s.
· Left Justify Left justifies the record number.
· Right Justify Right justifies the record number. If you select Right Justify, you also
have the option to "0" Fill.

Sample Shows you what a typical Source Code Field will look like.

Setup 77
78 MatchUp
4.2 Merge/Purge Setup

Select Merge/Purge if you want to take several tables, remove matching records, and put
them into a single table.

4.2.1 General

Setup 79
General: Matchcode

Select the matchcode that best suits your needs. A selection is supplied, and you will
probably find a suitable one from this collection. If not, you can click Matchcode Editor to
create (or edit) a matchcode.

The table above lists the properties of the currently selected matchcode. It is for viewing
purposes only.

Matching Rules: MatchUp uses up to 16 matchcode "combinations" which are evaluated

simultaneously. In the table, the X's in the columns marked 1 through 16 define component
combinations that should be considered a match. The Matching Rules section simply
displays them in an alternate form.

80 MatchUp
General: Ranking

Ranking Allows you to choose which records to favor when dupes are found:
· MatchUp will select the best record Take the record with the most complete
information.
· I will assign priorities or specify a priority field Allows you to choose which table or
records to favor when duplicate records are found. If you select this option, you will be
able to assign a rank or rank field for each input table on the Input Tables tab.
· Prioritization will be random for each record MatchUp will randomly select a record
to use as the best record when duplicate records are found.
· A uniform distribution of records will be selected Instructs MatchUp to try (as best
it can) to assign output records uniformly from all input files.
· Doesn't Matter Take the first record you come across.

What's the difference between Random and Uniform?

Random ranking tends to take an appropriate proportion of output records from each source.
Uniform Distribution tries to take an equal quantity of output records from each source. For
example, if there were 100 output records:

Total Records Random Uniform

File1 9,000 90 50
File2 1,000 10 50
Ascending MatchUp will select records with a lower rank over records with a higher rank
(i.e., "A" over "B", "1" over "2", and "03/10/68" over "04/10/68").

Descending MatchUp will select records with a higher rank over records with a lower rank
(i.e., "C" over "B", "3" over "2", and "05/10/68" over "04/10/68").

Setup 81
When matching records have the same rank If MatchUp can't decide which record to use
because they both have the same rank (ie, a tie), a backup plan can be selected:
· The first record processed will be selected
· The last record processed will be selected
· A random record will be selected

82 MatchUp
4.2.2 Input Tables

Input Tables: Input Sources

This screen displays the tables you have added.

You can change the placement of an input table by clicking and dragging it into the desired
position. Alternately, you can press CTRL+UP ARROW or CTRL+DOWN ARROW to change

Setup 83
the table's position. Note that the order of input tables has no impact on processing, but is
purely for your benefit.

Add... Allows you to add an additional table to the setup.

Remove Removes a table from the merge/purge setup.

Change... Allows you to replace a table in the merge/purge setup with another table.

View Although you cannot edit the data, you can view your table here.

Input Tables: Information about

Table Type Select how the highlighted table will be handled:

· Regular This table will be deduped against itself and any other Regular tables.
· Suppression Records in Regular, No Purge, and Self Purge tables which also appear
in a Suppression list will be rejected. Ideal for "do not mail" lists.
· Intersection Records in Regular, No Purge, and Self Purge tables which do not
appear on an Intersect list will be rejected. (opposite of Suppression).
· Self Purge This table will be matched against itself, Suppression, and Intersection
tables, but not against Regular, No Purge, or other Self Purge tables. Self Purge
tables are used when you would like to match several tables against a Suppression
list, and purge internal dupes, but not match the tables against each other.
· No Purge This table will not be deduped against any other table. It will, however, be
matched against Suppression and Intersection tables. No Purge tables are used when
you would like to match several tables against a Suppression list, but not dedupe the
tables internally or against each other.
· Advanced Click if the contents of a field dictate the Table Type.

Alias Enter a name you would like to refer to this table. This name will appear on all reports.

84 MatchUp
Priority Select a ranking method for the highlighted table:
· Assign Priority Enter a rank for this table. This option only appears if you selected I
will assign priorities or specify a priority field under the General tab.
· Contents of Field Enter the name of the rank field. This option only appears if you
selected I will assign priorities or specify a priority field under the General tab. This
field must already exist in your table.
· Advanced Rank Check this option when the values in the specified Field are not
easily compared.
· Custom Expression Enter a valid dBASE expression for evaluating rank. Use it when
you need a tighter criteria than a single field or table will allow. For example, if you
want to mail to the oldest female in a household, enter the expression: SEX +
DtoS(BIRTHDATE). Only appears if you selected I will assign priorities or specify a
priority field under the General tab.

Status Marking You can select methods of marking a record's disposition with these
options:
· Mark rejected records for deletion If checked, rejected records will be marked for
deletion. This can only be done with dBASE tables.
· Status Field (optional) Select a field name for capturing a record's disposition. During
processing, this field will be populated with a code indicating whether this record is a
dupe, has dupes, was suppressed, etc. See MatchUp Status Codes.
· Count Field (optional) Select a field name to receive a record's dupe count. During
processing, this count field will be populated with a number indicating how many
dupes a record has. See Counting Method.

Setup 85
Information about: Advanced Table Type

Sometimes tables are created by a Merge process and contain a code field specifying their
original source. Before MatchUp, to process different source codes as different table types,
users would have to 'undo' the merged file into separate files, then Merge/Purge them. Now
you can specify the table type as Advanced and specify the field containing the code
indicating how the record should be processed.

Base Table Type selection on contents of field The field containing a code indicating how
a record should be processed.

Number of characters to use from field How many characters in the above field should be
used in considering how a record should be processed.

Ignore case Whether or not case should be considered when looking up a code in the lists
below.

Table Type 'Regular' when code is A list of code(s) that indicate that a record should be
processed as a regular record.

86 MatchUp
Table Type 'Suppression' when code is A list of code(s) that indicate that a record should
be processed as a suppression record.

Table Type 'Intersection' when code is A list of code(s) that indicate that a record should
be processed as an intersection record.

Table Type 'Self Purge' when code is A list of code(s) that indicate that a record should be
processed as a self purge record.

Table Type 'No Purge' when code is A list of code(s) that indicate that a record should be
processed as a no purge record.

Any codes not in any of the above lists should be processed Other codes will be
processed in the following manner:
· Not Processed (filtered)
· Table Type 'Regular'
· Table Type 'Suppression'
· Table Type 'Intersection'
· Table Type 'Self Purge'
· Table Type 'No Purge'

Setup 87
Information about: Advanced Rank

When the field chosen to provide the ranking has values which can not easily be compared,
Advanced Ranking can provide an order to those values.

For example, a rank field containing Months would wind up with unusual prioritization if the
values were compared by character strings: Apr, Dec, Feb, Jan, Jul, Jun, etc.

When the list of prioritized ranks is quite long, it can be get tedious to manually enter a list of
ranks in the manner specified above. A more efficient method is to create a two-field lookup
table (dBASE III format) and specify it using the first option, Use an external file listing
ranks and priorities. In our example, the lookup table would look like the following:

88 MatchUp
 Month Priority
Jan 1
Feb 2
Mar 3
Apr 4
May 5
Jun 6
Jul 7
Aug 8
Sep 9
Oct 10
Nov 11
Dec 12

Number of characters to use from field How many characters in the specified rank field
should be used in considering a record's priority.

Ignore case Whether or not case should be considered when looking up a rank in the list
below.

Advanced rank assignment can be done in one of two ways:

· Use an External File If you have a pre-existing dBASE table with the desired ranks, or
have many potential codes - in which case creating a table is a good idea, use this
method. One advantage of this method is you can use the external table in other
setups.
· Rank in This Order Enter a string of codes in ranking order from lowest to highest
priority. When you don't have a lot of different codes in the selected field, or don't want
to bother creating a new dBASE file, use this method.

Codes that don't appear on the above list should be Specify what to do when a record
has a value in the Rank Field which is not present in the above list:
· Not Processed (filtered)
· Assigned the lowest priority
· Assigned the highest priority

Setup 89
4.2.3 Matchcode Mapping

Matchcode Mapping: Field Mapping for Matchcode

This screen tells MatchUp which field contains the data for each component of the
matchcode.

90 MatchUp
These components are matched with fields from your input table. MatchUp tries to determine
a field's type based on the field name. For example, MatchUp will default to "Address" if the
field's name starts with "ADD", or "STR". You can customize these defaults in Field Naming.
If MatchUp doesn't pick correctly, double-click on the incorrect field name and select the
correct field.

The purple field names are fields that need to be parsed to get the needed matchcode
component. So, even though this matchcode is using last name, MatchUp can use a full
name field and find the information it needs.

And, although this matchcode uses street number, street name and PO Box, MatchUp does
not ask for these components individually (as few databases provide data in that way).
Instead, MatchUp asks for up to three address lines which it will parse internally.

Matchcode Mapping: Matchcode Component

Describes the currently highlighted matchcode component:

Data Type The currently selected matchcode component.

Size The number of characters this matchcode uses from the data.

Label A label that is attached to this component. MatchUp does not itself use this label, but it
can be helpful in remembering what a particular General component was for. The label is
specified in the Matchcode Editor.

Start Specifies where the matchcode extracts the data: from the Left, Right, or a specific
position.

Match Strategy Lists the component's fuzzy matching capability (if any).

Short/Empty Should blank or short fields be considered as matching or not?

Setup 91
Swap If this component is part of a Swap Pair, the letter that was assigned to that pair ("A"
through "H") will be listed.

See also: Matching Strategies

Matchcode Mapping: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this source field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Data Type and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion? Indicates what type of conversion will be done to get from the highlighted
field's data type to the output field's data type. For example, an input field containing city,
state and zip will have to be split to compare Zip codes.

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

92 MatchUp
4.2.4 Output Field Mapping

Output Field Mapping: Output Field Mapping

The initial Output file's structure is determined in Default Output Table. You can modify the
structure here, but you will probably want to start with one that is close to what you want for
your output file.

Note that information sometimes has to be extracted from complex fields. For example,
Setup 93
under national.dbf above, First and Last will be extracted from Name.

Add Output Field Adds a new field(s) to the output file's structure

Remove Output Field Removes a field(s) from the output structure.

Check Mappings Use Check Mapping to ensure that you haven't forgotten any source file's
fields.

Check Truncations MatchUp warns you about the possibility of truncation here.

Global Scatter/Gather An easier way to Scatter/Gather...

View File Although you cannot edit the data, you can view your file here. This is a great way
to make sure you have the right file!

Field List Shows a list of each field in each table. You can use your mouse to drag and drop
fields on this list into the desired mappings on the Output Field Mapping Screen.

Output Field Mapping: Add Output Field

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field. Some Field Types have fixed sizes.

Decimals If the field type is numeric there might be a decimal.

94 MatchUp
Setup 95
Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR".

Output Field Mapping: Output Field

Describes the currently highlighted output field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly.

Gather Check Gather if you want data gathered for the highlighted output field.

Method This option is not available unless Gather is checked for the highlighted output field.
See Gathering Methods.

Change Output Field Allows you to change all the specifications of a field: Name, Field
Type, Size, Decimals, and Data Type.

96 MatchUp
Output Field Mapping: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Output Field and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion? Indicates what type of conversion will be done to get from the highlighted
field's data type to the output field's data type. For example, an input field containing city,
state and zip will have to be split if the output table has separate city, state and zip fields.

Gather Check Gather if you want data from the highlighted input field to contribute in the
gathering process for the highlighted output field. This option is not available if Gather is not
checked for the highlighted output field.

In most cases, when you are gathering to an output field, you will want to get the data from
each input file. But there are cases when this is not desirable. For instance, if you were
gathering phone numbers, and one contributing database had outdated area codes, you
would probably not want to gather phone numbers from that database.

Scatter to this field Check this box if you would like the deposit contents of the highlighted

Setup 97
output field to be deposited into the current field.

Scatter to alternate field Check this box if you would like to deposit the contents of the
highlighted output field to an another input field.

Scattering is used to update source files (for example, updating phone numbers in a
database). Note that you don't have to gather in order to scatter. In fact, it is often not
desirable (like in a change of address update). Scattering is not available for read-only tables
or Suppression tables.

Note that scattering is one of the only functions in MatchUp that will overwrite your data. Use
this option with caution and always maintain a backup!

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

98 MatchUp
4.2.5 Output Tables

Result Tables Allows you to specify where you'd like to put Ouput, Duplicate, Suppressed
and Non-Intersected records. You can use the Browse button for the easiest way to select a
table.
· Output Table (optional) Specify the table that will receive output records. Output
records consist of unique records and one record from each duplicate group.
· Duplicate Table (optional) Specify the table that will receive duplicate records.
· Suppression Table (optional) Specify the table that will receive suppressed records.
Suppressed records consist of Regular, Self Purge, and No Purge records that did not
match against a Suppression list.
· Non-Intersection Table (optional) Specify the table that will receive non-intersected
records. Non-intersected records consist of Regular, Self Purge, and No Purge
records that did not match against an Intersection list.

Setup 99
Status Field (optional) Select a field name for capturing a record's disposition. During
processing, this field will be populated with a code indicating whether this record is a dupe,
has dupes, was suppressed, etc. See MatchUp Status Codes.

Output Matchcode Field (Advanced) (optional) Enter a field name for the Output
Matchcode. During processing, MatchUp will put the record's matchcode in this field.

There are a few practical uses for the Output Matchcode Field:
· Reusing this matchcode speeds processing in future runs.
· You can use this matchcode as a lookup field in your own database programs.
· It can be used as one of those "secret codes" people sometimes print on "Anonymous"
Business Reply Cards.

Multi-Buyer Count Field (Advanced) (optional) Specify a field to receive a record's multi-
buyer count. During processing, this field will be populated with a number indicating how
many inter-file dupes a record has. However, see Counting Method for alternate multi-buyer
counting methods.

Source Code Field (Advanced) (optional) Specify a field that contains source information
that should be used from this file when generating the Source Code Reports (See Reports).

Source Code Format Allows you to select the format of this code.

Count Field (optional) Select a field name to receive a record's dupe count. During
processing, this count field will be populated with a number indicating how many dupes a
record has. See Counting Method for the various counting methods.

Dupe Group Field (Advanced) (optional) Enter a field name which will receive the dupe
group number for this record. Each original record that MatchUp comes across is assigned a
number. When a dupe is found, it is assigned the same number. See Dupe Groups.

Multi-Buyer Field (Advanced) (optional) Specify a field which will receive multi-buyer
information for this record. See Multi-Buyer format.

Multi-Buyer Format Allows you to select the format of this code.

CASS Allows you to CASS and/or DPV the output table.

100 MatchUp
Output Tables: Source Code Format

Source tracking gives you a way to track what table (and optionally record number) a record
in the output table came from. If you like, you can name a new field, and it will be added to
the table. Table Information and Record Information allow you to format the information in the
Source Code field.

Table Information Select how you would like to show which table a record is from:
· Table Name Uses up to 8 characters of the input table's name.
· Table Number Uses up to 2 numbers to store the input table number
· Table Letter Uses 1 letter to indicate which input table number

Record Information Select how you would like to indicate which record a record came from:
· Long Numbers Uses up to 8 numbers to store the input record number.
· Medium Numbers Uses up to 6 numbers to store the input record number.
· Short Numbers Uses up to 4 numbers to store the input record number.
· Left Justify Left justifies the record number.
· Right Justify Right justifies the record number. If you select Right Justify, you also
have the option to "0" Fill.
· "0" Fill Fills the unused record information spaces with "0"s.

Sample Shows you what the Source Code Field will look like.

Setup 101
Output Tables: Multi-Buyer Format

Multi-Buyer tracking gives you a way to track what input table(s) the data in the output table
came from.

Source Code A key will be retrieved from the Input Source Code Field (specified on the
Advanced Tab), not the Source Code Field that you selected on the Output Tables screen. If
no field was specified for a particular source table, a table letter will be used instead.

Table Letter Uses 1 letter to indicate the input table.

Table Number Uses up to 1, 2, or 3 digits to store the input table number. You must specify
the number of digits you wish to use.

Delimit sources with Select how you would like to show which table a record is from: no
delimited, space, or comma.

Omit repeated sources Only report a source once for each table.

Sample Shows you what the Multi-Buyer data will look like.

Using the Multi-Buyer Field:

102 MatchUp
Now that you have one, what do you do with it? Well for starters, it's great for counts. Say
you're wondering how many output records had dupes in all three files. A simple Tools |
Browse | File Control | Count "For" condition would be:

'A' $ MultiBuyer .AND. 'B' $ MultiBuyer .AND. 'C' $ MultiBuyer

You'll find the "$" ("is contained in") operator very useful for these queries.

How about how many output records came from the second file and have dupes in the third
file?

Left(MultiBuyer,1)='B' .AND. 'C' $ MultiBuyer

You can count on the first letter to represent the output record, and any following letters to
represent dupe records.

Setup 103
Output Tables: CASS

Our CASS module checks, corrects and verifies addresses, by comparing them to the USPS
national database. The verification process standardizes your address to USPS
specifications by correcting misspellings, fixing nonstandard abbreviations, and adding
missing data.

CASS: Input/Output Fields

CASS the Output Table Turns CASS on or off for the Output Table.

Input MatchUp determines which field corresponds with which input line. You must provide
at least one address line, a city and a state for CASS processing. The more input information
you provide, the better the results.
· Company Very large companies sometimes get their own Plus 4, so providing a
company can increase accuracy.
104 MatchUp
 · Address Line 1 & 2 The delivery address lines.
· Suite The suite name and number.
· City or City/St/Zip The city, or if it is not pre-split, the entire city/state/zip.
· State The 2 letter state abbreviation.
· Zip The 5-digit zip code.
· Plus 4 The zip plus 4 extension.
· Urbanization A description of an area, sector, or development within a geographic
area. It is commonly used in urban Puerto Rico, as it describes the location of a given
street.
· Carrier Route Delivery carrier route.
· Delivery Point Code A two digit string which describe the 10th and 11th positions of a
12-digit POSTNet barcode.
· Delivery Point Check Digit A single digit which describes the 12th position of a 12-
digit POSTNet barcode.

Output Specify the fields that will receive the updated address information. You can elect to
use the input fields or new fields:
· Company
· Address Line 1 & 2
· Suite
· City If you desire the output city, state and zip code to all be in the same field, specify
the same output field for City, State, Zip and Plus 4.
· State
· Zip
· Plus 4 If the Zip and Plus 4 are the same field, the field will contain both codes.
· Urbanization
· Carrier Route
· Delivery Point Code
· Delivery Point Check Digit If this field is the same as the Delivery Point Code, it will
be appended to that value.

Carrier Route:
The carrier route is a letter followed by three numbers (for example "R009" and "C039"). The
alphabetic character describes the type of delivery:
Code Description
B PO Box
C City Delivery
G General Delivery
H Highway Contract
R Rural Route

Setup 105
CASS: Additional Output Fields

MatchUp will populate any of these additional fields from the CASS database:

Demographics Geopolitical information:

· City Abbreviation City names longer than 13 letters will have the Postal Services'
official abbreviation. If the city is 13 letters or less, the full city name is returned. For
example, "Fort Lauderdale" for "Ft Lauderdale" and "Fredericksburg" for
"Fredericksbrg".
· County The name of the County the verified address resides in.
· County FIPS Federal Information Processing Standard. A 5 digit code defined by the
U.S. Census Bureau. The first two digits represent the state code, the last three
representing the residing county.
· Time Zone A description of the zip code's time zone. example: "Eastern Time"
· Time Zone Code A 1 or 2 digit code representing the verified addresses' time zone
(see below).
106 MatchUp
 · MSA Number Metropolitan Statistical Area Number. A 4 digit code for a metropolitan
area made up of one or more counties and meeting statistical criteria defined by the
U.S. Office of Management and Budget (OMB). example: "1922"
· PMSA Number Primary Metropolitan Statistical Area. A 4 digit code for metropolitan
regions with a population greater than one million.
· Congressional Dist A 2 digit number representing the congressional district
associated with the verified address.

Geographics Note that latitude and longitude are accurate only to the 5-digit zip level, and
not the plus 4.
· Latitude The 7-digit latitude coordinate (accurate to 4 decimal places) of the center of
a Zip code.
· Longitude The 9-digit longitude of the center of a Zip code.

Split Components See Matchcode Components for more information on the components of
a street address. The Split Components fields are populated regardless of whether or not an
address has been successfully CASSed. If a record has been CASSed, these components
will be populated with the parsed components as they exist in the USPS database. If not,
MatchUp's internal street splitter will parse the components instead.

Address Status
· LACS Locatable Address Conversion Service. The total number of records which were
flagged as having undergone a change to a city-style address (to allow emergency
services to locate these addresses more efficiently). The address change is not
reflected here.
· EWS Early Warning System Data. Count of new addresses scheduled to be included
in the next release of the bimonthly USPS national database. New construction
projects for example, are flagged by the EWS file.
· Address Type A single character code indicating the type of verified address.
· DPV Footnotes A set of one or more 2 character codes containing delivery point
information.

Setup 107
Time Zone Code:
Code Zone
(blank) Military
4 Atlantic Time
5 Eastern Time
6 Central Time
7 Mountain Time
8 Pacific Time
9 Alaska time
10 Hawaii Time
11 Samoa Time
13 Marshall Island Time
14 Guam Time
15 Palau Time
Address Type:
Code Description
(blank) unverified
F Company Address
G General Delivery
H High-Rise or business complex
P PO Box
R Rural Route
S Residential Address

108 MatchUp
DPV Footnotes:
Code Description
AA Input address matched to the Zip+4 file
A1 Input address not matched to the Zip+4 file
BB Input address matched to DPV (all components)
CC Input address primary number (street number) matched to DPV but secondary
number not matched
N1 Input address primary number matched to DPV but secondary number missing
M1 Input address primary number missing
M3 Input address primary number invalid
P1 Input address PO, RR or HC box number missing
P3 Input address PO, RR or HC box number invalid
RR Input address matched to CMRA
R1 Input address matched to CMRA but secondary number not present
F1 Address was coded to a military address
G1 Address was coded to a General Delivery address
U1 Address was coded to a unique Zip Code

Setup 109
CASS: Options

Output Error Code Select the field you want to use for the error code.

Output Status Code Select the field you want to use for the status code.

If there's a coding error What should MatchUp do if it comes across a coding error (you
can choose one or both):
· Clear the Input Plus 4 field
· Clear the Input Carrier Route field

City/State Delimiter If you elect to put city, state, and zip data into a single field, how would
you like them delimited?
· Delimit City & State with a space
· Delimit City & State with a comma

110 MatchUp
Zip/Plus 4 Delimiter Same deal, but separating the Zip and Plus 4:
· Delimit Zip & Plus 4 with a dash
· Delimit Zip & Plus 4 with a space
· Delimit Zip & Plus 4 with no delimiters

Perform Delivery Point Validation (DPV) Check if you want to verify each address's
delivery point during CASS verification.

Use CASSmate enhanced processing Whether or not to use the enhanced power of
CASSmate to get more CASS matches. Enhanced CASSmate processing is only used when
normal CASSing attempts have yielded no results.

Form 3553 information This is the general information necessary to fill out the USPS Form
3553 for the CASS certification. Be sure to fill in this information or your Form 3553 will not
print out complete.

Setup 111
Error Codes:
Code Description
(blank) No Error
M Multiple Matches More than one record matches the address and there is not
enough information available in the input address to break the tie between
multiple records. Passing more complete information, such as city names or
urbanization names, can help reduce the number of multiple match errors.
N No Street Data for Zip The Zip Code exists but no streets begin with the same
letter in that Zip Code.
R Range Error The address was found but the street number in the input address
was not between the low and high range in the CASS database.
T Component Error Either the directionals or the suffix field did not match the
CASS database, and there was more than one choice for correcting the address.
For example, if the given address was "100 Main St" and the only addresses
found were "100 E Main St" and 100 Main Ave" the error code "T" would be
returned because we do not know whether to add the directional "E" or to
change the suffix to "Ave".
U Unknown Street An exact street name match could not be found and
phonetically matching the street name resulted in either no matches or matches
to more than one street name.
X Non-Deliverable Address The physical location exists but there are no homes
on this street. One reason might be that there are railroad tracks or a river
running alongside this street, as they would prevent construction of homes in this
location.
Z Invalid Zip Code The Zip Code does not exist and could not be determined by
the city and state.
C Canadian Postal Code The Zip Code matches the format characteristics of a
Canadian Postal Code.
W EWS File Record The Zip Code was found in the Early Warning System Data.
These are new addresses scheduled to be included in the next USPS national
database.
EWS:
Having a EWS file prevents new addresses from being miscoded with information from the
(soon-to-be-outdated) USPS database. For example, an address of 44 Legacy Drive might
be changed to 44 Legacy Street if Legacy Drive does not exist in the EWS File. However, if
legacy Drive is in the EWS File, the address is left alone (no Plus 4) and coded with an EWS
Error. The next CASS database update should contain the new information and will code the
address correctly.

The EWS file is updated weekly every Thursday. To get the most current file, you can
download it from ftp://www.MelissaData.com/updates/ews.txt. Place the ews.txt file in the
same folder as your other CASS Databases (mdAddr.dat, mdAddr.lic, etc.) overwriting the
previous week's file.

Status Codes:

112 MatchUp
Code Description
D Demo Mode If processing with the demo CASS database, records not in the
state of Nevada will be coded with "D".
E Expired Database The CASS database has expired.
S Standardized but Not Coded Standardization means that some conversion
was done on the address (for example, changing "Post Office Box" to "PO
Box").
V Street Number Validated to DPV Level This record has been DPV validated.
You can check the DPV Footnotes for more information about the level of
validation.
X Address Not Coded Check the Error code and/or DPV Footnotes for more
information.
7 Multiple Matches There were multiple matches for the address that were all in
the same Zip Code and Carrier Route. The returned Zip Code and Carrier Route
will be correct but you will not get any Plus 4 information.
9 Fully Coded The address was fully CASS coded.

Setup 113
4.2.6 Advanced

View File Although you cannot edit the data, you can view your file here.

Input Matchcode (Advanced) (optional) If you have used MatchUp to process this file before
and saved the matchcode (See "Output Matchcode" below), you may use this matchcode
instead of having the program build it again. Why? For lists that you process all the time, it's
much faster to build a matchcode once and reuse it.

Input Matchcode Restrictions:

· The matchcode you used to build this matchcode must be the exact same one you are
using now - MatchUp cannot verify this for you.
· This matchcode field does not automatically update if you change the contents of a
record. You must adjust the matchcode accordingly. (With one exception - if you fill a
record's Input Matchcode Field with "*****" (five asterisks), MatchUp will rebuild the
Matchcode.

114 MatchUp
Output Matchcode (Advanced) (optional) Enter a field name for the Output Matchcode.
During processing, MatchUp will put the record's matchcode in this field.

There are a few practical uses for the Output Matchcode Field:
· Reusing this matchcode speeds processing in future runs.
· You can use this matchcode as a lookup field in your own database programs.
· It can be used as one of those "secret codes" people sometimes print on "Anonymous"
Business Reply Cards.

Output Multi-Buyer Count (Advanced) (optional) Enter a field name which will receive a
record's multi-buyer count. During processing, this field will be populated with a number
indicating how many inter-file dupes a record has. However, see Counting Method for
alternate multi-buyer counting methods.

Output Dupe Group (Advanced) (optional) Enter a field name which will receive the dupe
group number for this record. Each original record that MatchUp comes across is assigned a
number. When a dupe is found, it is assigned the same number. (See Dupe Groups)

Input Source Code (Advanced) (optional) Specify a field that contains source information
that should be used from this table when generating the Source Reports (See Reports).

Use leftmost ... characters (Advanced) Specify how many leading characters should be
extracted from the Input Source Field and used in generating the Source Reports (See
Reports).

Input Filter (Advanced) (optional) Enter a filter condition to limit records processed from the
currently highlighted input table. This filter condition must be specified in dBASE Syntax or
you can access the Expression Builder.

Setup 115
Result Tables Allows you to specify where you'd like to put Ouput, Duplicate, Suppressed
and Non-Intersected records. You can use the Browse button for the easiest way to select a
table. Records placed in these tables will be specific to the currently highlighted input table.
You can specify result table(s) for any input table(s).
· Output Table (Advanced) (optional) Specify the table that will receive output records
from the highlighted input table. Output records consist of unique records and one
record from each duplicate group.
· Duplicate Table (Advanced) (optional) Specify the table that will receive duplicate
records from the highlighted input table.
· Suppression Table (Advanced) (optional) Specify the table that will receive
suppressed records from the highlighted input table. Suppressed records consist of
Regular, Self Purge, and No Purge records that did not match against a Suppression
list.
· Non-Intersection Table (Advanced) (optional) Specify the table that will receive non-
intersected records from the highlighted input table. Non-intersected records consist of
Regular, Self Purge, and No Purge records that did not match against an Intersection
list.

116 MatchUp
Setup 117
4.3 Purge Setup

Select Purge if you want to take one (or several) tables and remove matching records.

4.3.1 General

118 MatchUp
General: Matchcode

Select the matchcode that best suits your needs. A selection is supplied, and you will
probably find a suitable one from this collection. If not, you can click Matchcode Editor to
create (or edit) a matchcode.

The table above lists the properties of the currently selected matchcode. It is for viewing
purposes only.

Matching Rules: MatchUp uses up to 16 matchcode "combinations" which are evaluated

simultaneously. In the table, the X's in the columns marked 1 through 16 define component
combinations that should be considered a match. The Matching Rules section simply
displays them in an alternate form.

Setup 119
General: Ranking

Ranking Allows you to choose which records to favor when dupes are found:
· MatchUp will select the best record Take the record with the most complete
information.
· I will assign priorities or specify a priority field Allows you to choose which table or
records to favor when duplicate records are found. If you select this option, you will be
able to assign a rank or rank field for each input table on the Input Tables tab.
· Prioritization will be random for each record MatchUp will randomly select a record
to use as the best record when duplicate records are found.
· A uniform distribution of records will be selected Instructs MatchUp to try (as best
it can) to assign output records uniformly from all input files.
· Doesn't Matter Take the first record you come across.

What's the difference between Random and Uniform?

Random ranking tends to take an appropriate proportion of output records from each source.
Uniform Distribution tries to take an equal quantity of output records from each source. For
example, if there were 100 output records:

Total Records Random Uniform

File1 9,000 90 50
File2 1,000 10 50
Ascending MatchUp will select records with a lower rank over records with a higher rank
(i.e., "A" over "B", "1" over "2", and "03/10/68" over "04/10/68").

Descending MatchUp will select records with a higher rank over records with a lower rank
(i.e., "C" over "B", "3" over "2", and "05/10/68" over "04/10/68").

120 MatchUp
When matching records have the same rank If MatchUp can't decide which record to use
because they both have the same rank (ie, a tie), a backup plan can be selected:
· The first record processed will be selected
· The last record processed will be selected
· A random record will be selected

4.3.2 Input Tables

Setup 121
Input Tables: Input Sources

This screen displays the tables you have added.

You can change the placement of an input table by clicking and dragging it into the desired
position. Alternately, you can press CTRL+UP ARROW or CTRL+DOWN ARROW to change
the table's position. Note that the order of input tables has no impact on processing, but is
purely for your benefit.

Add... Allows you to add an additional table to the setup.

Remove Removes a table from the merge/purge setup.

Change... Allows you to replace a table in the merge/purge setup with another table.

View Although you cannot edit the data, you can view your table here.

Input Tables: Information About

Table Type Select how the highlighted table will be handled:

· Regular This table will be deduped against itself and any other Regular tables.
· Suppression Records in Regular, No Purge, and Self Purge tables which also appear

122 MatchUp
 in a Suppression list will be rejected. Ideal for "do not mail" lists.
· Intersection Records in Regular, No Purge, and Self Purge tables which do not
appear on an Intersect list will be rejected. (opposite of Suppression).
· Self Purge This table will be matched against itself, Suppression, and Intersection
tables, but not against Regular, No Purge, or other Self Purge tables. Self Purge
tables are used when you would like to match several tables against a Suppression
list, and purge internal dupes, but not match the tables against each other.
· No Purge This table will not be deduped against any other table. It will, however, be
matched against Suppression and Intersection tables. No Purge tables are used when
you would like to match several tables against a Suppression list, but not dedupe the
tables internally or against each other.
· Advanced Click if the contents of a field dictate the Table Type.

Alias Enter a name you would like to refer to this table. This name will appear on all reports.

Priority Select a ranking method for the highlighted table:

· Assign Priority Enter a rank for this table. This option only appears if you selected I
will assign priorities or specify a priority field under the General tab.
· Contents of Field Enter the name of the rank field. This option only appears if you
selected I will assign priorities or specify a priority field under the General tab. This
field must already exist in your table.
· Advanced Rank Check this option when the values in the specified Field are not
easily compared.
· Custom Expression Enter a valid dBASE expression for evaluating rank. Use it when
you need a tighter criteria than a single field or table will allow. For example, if you
want to mail to the oldest female in a household, enter the expression: SEX +
DtoS(BIRTHDATE). Only appears if you selected I will assign priorities or specify a
priority field under the General tab.

Status Marking You can select methods of marking a record's disposition with these
options:
· Mark rejected records for deletion If checked, rejected records will be marked for
deletion. This can only be done with dBASE tables.
· Status Field (optional) Select a field name for capturing a record's disposition. During
processing, this field will be populated with a code indicating whether this record is a
dupe, has dupes, was suppressed, etc. See MatchUp Status Codes.
· Count Field (optional) Select a field name to receive a record's dupe count. During
processing, this count field will be populated with a number indicating how many
dupes a record has. See Counting Method.

Setup 123
Information About: Advanced Table Type

Sometimes tables are created by a Merge process and contain a code field specifying their
original source. Before MatchUp, to process different source codes as different table types,
users would have to 'undo' the merged file into separate files, then Merge/Purge them. Now
you can specify the table type as Advanced and specify the field containing the code
indicating how the record should be processed.

Base Table Type selection on contents of field The field containing a code indicating how
a record should be processed.

Number of characters to use from field How many characters in the above field should be
used in considering how a record should be processed.

Ignore case Whether or not case should be considered when looking up a code in the lists
below.

Table Type 'Regular' when code is A list of code(s) that indicate that a record should be
processed as a regular record.

124 MatchUp
Table Type 'Suppression' when code is A list of code(s) that indicate that a record should
be processed as a suppression record.

Table Type 'Intersection' when code is A list of code(s) that indicate that a record should
be processed as an intersection record.

Table Type 'Self Purge' when code is A list of code(s) that indicate that a record should be
processed as a self purge record.

Table Type 'No Purge' when code is A list of code(s) that indicate that a record should be
processed as a no purge record.

Any codes not in any of the above lists should be processed Other codes will be
processed in the following manner:
· Not Processed (filtered)
· Table Type 'Regular'
· Table Type 'Suppression'
· Table Type 'Intersection'
· Table Type 'Self Purge'
· Table Type 'No Purge'

Information About: Advanced Rank

When the field chosen to provide the ranking has values which can not easily be compared,
Advanced Ranking can provide an order to those values.

For example, a rank field containing Months would wind up with unusual prioritization if the
values were compared by character strings: Apr, Dec, Feb, Jan, Jul, Jun, etc.

Setup 125
When the list of prioritized ranks is quite long, it can be get tedious to manually enter a list of
ranks in the manner specified above. A more efficient method is to create a two-field lookup
table (dBASE III format) and specify it using the first option, Use an external file listing
ranks and priorities. In our example, the lookup table would look like the following:

Month Priority
Jan 1
Feb 2
Mar 3
Apr 4
May 5
Jun 6
Jul 7
Aug 8
Sep 9
Oct 10
Nov 11
Dec 12

126 MatchUp
Number of characters to use from field How many characters in the specified rank field
should be used in considering a record's priority.

Ignore case Whether or not case should be considered when looking up a rank in the list
below.

Advanced rank assignment can be done in one of two ways:

· Use an External File If you have a pre-existing dBASE table with the desired ranks, or
have many potential codes - in which case creating a table is a good idea, use this
method. One advantage of this method is you can use the external table in other
setups.
· Rank in This Order Enter a string of codes in ranking order from lowest to highest
priority. When you don't have a lot of different codes in the selected field, or don't want
to bother creating a new dBASE file, use this method.

Codes that don't appear on the above list should be Specify what to do when a record
has a value in the Rank Field which is not present in the above list:
· Not Processed (filtered)
· Assigned the lowest priority
· Assigned the highest priority

Setup 127
4.3.3 Matchcode Mapping

Matchcode Mapping: Field Mapping for Matchcode

This screen tells MatchUp which field contains the data for each component of the
matchcode.

128 MatchUp
These components are matched with fields from your input table. MatchUp tries to determine
a field's type based on the field name. For example, MatchUp will default to "Address" if the
field's name starts with "ADD", or "STR". You can customize these defaults in Field Naming.
If MatchUp doesn't pick correctly, double-click on the incorrect field name and select the
correct field.

The purple field names are fields that need to be parsed to get the needed matchcode
component. So, even though this matchcode is using last name, MatchUp can use a full
name field and find the information it needs.

And, although this matchcode uses street number, street name and PO Box, MatchUp does
not ask for these components individually (as few databases provide data in that way).
Instead, MatchUp asks for up to three address lines which it will parse internally.

Setup 129
Matchcode Mapping: Matchcode Component

Describes the currently highlighted matchcode component:

Data Type The currently selected matchcode component.

Size The number of characters this matchcode uses from the data.

Label A label that is attached to this component. MatchUp does not itself use this label, but it
can be helpful in remembering what a particular General component was for. The label is
specified in the Matchcode Editor.

Start Specifies where the matchcode extracts the data: from the Left, Right, or a specific
position.

Match Strategy Lists the component's fuzzy matching capability (if any).

Short/Empty Should blank or short fields be considered as matching or not?

Swap If this component is part of a Swap Pair, the letter that was assigned to that pair ("A"
through "H") will be listed.

See also: Matching Strategies

130 MatchUp
Matchcode Mapping: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this source field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Data Type and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion? Indicates what type of conversion will be done to get from the highlighted
field's data type to the output field's data type. For example, an input field containing city,
state and zip will have to be split to compare Zip codes.

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

Setup 131
4.3.4 Advanced

View File Although you cannot edit the data, you can view your file here.

Input Matchcode (Advanced) (optional) If you have used MatchUp to process this file before
and saved the matchcode (See "Output Matchcode" below), you may use this matchcode
instead of having the program build it again. Why? For lists that you process all the time, it's
much faster to build a matchcode once and reuse it.

Input Matchcode Restrictions:

· The matchcode you used to build this matchcode must be the exact same one you are
using now - MatchUp cannot verify this for you.
· This matchcode field does not automatically update if you change the contents of a
record. You must adjust the matchcode accordingly. (With one exception - if you fill a
record's Input Matchcode Field with "*****" (five asterisks), MatchUp will rebuild the
Matchcode.

132 MatchUp
Output Matchcode (Advanced) (optional) Enter a field name for the Output Matchcode.
During processing, MatchUp will put the record's matchcode in this field.

There are a few practical uses for the Output Matchcode Field:
· Reusing this matchcode speeds processing in future runs.
· You can use this matchcode as a lookup field in your own database programs.
· It can be used as one of those "secret codes" people sometimes print on "Anonymous"
Business Reply Cards.

Output Multi-Buyer Count (Advanced) (optional) Enter a field name which will receive a
record's multi-buyer count. During processing, this field will be populated with a number
indicating how many inter-file dupes a record has. However, see Counting Method for
alternate multi-buyer counting methods.

Output Dupe Group (Advanced) (optional) Enter a field name which will receive the dupe
group number for this record. Each original record that MatchUp comes across is assigned a
number. When a dupe is found, it is assigned the same number. (See Dupe Groups)

Input Source Code (Advanced) (optional) Specify a field that contains source information
that should be used from this table when generating the Source Reports (See Reports).

Use leftmost ... characters (Advanced) Specify how many leading characters should be
extracted from the Input Source Field and used in generating the Source Reports (See
Reports).

Input Filter (Advanced) (optional) Enter a filter condition to limit records processed from the
currently highlighted input table. This filter condition must be specified in dBASE Syntax or
you can access the Expression Builder.

Setup 133
Result Tables Allows you to specify where you'd like to put Output, Duplicate, Suppressed
and Non-Intersected records. You can use the Browse button for the easiest way to select a
table. Records placed in these tables will be specific to the currently highlighted input table.
You can specify result table(s) for any input table(s).
· Output Table (Advanced) (optional) Specify the table that will receive output records
from the highlighted input table. Output records consist of unique records and one
record from each duplicate group.
· Duplicate Table (Advanced) (optional) Specify the table that will receive duplicate
records from the highlighted input table.
· Suppression Table (Advanced) (optional) Specify the table that will receive
suppressed records from the highlighted input table. Suppressed records consist of
Regular, Self Purge, and No Purge records that did not match against a Suppression
list.
· Non-Intersection Table (Advanced) (optional) Specify the table that will receive non-
intersected records from the highlighted input table. Non-intersected records consist of
Regular, Self Purge, and No Purge records that did not match against an Intersection
list.

134 MatchUp
Setup 135
4.4 File Update
File Update allows you to transfer data from one table to another. For example, if you have
one table that contains addresses and phone numbers and would like to transfer these
phone numbers to another table (containing matching addresses), you can do this through
File Update.

Another popular usage for File Update is the updating of NCOA information. Say you send a
table to the USPS for an NCOA or Move Update. Often, you won't send every field in your
table, just the name and address fields. When you receive the updated list, File Update can
be used to transfer addresses from the update list back into the original table. In these
cases, it is advantageous to have a unique ID field in both tables, as address matching isn't a
good idea for NCOA or Move updates.

136 MatchUp
4.4.1 General

Setup 137
General: Matchcode

Select the matchcode that best suits your needs. A selection is supplied, and you will
probably find a suitable one from this collection. If not, you can click Matchcode Editor to
create (or edit) a matchcode.

The table above lists the properties of the currently selected matchcode. It is for viewing
purposes only.

Matching Rules: MatchUp uses up to 16 matchcode "combinations" which are evaluated

simultaneously. In the table, the X's in the columns marked 1 through 16 define component
combinations that should be considered a match. The Matching Rules section simply
displays them in an alternate form.

138 MatchUp
General: Ranking

Ranking Allows you to choose which records to favor when dupes are found:
· MatchUp will select the best record Take the record with the most complete
information.
· I will assign priorities or specify a priority field Allows you to choose which table or
records to favor when duplicate records are found. If you select this option, you will be
able to assign a rank or rank field for each input table on the Input Tables tab.
· Prioritization will be random for each record MatchUp will randomly select a record
to use as the best record when duplicate records are found.
· A uniform distribution of records will be selected Instructs MatchUp to try (as best
it can) to assign output records uniformly from all input files.
· Doesn't Matter Take the first record you come across.

What's the difference between Random and Uniform?

Random ranking tends to take an appropriate proportion of output records from each source.
Uniform Distribution tries to take an equal quantity of output records from each source. For
example, if there were 100 output records:

Total Records Random Uniform

File1 9,000 90 50
File2 1,000 10 50
Ascending MatchUp will select records with a lower rank over records with a higher rank
(i.e., "A" over "B", "1" over "2", and "03/10/68" over "04/10/68").

Descending MatchUp will select records with a higher rank over records with a lower rank
(i.e., "C" over "B", "3" over "2", and "05/10/68" over "04/10/68").

Setup 139
When matching records have the same rank If MatchUp can't decide which record to use
because they both have the same rank (ie, a tie), a backup plan can be selected:
· The first record processed will be selected
· The last record processed will be selected
· A random record will be selected

140 MatchUp
4.4.2 Input Tables

Input Tables: Input Sources

This screen displays the tables you have added.

You can change the placement of an input table by clicking and dragging it into the desired
position. Alternately, you can press CTRL+UP ARROW or CTRL+DOWN ARROW to change
the table's position. Note that the order of input tables has no impact on processing, but is
purely for your benefit.

Setup 141
Add... Allows you to add an additional table to the setup.

Remove Removes a table from the merge/purge setup.

Change... Allows you to replace a table in the merge/purge setup with another table.

View Although you cannot edit the data, you can view your table here.

Input Tables: Information About

Table Type Select how the highlighted table will be handled:

· Input This table will supply the data to update other tables.
· Output This table will updated with the data from the Input table(s). NOTE: You only
get one(1) output table.

Alias Enter a name you would like to refer to this table. This name will appear on all reports.

Priority Select a ranking method for the highlighted table:

· Assign Priority Enter a rank for this table. This option only appears if you selected I
will assign priorities or specify a priority field under the General tab.
· Contents of Field Enter the name of the rank field. This option only appears if you
selected I will assign priorities or specify a priority field under the General tab. This
field must already exist in your table.
· Advanced Rank Check this option when the values in the specified Field are not
easily compared.
· Custom Expression Enter a valid dBASE expression for evaluating rank. Use it when
you need a tighter criteria than a single field or table will allow. For example, if you
want to mail to the oldest female in a household, enter the expression: SEX +
DtoS(BIRTHDATE). Only appears if you selected I will assign priorities or specify a
priority field under the General tab.

142 MatchUp
4.4.3 Matchcode Mapping

Matchcode Mapping: Field Mapping for Matchcode

This screen tells MatchUp which field contains the data for each component of the
matchcode.

Setup 143
These components are matched with fields from your input table. MatchUp tries to determine
a field's type based on the field name. For example, MatchUp will default to "Address" if the
field's name starts with "ADD", or "STR". You can customize these defaults in Field Naming.
If MatchUp doesn't pick correctly, double-click on the incorrect field name and select the
correct field.

The purple field names are fields that need to be parsed to get the needed matchcode
component. So, even though this matchcode is using last name, MatchUp can use a full
name field and find the information it needs.

And, although this matchcode uses street number, street name and PO Box, MatchUp does
not ask for these components individually (as few databases provide data in that way).
Instead, MatchUp asks for up to three address lines which it will parse internally.

144 MatchUp
Matchcode Mapping: Matchcode Component

Describes the currently highlighted matchcode component:

Data Type The currently selected matchcode component.

Size The number of characters this matchcode uses from the data.

Label A label that is attached to this component. MatchUp does not itself use this label, but it
can be helpful in remembering what a particular General component was for. The label is
specified in the Matchcode Editor.

Start Specifies where the matchcode extracts the data: from the Left, Right, or a specific
position.

Match Strategy Lists the component's fuzzy matching capability (if any).

Short/Empty Should blank or short fields be considered as matching or not?

Swap If this component is part of a Swap Pair, the letter that was assigned to that pair ("A"
through "H") will be listed.

See also: Matching Strategies

Setup 145
Matchcode Mapping: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this source field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Data Type and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion? Indicates what type of conversion will be done to get from the highlighted
field's data type to the output field's data type. For example, an input field containing city,
state and zip will have to be split to compare Zip codes.

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

146 MatchUp
4.4.4 Update

Input Sources These tables will supply the data to output table.

Output Source This table will be updated with the data from the Input table(s).

Global Update An easier way to make multiple update settings.

View Although you cannot edit the data, you can view your table here.

Setup 147
Update: Input Field

Describes the currently highlighted input field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly. MatchUp tries to determine a field's type based on the field
name. For example, MatchUp will default to "Address" if the field's name starts with "ADD",
or "STR". If MatchUp doesn't pick correctly, click Change Output Field and select the
description which best suits this field. You can customize these defaults in Field Naming.

Conversion? Indicates what type of conversion will be done to get from the highlighted
field's data type to the output field's data type. For example, an input field containing city,
state and zip will have to be split if the output table has separate city, state and zip fields.

Change Data Type Allows you to change the Data Type of an input field. Single right-click
for a pop-up menu or double right-click for a dialog box.

148 MatchUp
Update: Output Field

Describes the currently highlighted output field:

Name Name of the field.

Field Type Valid types are: Fixed Character, Variable Character, Integer, Float, Decimal,
Logical, and DateTime.

Size Size of the field.

Decimals If the field type is numeric there may be a decimal.

Data Type The type of information contained in this output field. Make sure this is correct or
the fields will not merge properly.

Update Method Select the method of updating your data. See Gathering Methods for more
information.

Change Data Type Allows you to change the Data Type of an output field. Single right-click
for a pop-up menu or double right-click for a dialog box.

Setup 149
Update: Global Update

When you need to set several update settings, it can get quite tedious to set up each table.
To make this process less painful, a global update dialog is provided.

Once you select a field(s) you want to update, check (or uncheck) Update and then select
the Method you want to use. To execute your selections, click Change!

150 MatchUp
Setup 151
4.5 CASS
Our CASS module checks, corrects and verifies addresses, by comparing them to the USPS
national database. The verification process standardizes your address to USPS
specifications by correcting misspellings, fixing nonstandard abbreviations, and adding
missing data.

4.5.1 Input/Output Fields

Input Table Click Browse to select the table that you would like to CASS verify.

Input MatchUp determines which field corresponds with which input line. You must provide
at least one address line, a city and a state for CASS processing. The more input information
you provide, the better the results.
· Company Very large companies sometimes get their own Plus 4, so providing a
company can increase accuracy.
· Address Line 1 & 2 The delivery address lines.
152 MatchUp
 · Suite The suite name and number.
· City or City/St/Zip The city, or if it is not pre-split, the entire city/state/zip.
· State The 2 letter state abbreviation.
· Zip The 5-digit zip code.
· Plus 4 The zip plus 4 extension.
· Urbanization A description of an area, sector, or development within a geographic
area. It is commonly used in urban Puerto Rico, as it describes the location of a given
street.
· Carrier Route Delivery carrier route.
· Delivery Point Code A two digit string which describe the 10th and 11th positions of a
12-digit POSTNet barcode.
· Delivery Point Check Digit A single digit which describes the 12th position of a 12-
digit POSTNet barcode.

Output Specify the fields that will receive the updated address information. You can elect to
use the input fields or new fields:
· Company
· Address Line 1 & 2
· Suite
· City If you desire the output city, state and zip code to all be in the same field, specify
the same output field for City, State, Zip and Plus 4.
· State
· Zip
· Plus 4 If the Zip and Plus 4 are the same field, the field will contain both codes.
· Urbanization
· Carrier Route
· Delivery Point Code
· Delivery Point Check Digit If this field is the same as the Delivery Point Code, it will
be appended to that value.

Carrier Route:
The carrier route is a letter followed by three numbers (for example "R009" and "C039"). The
alphabetic character describes the type of delivery:
Code Description
B PO Box
C City Delivery
G General Delivery
H Highway Contract
R Rural Route

Setup 153
4.5.2 Additional Output Fields

MatchUp will populate any of these additional fields from the CASS database:

Demographics Geopolitical information:

· City Abbreviation City names longer than 13 letters will have the Postal Services'
official abbreviation. If the city is 13 letters or less, the full city name is returned. For
example, "Fort Lauderdale" for "Ft Lauderdale" and "Fredericksburg" for
"Fredericksbrg".
· County The name of the County the verified address resides in.
· County FIPS Federal Information Processing Standard. A 5 digit code defined by the
U.S. Census Bureau. The first two digits represent the state code, the last three
representing the residing county.
· Time Zone A description of the zip code's time zone. example: "Eastern Time"

154 MatchUp
· Time Zone Code A 1 or 2 digit code representing the verified addresses' time zone
(see below).

Setup 155
 · MSA Number Metropolitan Statistical Area Number. A 4 digit code for a metropolitan
area made up of one or more counties and meeting statistical criteria defined by the
U.S. Office of Management and Budget (OMB). example: "1922"
· PMSA Number Primary Metropolitan Statistical Area. A 4 digit code for metropolitan
regions with a population greater than one million.
· Congressional Dist A 2 digit number representing the congressional district
associated with the verified address.

Geographics Note that latitude and longitude are accurate only to the 5-digit zip level, and
not the plus 4.
· Latitude The 7-digit latitude coordinate (accurate to 4 decimal places) of the center of
a Zip code.
· Longitude The 9-digit longitude of the center of a Zip code.

Split Components See Matchcode Components for more information on the components of
a street address. The Split Components fields are populated regardless of whether or not an
address has been successfully CASSed. If a record has been CASSed, these components
will be populated with the parsed components as they exist in the USPS database. If not,
MatchUp's internal street splitter will parse the components instead.

Address Status
· LACS Locatable Address Conversion Service. The total number of records which were
flagged as having undergone a change to a city-style address (to allow emergency
services to locate these addresses more efficiently). The address change is not
reflected here.
· EWS Early Warning System Data. Count of new addresses scheduled to be included
in the next release of the bimonthly USPS national database. New construction
projects for example, are flagged by the EWS file.
· Address Type A single character code indicating the type of verified address.
· DPV Footnotes A set of one or more 2 character codes containing delivery point
information.

156 MatchUp
Time Zone Code:
Code Zone
(blank) Military
4 Atlantic Time
5 Eastern Time
6 Central Time
7 Mountain Time
8 Pacific Time
9 Alaska time
10 Hawaii Time
11 Samoa Time
13 Marshall Island Time
14 Guam Time
15 Palau Time
Address Type:
Code Description
(blank) unverified
F Company Address
G General Delivery
H High-Rise or business complex
P PO Box
R Rural Route
S Residential Address

Setup 157
DPV Footnotes:
Code Description
AA Input address matched to the Zip+4 file
A1 Input address not matched to the Zip+4 file
BB Input address matched to DPV (all components)
CC Input address primary number (street number) matched to DPV but secondary
number not matched
N1 Input address primary number matched to DPV but secondary number missing
M1 Input address primary number missing
M3 Input address primary number invalid
P1 Input address PO, RR or HC box number missing
P3 Input address PO, RR or HC box number invalid
RR Input address matched to CMRA
R1 Input address matched to CMRA but secondary number not present
F1 Address was coded to a military address
G1 Address was coded to a General Delivery address
U1 Address was coded to a unique Zip Code

158 MatchUp
4.5.3 Options

Output Error Code Select the field you want to use for the error code.

Output Status Code Select the field you want to use for the status code.

If there's a coding error What should MatchUp do if it comes across a coding error (you
can choose one or both):
· Clear the Input Plus 4 field
· Clear the Input Carrier Route field

City/State Delimiter If you elect to put city, state, and zip data into a single field, how would
you like them delimited?
· Delimit City & State with a space

Setup 159
· Delimit City & State with a comma

160 MatchUp
Zip/Plus 4 Delimiter Same deal, but separating the Zip and Plus 4:
· Delimit Zip & Plus 4 with a dash
· Delimit Zip & Plus 4 with a space
· Delimit Zip & Plus 4 with no delimiters

Input Filter (Advanced) (optional) Enter the input filtering criteria for each source table.
MatchUp will allow you to use a different filter for each input table. The filter must be entered
in dBASE syntax. Instead of typing in a dBASE expression, or if you're unfamiliar with dBASE
syntax, you can use the Expression Builder.

Perform Delivery Point Validation (DPV) Check if you want to verify each address's
delivery point during CASS verification.

Use CASSmate enhanced processing Whether or not to use the enhanced power of
CASSmate to get more CASS matches. Enhanced CASSmate processing is only used when
normal CASSing attempts have yielded no results.

Process via Zip Index CASSmate generally processes faster in Zip order. But, often your
table is not in Zip order (and you don't want it to be). With this option checked, MatchUp will
create a temporary index that will cause your table to be CASSed in Zip order (but it's
physical order will remain untouched). This option is disabled for all table types exact dBASE
(the speed gain is countered by random access slowdowns).

Setup 161
Form 3553 information This is the general information necessary to fill out the USPS Form
3553 for the CASS certification. Be sure to fill in this information or your Form 3553 will not
print out complete.

Error Codes:
Code Description
(blank) No Error
M Multiple Matches More than one record matches the address and there is not
enough information available in the input address to break the tie between
multiple records. Passing more complete information, such as city names or
urbanization names, can help reduce the number of multiple match errors.
N No Street Data for Zip The Zip Code exists but no streets begin with the same
letter in that Zip Code.
R Range Error The address was found but the street number in the input address
was not between the low and high range in the CASS database.
T Component Error Either the directionals or the suffix field did not match the
CASS database, and there was more than one choice for correcting the address.
For example, if the given address was "100 Main St" and the only addresses
found were "100 E Main St" and 100 Main Ave" the error code "T" would be
returned because we do not know whether to add the directional "E" or to
change the suffix to "Ave".
U Unknown Street An exact street name match could not be found and
phonetically matching the street name resulted in either no matches or matches
to more than one street name.
X Non-Deliverable Address The physical location exists but there are no homes
on this street. One reason might be that there are railroad tracks or a river
running alongside this street, as they would prevent construction of homes in this
location.
Z Invalid Zip Code The Zip Code does not exist and could not be determined by
the city and state.
C Canadian Postal Code The Zip Code matches the format characteristics of a
Canadian Postal Code.
W EWS File Record The Zip Code was found in the Early Warning System Data.
These are new addresses scheduled to be included in the next USPS national
database.
EWS:
Having a EWS file prevents new addresses from being miscoded with information from the
(soon-to-be-outdated) USPS database. For example, an address of 44 Legacy Drive might
be changed to 44 Legacy Street if Legacy Drive does not exist in the EWS File. However, if
legacy Drive is in the EWS File, the address is left alone (no Plus 4) and coded with an EWS
Error. The next CASS database update should contain the new information and will code the
address correctly.

162 MatchUp
The EWS file is updated weekly every Thursday. To get the most current file, you can
download it from ftp://www.melissadata.com/updates/ews.txt. Place the ews.txt file in the
same folder as your other CASS Databases (mdAddr.dat, mdAddr.lic, etc.) overwriting the
previous week's file.

Status Codes:
Code Description
D Demo Mode If processing with the demo CASS database, records not in the
state of Nevada will be coded with "D".
E Expired Database The CASS database has expired.
S Standardized but Not Coded Standardization means that some conversion
was done on the address (for example, changing "Post Office Box" to "PO
Box").
V Street Number Validated to DPV Level This record has been DPV validated.
You can check the DPV Footnotes for more information about the level of
validation.
X Address Not Coded Check the Error code and/or DPV Footnotes for more
information.
7 Multiple Matches There were multiple matches for the address that were all in
the same Zip Code and Carrier Route. The returned Zip Code and Carrier Route
will be correct but you will not get any Plus 4 information.
9 Fully Coded The address was fully CASS coded.

Setup 163
164 MatchUp
5 Processing
5.1 Merge

Processing Statistics:
Elapsed Time How much time has elapsed during processing.
Estimated Time Remaining Approximately how much longer processing will take.
Estimated Completion Approximately when the job will be done.
Total Records Total number of records in the table being processed.
Records Processed Number of records that have been processed.
Records per Hour Number of records processed per hour (your mileage may vary).

Processing 165
Merging Statistics:
Source Lists each table that was processed.
Total Total number of records in this table.
Processed The total number of records output from all source tables.
Filtered Number of records filtered from this table (Input Filter is specified under
Information about).

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

166 MatchUp
Processing 167
5.2 Merge/Purge

Processing Statistics:
Elapsed Time How much time has elapsed during processing.
Estimated Time Remaining Approximately how much longer processing will take.
Estimated Completion Approximately when the job will be done.
Total Records Total number of records in the table being processed.
Records Processed Number of records that have been processed.
Records per Hour Number of records processed per hour (your mileage may vary).

Matching Statistics:
Source Lists each table that was processed.
Total Total number of records in this table.
Processed The number of records processed in this table.
Filtered Number of records filtered from this table (Input Filter is specified under
Advanced).
Output Number of records in this table that were unique or selected as the output
record in a group of duplicates.
168 MatchUp
 Duplicates Number of records in this table that were duplicates of other records - in a
group of duplicate records, these records were not selected as the output record, i.e.,
didn't end up in the Output table.
Suppressions Number of records in this table that matched against records in any
suppression list. Suppression records themselves don't contribute to this count.
Non-Intersections Number of records in this table that didn't match against any record
in any intersection list. Intersection records themselves don't contribute to this count.
CASS Coded Number of records that were CASS certified.
DPV Coded Number of records that were Delivery Point Validated.

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

Processing 169
170 MatchUp
5.3 Purge

Processing Statistics:
Elapsed Time How much time has elapsed during processing.
Estimated Time Remaining Approximately how much longer processing will take.
Estimated Completion Approximately when the job will be done.
Total Records Total number of records in the table being processed.
Records Processed Number of records that have been processed.
Records per Hour Number of records processed per hour (your mileage may vary).

Matching Statistics:
Source Lists each table that was processed.
Total Total number of records in this table.
Processed The number of records processed in this table.
Filtered Number of records filtered from this table (Input Filter is specified under
Advanced).
Output Number of records in this table that were unique or selected as the output
record in a group of duplicates.

Processing 171
 Duplicates Number of records in this table that were duplicates of other records - in a
group of duplicate records, these records were not selected as the output record, i.e.,
didn't end up in the Output table.
Suppressions Number of records in this table that matched against records in any
suppression list. Suppression records themselves don't contribute to this count.
Non-Intersections Number of records in this table that didn't match against any record
in any intersection list. Intersection records themselves don't contribute to this count.

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

172 MatchUp
Processing 173
5.4 File Update

Processing Statistics:
Elapsed Time How much time has elapsed during processing.
Estimated Time Remaining Approximately how much longer processing will take.
Estimated Completion Approximately when the job will be done.
Total Records Total number of records in the table being processed.
Records Processed Number of records that have been processed.
Records per Hour Number of records processed per hour (your mileage may vary).

Matching Statistics:
Source Lists each table that was processed.
Total Total number of records in this table.
Processed The number of records processed in this table.
Filtered Number of records filtered from this table (Input Filter is specified under
Advanced).
Output Number of records in this table that were unique or selected as the output
record in a group of duplicates.

174 MatchUp
 Duplicates Number of records in this table that were duplicates of other records - in a
group of duplicate records, these records were not selected as the output record, i.e.,
didn't end up in the Output table.
Suppressions Number of records in this table that matched against records in any
suppression list. Suppression records themselves don't contribute to this count.
Non-Intersections Number of records in this table that didn't match against any record
in any intersection list. Intersection records themselves don't contribute to this count.

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

Processing 175
176 MatchUp
5.5 CASS

Processing Statistics:
Elapsed Time How much time has elapsed during processing.
Estimated Time Remaining Approximately how much longer processing will take.
Estimated Completion Approximately when the job will be done.
Total Records Total number of records in the table being processed.
Records Processed Number of records that have been processed.
Records per Hour Number of records processed per hour (your mileage may vary).

CASS Statistics:

Processing 177
 5-Digit Coded Records having a valid 5-digit Zip Code.
Zip+4 Coded Records coded with a Plus 4 extension.
CASSmate Match Records coded with a Plus 4 extension through the assistance of
CASSMATE. This count is included in the Zip+4 Coded total.
CRRT Coded Records coded with a Carrier Route.
DPV Coded Records with a Delivery Point verified address.

CASS Errors:
Multiple Matches (M) More than one record matches the address and there is not
enough information available in the input address to break the tie between multiple
records. Passing more complete information, such as city names or urbanization names,
can help reduce the number of multiple match errors.
No Street Data (N) The Zip Code exists but no streets begin with the same letter in that
Zip Code.
Range Error (R) The address was found but the street number in the input address was
not between the low and high range in the CASS database.
Component Mismatch (T) Either the directionals or the suffix field did not match the
CASS database, and there was more than one choice for correcting the address. For
example, if the given address was "100 Main St" and the only addresses found were
"100 E Main St" and 100 Main Ave" the error code "T" would be returned because we do
not know whether to add the directional "E" or to change the suffix to "Ave".
Unknown Street (U) An exact street name match could not be found and phonetically
matching the street name resulted in either no matches or matches to more than one
street name.
Non-Deliverable (X) The physical location exists but there are no homes on this street.
One reason might be that there are railroad tracks or a river running alongside this
street, as they would prevent construction of homes in this location.
Zip Code Error (Z) The Zip Code does not exist and could not be determined by the city
and state.

DPV Footnotes:
Zip+4 Coded (AA) Input address matched to the Zip+4 file
Zip+4 Not Coded (A1) Input address not matched to the Zip+4 file
DPV Coded (BB) Input address matched to DPV (all components)
Invalid Secondary (CC) Input address primary number (street number) matched to
DPV but secondary number not matched
Missing Secondary (N1) Input address primary number matched to DPV but secondary
number missing
Missing Number (M1) Input address primary number missing
Invalid Number (M3) Input address primary number invalid
Missing PO/RR/HC (P1) Input address PO, RR or HC box number missing
Invalid PO/RR/HC (P3) Input address PO, RR or HC box number invalid
CMRA Coded (RR) Input address matched to CMRA

178 MatchUp
 Missing Secondary (R1) Input address matched to CMRA but secondary number not
present
Military Coded (F1) Address was coded to a military address
Genl Delivery Coded (G1) Address was coded to a General Delivery address
Unique Zip Coded (U1) Address was coded to a unique Zip Code

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

Processing 179
180 MatchUp
6 Analyze and Report
Analyzing gives you the chance to view and correct errors that occurred during processing.
Analyze is not available for Merge or CASS setups.

Analyze and Report 181

6.1 Edit Results

Status This information is also available in plain English across the bottom of the Analyze
screen:
· U A unique record.
· O An output record that has dupes (i.e. is not unique).
· D A duplicate record that was rejected.
· S A suppression table record (neither output nor reject). A lowercase s means that this
record was used to suppress records. An uppercase S means that this record didn't
find any matching record to suppress.
· X A Regular, Self Purge, or No-Purge record that was suppressed (rejected).
· I An intersection table record (neither output nor reject). A lowercase i means that this
record was used to intersect records. An uppercase I means that this record didn't find
any matching record to intersect.
· Y A Regular, Self Purge, or No-Purge record that had no intersection hits (rejected).

Table The source table for this record.

Record The original record number for this record.

Matchcode Hits The matchcode combination(s) that caused this match. If you have the
mouse over this field, a more verbose representation will be displayed in a tooltip.
182 MatchUp
Key The matchcode for this record.

Dupe Group Every original record encountered by MatchUp gets a unique number. When a
duplicate record is found, it is assigned the same number as the original. This number is
known as the dupe group.

Colors:
Output, Rejected, Suppressed, and Non-Intersected records come in two background colors.
There's no difference between the two: we just use the color variation to "break up" duplicate
groups, much like green bar paper. The color coding can be changed in Analyze | Options.

Analyze:
Next Dupe Group Take you to the next group of duplicate records.
Locate Row Search for a text string in the matchcode key or any displayed column(s).
Locate Again Search for the next record matching the previously specified search
criteria.
Show/Hide Rows Select the type of records and matches you want to view in the
analyzer.
Sort Rows Alter the sequence of the displayed rows.
Mark as Output Moves the highlighted record to the Output file. You can also press "O"
on your keyboard to accomplish the same. The source column shows where the record
is currently stored. This option is disabled (dimmed) if you specified no Output file or the
record is already marked as output.
Mark as Duplicate Moves the highlighted record to the Duplicate file. You can also
press "D" on your keyboard to accomplish the same. The source column shows where
the record is currently stored. This option is disabled (dimmed) if you specified no Dupe
file or the record is already marked as duplicate.
Mark as Suppressed Moves the highlighted record to the Suppression file. You can
also press "S" on your keyboard to accomplish the same. The source column shows
where the record is currently stored. This option is disabled (dimmed) if you specified no
Suppression file or the record is already marked as suppressed.
Print Rows Print the analyzer results.
Vertical Display Display the analyzer results vertically.
Display Output Display the output record vertically.

Analyze and Report 183

Column:
Add Column Display a new column in the browser.
Remove Column Remove an existing column from the browser.
Change Column Change what data is displayed in a column.
Auto-Size Column Change the width of a column to fit the longest string.
Auto-Size All Columns Change the width of each column to fit its longest string.
All Columns This Size Change the width of each column to the size of a column.
Reset Columns Reset all column widths to their original sizes. Show/Hide settings are
also reset to show all records.

Options:
Settings Modify the analyzer's display settings.
No Toolbar Hide the analyzer's toolbar.
Small Toolbar Display the small toolbar.
Medium Toolbar Display the medium toolbar.
Large Toolbar Display the large toolbar.
Overview Toggle the narrow browser overview band that appears just to the right of the
vertical scrollbar.

184 MatchUp
6.1.1 Locate Row

Search for The word or phrase that you want to locate.

Ignore case Check if you want to locate the text regardless of the case.

Search Select how you would like to search:

· Matchcode Key Search the matchcode keys that were generated during processing
(ie, the Key column).
· Selected Column(s) The checked fields will be searched for the specified text.

Search from start Start searching from the first row.

Search from here Start searching from the current record.

Note:
Only rows that are currently being displayed are searched. If you have used Show/Hide to
hide some of the rows, you may not find what you're looking for.

Analyze and Report 185

6.1.2 Show/Hide Rows

Records from Regular, Self-Purge and No-Purge List(s):

· Unique Records Show records with no matches.
· Records having Duplicates Show all records that have matched other records.
· Duplicate Records Show duplicate records (this doesn't include suppressed or non-
intersected records).
· Suppressed Records Show records that were matched against a suppression list(s).
· Non-Intersected Records Show records that didn't match against any intersection
list(s).

186 MatchUp
Records from Suppression List(s):
· None Don't show any records from the suppression list(s).
· Only Hits Only show records that were actually used to suppress other records.
· All Records Show all records from the suppression list(s).

Records from Intersection List(s):

· None Don't show any records from the intersection list(s).
· Only Hits Only show records that were actually used to intersect other records.
· All Records Show all records from the intersection list(s).

Matchcode Combinations Check the matchcode combination(s) that you want to view.
Matches are filtered based on which combinations are checked as well as the selection:
· Show matches having ANY of the above checked combinations
· Show matches having ALL of the above checked combinations
· Show matches having EXACTLY the above checked combinations

Records from Sources Check the source table(s) you want to view.

Analyze and Report 187

6.1.3 Sort Rows

This option lets you display the records in the analyzer in your preferred order.

Sort by Select the sorting method:

· Dupe Group Order the records in groups of duplicates so that you can easily see and
compare them (this is the default).
· Matchcode Component(s) Order the records by the contents of the specified
component(s). To change this order, click a component and drag it to the desired
position. Sorting will be done first on the top component, then the second, etc.

188 MatchUp
6.1.4 Print Rows

Start at row Specify the number of the row that you want printing to start at.

Number of Rows to Print Specify the number of rows to be printed. If you enter "0", all rows
from the starting row will be printed.

Separate Dupe Groups Select what should visually separate one group of matching records
from the next:
· No separator
· Solid line
· Blank line

Additional Information A few more options:

· Page headings and page numbers
· Field headings
· Colors

Analyze and Report 189

Output Options Select the method of output:
· Output to Printer Output will be sent to the selected printer. You can select the Font
that will be used for printing. Additionally, you can click Properties to change the
printer's properties and Page Setup to change the page size and orientation.
· Output to File Output will be sent to the specified file. You can select the Format of
the file (dBASE III, Excel, ASCII SDF, ASCII Comma Delimited and ASCII Tab
Delimited).

Note:
Excel formatted files, while eye-appealing, can take a long time to print. If you're using this as
a proof of delivery to your customer, you may consider only providing the first 100 or so rows.

190 MatchUp
6.1.5 Vertical Display

Displays a record vertically.

Note:
You can't edit records here.

Keyboard Shortcuts:
PAGEUP, PAGEDOWN Move to the previous/next row.
CTRL+HOME, CTRL+END Move to the first/last row.

Analyze and Report 191

6.1.6 Output Display

Displays an output, duplicate, suppression or non-intersection record vertically. This

displays the output record that was created during processing. This is an excellent way to
ensure that your gather and scatter settings are correct.

Note:
You can't edit records here.

Keyboard Shortcuts:
PAGEUP, PAGEDOWN Move to the previous/next row.
CTRL+HOME, CTRL+END Move to the first/last row.

192 MatchUp
6.1.7 Add Column

Heading/Size The name and size of the column you want to create.

Column Contents What should be displayed in the column:

· Status Display the disposition of the record.
· Table Display which table this record came from. You can elect to display the full table
name, its number or letter.
· Record Number Display the source table's record number.
· Matchcode Hit Combination(s) Display the matchcode combination(s) which caused
this record to match others (see also Matchcode Combinations).
· Matchcode Key Display the record's matchcode key.
· Dupe Group Number Display the record's dupe group number.
· Matchcode Mapped Data Display data as it was mapped into the matchcode. Select
the matchcode component that contains the data you want to display. This is data as it
is used in the matchcode (i.e., after any necessary conversion has been made).
· Source Data Display data directly from the source table(s). Select the field from each
input table you want to display.

Analyze and Report 193

Status Codes:
U A unique record.
O An output record that has dupes (i.e. is not unique).
D A duplicate record that was rejected.
S A suppression table record (neither output nor reject). A lowercase s means that this
record was used to suppress records. An uppercase S means that this record didn't
find any matching record to suppress.
X A Regular, Self Purge, or No-Purge record that was suppressed (rejected).
I An intersection table record (neither output nor reject). A lowercase i means that this
record was used to intersect records. An uppercase I means that this record didn't find
any matching record to intersect.
Y A Regular, Self Purge, or No-Purge record that had no intersection hits (rejected).

194 MatchUp
6.1.8 Change Column

Heading/Size The name and size of the column.

Column Contents What should be displayed in the column:

· Status Display the disposition of the record.
· Table Display which table this record came from. You can elect to display the full table
name, its number or letter.
· Record Number Display the source table's record number.
· Matchcode Hit Combination(s) Display the matchcode combination(s) which caused
this record to match others (see also Matchcode Combinations).
· Matchcode Key Display the record's matchcode key.
· Dupe Group Number Display the record's dupe group number.
· Matchcode Mapped Data Display data as it was mapped into the matchcode. Select
the matchcode component that contains the data you want to display. This is data as it
is used in the matchcode (i.e., after any necessary conversion has been made).
· Source Data Display data directly from the source table(s). Select the field from each
input table you want to display.

Analyze and Report 195

6.1.9 Settings

You can change both the display and printer fonts for the Analyzer at the top of this screen.

The rest of the screen is devoted to the color schemes you want to use on the Analyze
screen. Many types of records have an alternate color that you can select. There is
absolutely no difference between the original and the alternate. The only reason they exist is
to provide alternating colors as a visual aid.

196 MatchUp
Analyze and Report 197
6.2 View Reports
Depending on your processing options, MatchUp produces up to 18 different reports. The
reports can be viewed on screen through Analyze | Reports, as well as printed in a variety
of ways.

Reports:
Processing Summary
File Summary
Multiples
Matrices:
Dupe Matrix
Multi-Buyer Dupe Matrix
Quality Matrix
Multi-Buyer Quality Matrix
Matchcode:
Matchcode Matrix
Matchcode Quality
Source Codes:
File Summary
Multiples
Matrices:
Dupe Matrix
Multi-Buyer Dupe Matrix
Quality Matrix
Multi-Buyer Quality Matrix
Merge Matrix
CASS

198 MatchUp
Printing:
When you click the Print button, the following dialog is displayed. It doesn't matter which
report you are currently viewing, the Report list is always the same (i.e., you can print one
report while viewing another).

Report Format Select the type of report you want to create: Excel, Word or Template (see
below).

Print Report(s) Check the report(s) that you want to create. The contents of this list will
depend on the selected Report Format as well as the type of processing (CASS,
Merge/Purge, etc).

Print Charts and Graphs Whether or not you want to waste your ink on the pretty charts
and graphs.

Analyze and Report 199

Font/Size The font to be used for printing. Although Excel and Word formatted reports can
be printed with any font, Template formatted reports should be printed with a mono-spaced
(a.k.a. fixed-pitch) font such as Courier New.

Output to Printer Output will be sent to the selected printer. You can select the Font that
will be used for printing. Additionally, you can click Properties to change the printer's
properties and Page Setup to change the page size and orientation.

Output to File Output will be sent to the specified file. For Template formatted reports, this
means no charts and graphs.

Report Formats:
· Excel The Excel format will produce a report generated through Microsoft Excel. If you
send this report to a file, you will find that this produces a single spreadsheet, with
tabbed sheets (one for each report). Excel 2000 or later must be installed on your
computer.
· Word The Word format will produce a report generated through Microsoft Word. If you
send this report to a file, you will find that this produces a single document, each report
starting on a new page. The page orientation (portrait or landscape) may change from
page to page. Word 2000 or later must be installed on your computer.
· Template The Template format will produce a report generated through MatchUp.
This format is nearly the same as the one used in DoubleTake 2 (in fact, DoubleTake
2 report templates will still work with no modification in most cases). If your computer
doesn't have Word or Excel installed, this is the way to go.

200 MatchUp
6.2.1 Processing Summary

The Processing Summary contains processing statistics and information about your setup.
Note that not all statistics are shown for all processes.

Setup File The name of the setup file used for processing.

Processing Statistics:
· Date Processed The latest processing date.
· Elapsed Time How long processing took.
· Total Records Processed The total number of records run through the process.
· Records per Hour The number of records processed in an hour.
· Maximum Records per Cluster Largest cluster.
· Minimum Records per Cluster Smallest cluster.
· Average Records per Cluster The average number of records in each cluster.
· Clusters Processed How many clusters were used in the process (see Concepts for
information on clustering).
· Clustering Key Size The maximum number of common characters in each cluster
load.

Output Tables:
· Output
· Duplicate
· Suppressed

Analyze and Report 201

· Non-Intersected

202 MatchUp
6.2.2 File Summary

The File Summary contains a detailed listing of what happened with each table. Counts for
Output, Duplicates, Suppressions, etc. are given.

Source Lists each table that was processed.

Table Type Indicates the type of table processed: Regular, Suppression, Intersection, Self
Purge, or No Purge.

Total Total number of records in this table.

Processed Number of records processed from this table.

Filtered Number of records filtered from this table.

Output Number of records in this table that were unique or selected as the output record in a
group of duplicates.

Duplicates Number of records in this table that were duplicates of other records, i.e., didn't
end up in the Output table.

Suppressions Number of records in this table that matched against records in a

suppression list. Suppression records themselves don't contribute to this count.

Analyze and Report 203

Not Intersected Number of records in this table that didn't match against any record in any
intersection list. Intersection records themselves don't contribute to this count.

Inter-File Number of dupe records from other tables that matched records output from this
table.

Intra-File Number of dupe records from this table that matched records output from this
table.

Multi-Buyer Number of records output from this table that matched against records from
another table. See Report Settings on the different ways that this statistic can be calculated.

How the Numbers Add Up:

· Output + Not Intersected + Suppressions + Duplicates = Processed
· Processed + Filtered = Total

204 MatchUp
6.2.3 Multiples

The Multiples report gives you counts of multiple dupe counts for each source table.

Source Lists each table that was processed.

1x Number of unique records (being unique, they ended up in the Output table).

2x Number of records that were selected as the output record (and ended up in the Output
table) in a group of two duplicates.

3x-10x Number of records that were selected as the output record in a group of three, four,
etc. duplicates.

10+x Number of records that were selected as the output record in a group of eleven or
more duplicates (we had to stop sometime).

How the Numbers Add Up:

· 1x + 2x + 3x + 4x + 5x + 6x + 7x + 8x + 9x + 10x + 10+x = Output
· 2x + (3x*2) + (4x*3) + (5x*4) + (6x*5) + (7x*6) + (8x*7) + (9x*8) + (10x*9) = Total
Dupes

The last equation only applies for the Total row. Also, if you have any 10+x counts, the
formula can't be used, as we have no way of knowing its multiplier.

Analyze and Report 205

6.2.4 Matrices
Matrices: Dupe Matrix

The Dupe Matrix gives record counts of inter- and intra-file dupes. The pie chart illustrates
the highlighted row.

Detailed explanation:
· For all records output from national, there were 7538 dupes of these records in
national (intra-file dupes), 130 dupes in in inhouse (inter-file dupes), and 0 in any other
table.
· For all records output from inhouse, there were 18 dupes of these records in national
(inter-file dupes), 3512 dupes in inhouse (intra-file dupes), and 0 in any other table.
· For all chosen records from nonprofit (ending up in the Output File), there were 46
dupes of these records in nonprofit, and 494 in household.

The important thing to remember about this report is that all dupes are counted with respect
to which record was output. This is the key difference between this report and the Quality
Matrix. The cells where a file intersects itself (the diagonal running from top left to bottom
right) are commonly known as the intra-list counts, whereas the others are known as inter-list
counts ("intra" meaning "within" and "inter" meaning "between")

206 MatchUp
Still don't get it? Let's try this example:

Demo1 Demo2 Demo3

Joe Joe Joe
Joe Joe Marge
Joe Melissa
Melissa Melissa
Marge
Marge
The bold entries indicate which records were selected as output records. A Dupe Matrix for
this run would look like:

Demo1 Demo2 Demo3

Demo1 2 2 1 2 dupe Joes in 1, 2 in 2, 1 in 3
Demo2 1 1 0 1 dupe Melissa in 1, 1 in 2, 0 in 3
Demo3 2 0 0 2 dupe Marges in 1, 0 in 2, 0 in 3
Notice that the output records themselves are not counted, just duplicates. That's why we
called it the Dupe Matrix.

How the Numbers Add Up:

Each column of numbers add up to the Dupe count for that column's table in the File
Summary.

One thing that is not immediately obvious is that a chart's counts are often "crowded"
towards a corner. This is very typical if Demo1 is ranked highest in priority, Demo3 lowest.
Why? Consider the case when there's a pair of duplicates, one from Demo1, one from
Demo3. Because of the ranking, Demo1 gets output, Demo3 duped. So the Demo3 row will
never have any counts in Demo1, because there will never be a situation where Demo3 is
output and Demo1 is duped.

Analyze and Report 207

Matrices: Multi-Buyer Dupe Matrix

The Multi-Buyer Dupe Matrix gives record counts of inter- and intra-file dupes. The
difference between this report and the Dupe Matrix is that in this report no more than one
record is counted per source. The pie chart illustrates the highlighted row.

Detailed explanation:
· For all records output from national, there were 3087 records with at least one
matching record from national. There were 125 records output from national that had
at least one additional matching record from inhouse.
· For all records output from inhouse, there were 10 records with at least one matching
record from national. There were 1731 records output from inhouse that had at least
one additional matching record from inhouse.
· For all chosen records from nonprofit (ending up in the Output File), 46 of these
records had one or more additional record also in nonprofit, and 449 in household.

The important thing to remember about this report is that all dupes are counted with respect
to which record was output and that duplicates coming from the same source table are only
tallied once per dupe group.

208 MatchUp
Still don't get it? Let's try this example:

Demo1 Demo2 Demo3

Joe Joe Joe
Joe Joe Marge
Joe Melissa
Melissa Melissa
Marge
Marge
The bold entries indicate which records were selected as output records. The dupe groups
would look like:

Group 1 Group 2 Group 3

Joe (Demo1) Melissa (Demo2) Marge (Demo3)
Joe (Demo1) Melissa (Demo1) Marge (Demo1)
Joe (Demo1) Melissa (Demo2) Marge (Demo1)
Joe (Demo2)
Joe (Demo2)
Joe (Demo3)
A Multi-Buyer Dupe Matrix for this run would look like:

Demo1 Demo2 Demo3

Demo1 1 1 1 Group 1 has matches in Demo1, Demo2 and Demo3
Demo2 1 1 0 Group 2 has matches in Demo1 and Demo2
Demo3 1 0 0 Group 3 has matches in Demo1
How the Numbers Add Up:
There's no correlation between these numbers and any other report's numbers. However,
you can be sure that these numbers will be less than or equal to the numbers in the Dupe
Matrix.

Analyze and Report 209

What's so Great About this Report?
The Multi-Buyer Dupe Matrix gives you an unbiased indication of how two lists intersect. A
high number in a given cell would indicate that the two lists have a number of common
records. Although the Dupe Matrix (non-Multi-Buyer) also reflects such a situation, the Dupe
Matrix can be fooled by lists with high internal duplicate counts. For example, say Demo1
versus Demo2 has a Dupe Matrix count of 3000 (possibly indicating lots of common contacts
between these two lists). However, the Multi-Buyer Dupe Matrix count is only 200. This
would indicate that the lists are not very similar, and that the Demo2 list has a lot of internal
dupes (which skewed the Dupe Matrix count).

210 MatchUp
Matrices: Quality Matrix

The Quality Matrix gives duplicate counts regardless of where an output record came from.
The pie chart illustrates the highlighted row.

For example:

Demo1 Demo2 Demo3

Demo1 39 69 22
Demo2 70 52 73
Demo3 22 72 46
· There are 39 records in Demo1 that match other records in Demo1, 69 records in
Demo2 that match records in Demo1, and 22 records in Demo3 that match records in
Demo1.
· There are 70 records in Demo1 that match records in Demo2, 52 records in Demo2
that match other records in Demo2, and 73 records in Demo3 that match records in
Demo2.
· There are 22 records in Demo1 that match records in Demo3, 72 records in Demo2
that match records in Demo3, and 46 records in Demo3 that match other records in
Demo3.

The intra- and inter- terminology does not apply to the Quality Matrix, as the counts are not
generated with any regard to which record was output. What this report does tell you is the

Analyze and Report 211

"cross-saturation" between tables. Often, this is hidden in the Dupe Matrix. For example,
using the sample report seen in the Dupe Matrix, a count of 0 is given in the Demo3 versus
Demo2 cell. But here, we see a 72 - meaning that there are 72 records in common between
the two tables. This count was "hidden" in the Dupe Matrix because the ranking caused all of
these matches to accumulate into the Demo1 or Demo2 rows (because Demo1 was
prioritized over Demo2 which was prioritized over Demo3)

This is a tough report to understand, so let's try this example (it's the same set of records
used in the Dupe Matrix example):

Demo1 Demo2 Demo3

Joe Joe Joe
Joe Joe Marge
Joe Melissa
Melissa Melissa
Marge
Marge
No indicator of which record was selected as the output record is given for this report for the
sake of clarity (you can assume that they are the same as the ones in the Dupe Matrix
example). A Quality Matrix for this run would look like:

Demo1 Demo2 Demo3

Demo1 3 4 22 dupe Joes + 1 dupe Marge in 1
2 dupe Joes + 2 dupe Melissa in 2
1 dupe Joe + 1 dupe Marge in 3
Demo2 4 2 1 3 dupe Joes + 1 dupe Melissa in 1
1 dupe Joe + 1 dupe Melissa in 2
1 dupe Joe in 3
Demo3 5 2 0 3 dupe Joes + 2 dupe Marges in 1
2 dupe Joes in 2
No dupes of anything in 3
Basically, you look at each unique record in a table and dedupe it against everything else.
The unique record does not get figured into the counts (just duplicates of it). Using Brand X
software, you would have to run 3 dedupe runs, assigning different priorities each time: first
to Demo1, then to Demo2, and finally to Demo3.

How the Numbers Add Up:

Good luck! Try as you might, there is no correlation between the numbers produced in the
Quality Matrix and any other counts. Don't call us and ask us if we've "fixed" this report, or
made some sort of scientific advance - the numbers don't add up because of what they
represent; there's nothing wrong with your television.

212 MatchUp
Matrices: Multi-Buyer Quality Matrix

The Multi-Buyer Quality Matrix gives duplicate counts regardless of where an output record
came from. The difference between this report and the Quality Matrix is that in this report no
more than one record is counted per source. The pie chart illustrates the highlighted row.

For example:

Demo1 Demo2 Demo3

Demo1 39 69 22
Demo2 69 52 72
Demo3 22 72 46
· There are 39 dupe groups with 2 or more records from Demo1. 69 dupe groups with
records from Demo1 and Demo2, and 22 dupe groups with records from Demo1 and
Demo3.
· There are 69 dupe groups with records from Demo2 and Demo1, 52 dupe groups with
2 or more records from Demo2, and 72 dupe groups with records from Demo2 and
Demo3.
· There are 22 dupe groups with records from Demo3 and Demo1, 72 dupe groups with
records from Demo3 and Demo2, and 46 dupe groups with 2 or more records from
Demo3.

Analyze and Report 213

What this report does tell you is the "cross-saturation" between tables. Often, this is hidden in
the Dupe Matrix. For example, using the sample report seen in the Dupe Matrix, a count of 0
is given in the Demo3 versus Demo2 cell. But here, we see a 72 - meaning that there are 72
records in common between the two tables. This count was "hidden" in the Dupe Matrix
because the ranking caused all of these matches to accumulate into the Demo1 or Demo2
rows (because Demo1 was prioritized over Demo2 which was prioritized over Demo3)

In some ways, this report can be more useful than the Quality Matrix. Because intra-file
dupes are not counted, lists with lots of internal dupes do not appear to have more 'real'
matches than they actually do.

This is a tough report to understand, so let's try this example (it's the same set of records
used in the Dupe Matrix example):

Demo1 Demo2 Demo3

Joe Joe Joe
Joe Joe Marge
Joe Melissa
Melissa Melissa
Marge
Marge
The dupe groups would look like:

Group 1 Group 2 Group 3

Joe (Demo1) Melissa (Demo2) Marge (Demo3)
Joe (Demo1) Melissa (Demo1) Marge (Demo1)
Joe (Demo1) Melissa (Demo2) Marge (Demo1)
Joe (Demo2)
Joe (Demo2)
Joe (Demo3)

214 MatchUp
A Multi-Buyer Quality Matrix for this run would look like:

Demo1 Demo2 Demo3

Demo1 2 2 2 Groups 1 and 3 have double matches in Demo1
Groups 1 and 2 have records from both Demo1 and
Demo2
Groups 1 and 3 have records from both Demo1 and
Demo3
Demo2 2 2 1 Groups 1 and 2 have records from both Demo2 and
Demo1
Groups 1 and 2 have double matches in Demo2
Group 1 has records from both Demo2 and Demo3
Demo3 2 1 0 Groups 1 and 3 have records from both Demo3 and
Demo1
Group 1 has records from both Demo3 and Demo2
No group has double matches in Demo3
How the Numbers Add Up:
They don't; there's no correlation between these numbers and any other report's numbers.
However, you can be sure that these numbers will be less than or equal to the numbers in
the Quality Matrix.

What's so Great About this Report?

The Multi-Buyer Quality Matrix gives you an unbiased indication of how two lists intersect. A
high number in a given cell would indicate that the two lists have a number of common
records. Although the Quality Matrix (non-Multi-Buyer) also reflects such a situation, the
Quality Matrix can be fooled by lists with high internal duplicate counts. For example, say
Demo1 versus Demo2 has a Quality Matrix count of 3000 (possibly indicating lots of
common contacts between these two lists). However, the Multi-Buyer Quality Matrix count is
only 100. This would indicate that the lists are not very similar, and that the Demo2 list has a
lot of internal dupes (which skewed the Quality Matrix count).

This report also has an advantage over the Multi-Buyer Dupe Matrix, in that the output record
doesn't skew the counts in any way. Because of this, commonality between two lists that is
obscured by file ranking can be seen.

Analyze and Report 215

6.2.5 Matchcode
Matchcode: Matchcode Matrix

The Matchcode Matrix lists the number of matches made based on each matchcode
combination. The bar chart illustrates the highlighted row, while the pie chart illustrates the
highlighted column.

For example, say combination 1 is Zip+LN+POB and combination 2 is Zip+LN+S#+SN

(combinations 3 through 16 weren't used). Using the example below:

1 2
Demo1 20 174
Demo2 21 225
Demo3 15 175

216 MatchUp
 · There were 20 records in Demo1 that matched other records (in any of the three
tables) using the first combination (Zip+LN+POB). There were 174 records in Demo1
that matched other records using the second combination (Zip+LN+S#+SN).
· There were 21 records in Demo2 that matched other records using the first
combination (Zip+LN+POB) and 225 records in Demo2 that matched other records
using the second combination (Zip+LN+S#+SN).
· There were 15 records in Demo3 that matched other records using the first
combination and 175 records in Demo3 that matched other records using the second
combination.

Note that the output record is not used in these calculations. Also, if matches are made using
both combinations, both columns' counts will reflect that match.

What this report can tell you is how well a matchcode combination is working. If a particular
combination is generating low counts, that combination may not be working for you. Don't
regard these counts as an exact indicator of performance, though. In our above example, we
certainly should expect the PO Box combination to generate substantially less hits than the
Street Address combination - after all, a large part of our population likes mail delivered to
the door!

Here's an example of how this report is generated (bold indicating the output record).
Warning: this is really involved:

Demo1 Demo2 Demo3

123 Main 123 Main 123 Main
POB 10 POB 10 POB 67
POB 67/10 Main 10 Main POB 67/10 Main
During deduping, we end up with 3 dupe groups:

Group 1:
123 Main from Demo1 (had Street # matches)
123 Main from Demo2 (matched because of Street #)
123 Main from Demo3 (matched because of Street #)

Group 2:
POB 10 from Demo2 (matched because of PO Box)
POB 10 from Demo1 (matched because of PO Box)

Group 3:
POB 67/10 Main from Demo3 (had both Street # and PO Box matches)
POB 67/10 Main from Demo1 (matched because of both Street # and PO Box)
10 Main from Demo2 (matched because of Street #)
POB 67 from Demo3 (matched because of PO Box)

Analyze and Report 217

The Matchcode Matrix for this run would look like (assume combination 1 is Zip+LN+POB
and combination 2 is Zip+LN+S#+SN):

1 2
Demo1 2 2Group 1: 1 Street #
Group 2: 1 POB
Group 3: 1 POB and Street #
Demo2 1 2 Group 1: 1 Street #
Group 2: 1 POB
Group 3: 1 Street #
Demo3 2 2 Group 1: 1 Street #
Group 2:
Group 3: 2 POBs and 1 Street #
Notice how the "POB 67/10 Main" produced a hit in both columns. Even if MatchUp finds a
match due to combination 1, it will still check out combination 2.

How the Numbers Add Up:

Sometimes they do. The main reason that they usually don't is the case when a record
matches other record(s) using more than one combination - a "double hit" is accumulated.
But, when this is not the case (and in the Street #/PO Box situation, this is often true), this
will work:

· Total Output - Total 1x + Total Dupe = Sum of Combinations

If this equation doesn't apply, it means you have some multiple combination matches. The
accumulation of multiple combination matches was intentional. If we wanted these numbers
to always add up, we would have had to decide which combination(s) should be left
uncounted in multiple combination matches. It was more important to us to generate
accurate counts rather than give you that "everything adds up" warm and fuzzy feeling. If that
bugs you, feel free to "fudge" the numbers. We won't tell (unless someone asks).

218 MatchUp
Matchcode: Matchcode Quality

The Matchcode Quality report gives you counts of how well your records were populated
with the matchcode components. The bar graph illustrates the highlighted row.

This report is refreshingly simple. For each table, records having each matchcode
component are counted. These counts can indicate problems with a database (like missing
Zip Codes), as well as bad setup mappings (i.e., trying to map a First Name field as a Full
Name).

The counts are generated after any on-the-fly splitting has been performed, so don't be
alarmed because Street # and Street Name are listed even though your database doesn't
have these fields - this information was extracted from the specified address lines.

How the Numbers Add Up:

You would love to see each count match the number of records processed for that table.
That, of course, is a dream. Bad records are sure to screw things up and generate missing
Zip Codes, Street Names, and so on. But you can dream!

This is a great report to give to your customer when he claims that his data is pristine and
had been input by a sect of data entry monks, when in reality it was entered by a roomful of
monkeys trying to type Macbeth.

Analyze and Report 219

6.2.6 Source Codes

If you specified an Input Source Code Field for one or more of your input tables, Source
Code reports will be available.

The Input Source Code Field is specified on the Advanced tab in Setup | Edit and is specific
to each Input Table.

MatchUp uses the contents of your Input Source Code Field to generate report counts. This
gives you counts based on a set of keycodes contained in your tables.

This report is generated whenever you have specified at least one Source Code Field
(Merge/Purge Setup: Advanced or Purge Setup: Advanced). For each source code detected
in the process, a count is accumulated for each table. The pie chart illustrates the highlighted
column.

But, you ask, what about the times when I've only specified a Source Code Field for only
some of my files? They get their own special source code: "*SF" followed by the number of
the file from which the record came. Because of the leading "*", these codes will generally
appear at the top of the list.

Source codes are counted during the Purge pass. For this reason, source codes are always
based on post-filtered records (as filters are applied during the Key Building pass).

220 MatchUp
Source Code Lookup Table:
Sometimes source codes aren't very descriptive. You can provide MatchUp with a list of
source codes and their descriptions in the "Source Code Lookup Table" prompt. This list
must be a dBASE III table with at least two fields. The first field should contain the source
codes (as they appear in your data), the second field should contain the source code's
longer description. If you forget any source codes, the original (non-descriptive) source code
will be used instead.

How the Numbers Add Up:

· Total Source Codes for a File = Records Processed for that File

Analyze and Report 221

Source Codes: File Summary

The Source Code File Summary shows processing results by source code. It is just like the
File Summary with that one difference. Because this report is oriented to source codes, you
will only get it when you have specified at least one Source Code Field (Merge/Purge Setup:
Advanced or Purge Setup: Advanced)

Source Code Lists each source code that was processed.

Total Total number of records with this source code.

Output Number of records with this source code that were unique or selected as the output
record in a group of duplicates.

Duplicates Number of records with this source code that were duplicates.

Suppressions Number of records with this source code that were suppressed. Suppression
records themselves don't contribute to this count.

Not Intersected Number of records with this source code that didn't match against any
record in any intersection list. Intersection records themselves don't contribute to this count.

Inter-File Number of dupe records from other sources that matched records from this
source.

222 MatchUp
Intra-File Number of dupe records from this source that matched records output from this
source.

Multi-Buyer Number of records output from this table that matched against records from
another table. See Report Settings on the different ways that this statistic can be calculated.

Why no Table Type and Data Type columns? Because source codes aren't table based -
they can be scattered across many tables. How about the Processed and Filtered columns?
Because source codes are accumulated during the Purge pass and filters are applied during
the Key Building pass, all source codes counted have survived any filter conditions.

How the Numbers Add Up:

· Output + No Intersect + Suppress + Dupe = Total

Analyze and Report 223

Source Codes: Multiples

The Multiples report gives you counts of multiple dupe counts for each source table.

Source Code Lists each source that was processed.

1x Number of unique records (being unique, they ended up in the Output table).

2x Number of records that were selected as the output record (and ended up in the Output
table) in a group of two duplicates.

3x-10x Number of records that were selected as the output record in a group of three, four,
etc. duplicates.

10+x Number of records that were selected as the output record in a group of eleven or
more duplicates (we had to stop sometime).

How the Numbers Add Up:

· 1x + 2x + 3x + 4x + 5x + 6x + 7x + 8x + 9x + 10x + 10+x = Output
· 2x + (3x*2) + (4x*3) + (5x*4) + (6x*5) + (7x*6) + (8x*7) + (9x*8) + (10x*9) = Total
Dupes

The last equation only applies for the Total row. Also, if you have any 10+x counts, the
formula can't be used, as we have no way of knowing its multiplier.

224 MatchUp
Source Codes: Matrices
Matrices: Dupe Matrix

The Source Code Dupe Matrix gives record counts of inter- and intra-source dupes. The
pie chart illustrates highlighted row. This report is identical to the Dupe Matrix, except that
counts are based on source codes, rather than source tables.

This report is generated whenever you have specified at least one Source Code.

The important thing to remember about this report is that all dupes are counted with respect
to which record was output. This is the key difference between this report and the Quality
Matrix. The cells where a table intersects itself (the diagonal running from top left to bottom
right) are commonly known as the intra-list counts, whereas the others are known as inter-list
counts.

How the Numbers Add Up:

Each column of numbers add up to the Dupe count for that column's source code in the
Source Code File Summary.

Analyze and Report 225

Matrices: Multi-Buyer Dupe Matrix

The Source Code Multi-Buyer Dupe Matrix gives record counts of inter- and intra-file
dupes. The difference between this report and the Dupe Matrix is that in this report no more
than one record is counted per source.

This report is generated whenever you have specified at least one Source Code. See Multi-
Buyer Dupe Matrix for a detailed explanation about this report.

226 MatchUp
Matrices: Quality Matrix

The Source Code Quality Matrix gives duplicate counts regardless of where an output
record came from. The pie chart illustrates the highlighted row.

This report is identical to the Quality Matrix, except that counts are based on source codes,
rather than source tables.

How the Numbers Add Up:

Like the Quality Matrix, the Source Code Quality Matrix does not produce counts that will
"jibe" with any other numbers. They're just not supposed to.

Analyze and Report 227

Matrices: Multi-Buyer Quality Matrix

The Multi-Buyer Quality Matrix duplicate counts regardless of where an output record
came from. The difference between this report and the Quality Matrix is that in this report no
more than one record is counted per source.

This report is generated whenever you have specified at least one Source Code. See Multi-
Buyer Quality Matrix for a detailed explanation about this report.

228 MatchUp
6.2.7 Merge Matrix

The Merge Matrix contains a detailed listing of what happened with each table during a
Merge process.

The pie chart illustrates the highlighted table.

Source Lists each table that was processed.

Total Total number of records in this table.

Processed Number of records processed from this table.

Filtered Number of records filtered from this table.

Analyze and Report 229

6.2.8 CASS

Processing Statistics:
Setup File The name of the setup file used for processing.
Table The table processed.
Date Processed The latest processing date.
Elapsed Time How long processing took.
Records Processed The total number of records run through the process.

Counts:
5-Digit Coded Records having a valid 5-digit Zip Code.
Zip+4 Coded Records coded with a Plus 4 extension.
CASSmate Match Records coded with a Plus 4 extension through the assistance of
CASSMATE. This count is included in the Zip+4 Coded total.
CRRT Coded Records coded with a Carrier Route.
DPV Coded Records with a Delivery Point verified address.

230 MatchUp
Errors:
Multiple Matches (M) More than one record matches the address and there is not
enough information available in the input address to break the tie between multiple
records. Passing more complete information, such as city names or urbanization names,
can help reduce the number of multiple match errors.
No Street Data (N) The Zip Code exists but no streets begin with the same letter in that
Zip Code.
Range Error (R) The address was found but the street number in the input address was
not between the low and high range in the CASS database.
Component Mismatch (T) Either the directionals or the suffix field did not match the
CASS database, and there was more than one choice for correcting the address. For
example, if the given address was "100 Main St" and the only addresses found were
"100 E Main St" and 100 Main Ave" the error code "T" would be returned because we do
not know whether to add the directional "E" or to change the suffix to "Ave".
Unknown Street (U) An exact street name match could not be found and phonetically
matching the street name resulted in either no matches or matches to more than one
street name.
Non-Deliverable (X) The physical location exists but there are no homes on this street.
One reason might be that there are railroad tracks or a river running alongside this
street, as they would prevent construction of homes in this location.
Zip Code Error (Z) The Zip Code does not exist and could not be determined by the city
and state.

DPV Footnotes:
Zip+4 Coded (AA) Input address matched to the Zip+4 file
Zip+4 Not Coded (A1) Input address not matched to the Zip+4 file
DPV Coded (BB) Input address matched to DPV (all components)
Invalid Secondary (CC) Input address primary number (street number) matched to
DPV but secondary number not matched
Missing Secondary (N1) Input address primary number matched to DPV but secondary
number missing
Missing Number (M1) Input address primary number missing
Invalid Number (M3) Input address primary number invalid
Missing PO/RR/HC (P1) Input address PO, RR or HC box number missing
Invalid PO/RR/HC (P3) Input address PO, RR or HC box number invalid
CMRA Coded (RR) Input address matched to CMRA
Missing Secondary (R1) Input address matched to CMRA but secondary number not
present
Military Coded (F1) Address was coded to a military address
Genl Delivery Coded (G1) Address was coded to a General Delivery address
Unique Zip Coded (U1) Address was coded to a unique Zip Code

Analyze and Report 231

Statistics:
High Rise Default The number of multi-delivery point addresses (apartment/suite)
where the apartment/suite number is not present or is invalid.
High Rise Exact The number of multi-delivery point addresses (apartment/suite) where
the apartment/suite number is present and matches a ZIP+4 range.
Rural Route Default The number of Rural or Highway Contract Route addresses where
the box number is not present or is invalid.
Rural Route Exact The number of Rural Route or Highway Contract addresses where
the box number is present and matches a ZIP+4 range.
LACS Locatable Address Conversion Service. The total number of records which were
flagged as having undergone a change to a city-style address (to allow emergency
services to locate these addresses more efficiently). The address change is not reflected
here.
EWS Early Warning System Data. Count of new addresses scheduled to be included in
the USPS national database, new construction projects for example, which were flagged
by the EWS file.

CASS database date The date of your CASS database files.

CASS database expires The expiration date of your CASS database files.

232 MatchUp
6.2.9 Report Settings

The report settings dialog gives you the ability to change many of the ways MatchUp
presents its reports. Of most interest is the General branch, which allows you to change the
display and printer font, as well as specify the bitmap file that should be used as a logo at the
top of your Template format reports.

The Multi-Buyer count that appears on the File Summary and Source Code File Summary
can be reported in a variety of ways, as shown in the examples on the screen.

Analyze and Report 233

The other branches on this dialog allow you to change the colors used in display the various
charts and graphs on the reporting screen.

234 MatchUp
6.2.10 Report Templates

All of MatchUp's reports can be modified to a certain extent. Each format (Excel, Word, or
Template) is handled differently. For more information on this topic, please see the Help files
section on Report Templates.

Report Templates: Multi-Buyer Counting

The multi-buyer statistic is relatively new to MatchUp. One of the problems we had with
introducing this statistic is that there are several differing opinions about exactly what this
statistic should represent. When creating a Report Template, you should use the statistic that
applies to your situation.

In the examples below, assume that we are talking about a single set of 6 matching records,
from 3 different tables (FileA, FileB, and FileC). The output record is from FileA. Bold
records indicates records that are counted in the Multi-Buyer statistic.

Multi-Buy 0:
This is what most people think of when you say "Multi-Buyer Count". It is simply a total of the
number of output records from a table that have dupes from another table.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 1.

Multi-Buy 1:
Records coming from the same table as the output record are not counted.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 3.

Multi-Buy 2:
This is a popular alternative to Multi-Buy 1. It is generated the same way with the exception
that records coming from the same table are not counted more than once.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 2.

Multi-Buy 3:
This spin-off is similar to Multi-Buy 2 except that dupe records coming from the same table
as the output record are counted.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 3.

Multi-Buy 4:
Most people don't consider this count a multi-buyer count at all. It is generated the same way
as Multi-Buy 3 except that multiple records from the same table are counted.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 5

Analyze and Report 235

Multi-Buy 5:
This one isn't really a multi-buyer count either. Basically, all records in the dupe group are
counted.
FileA FileA FileA FileB FileB FileC Multi-Buyer Count is 6.

Multi-Buy:
The MultiBuy macro having no number tells MatchUp to use whatever multi-buyer setting
has been selected in Analyze | Reports | Settings. This is the selection provided in
MatchUp's 'canned' templates.

236 MatchUp
Analyze and Report 237
7 Tools
7.1 Browse

Tools | Browse allows you to view and modify a table's contents. Each row of the table
represents one record.

Depending on the type of table you are browsing, you may or may not be allowed to edit its
contents. If MatchUp cannot allow you to edit records, you will be warned when the browser
window is first displayed.

Many of the Browse commands use a dBASE expression in one way or another. Even if your
table is not a dBASE formatted table (i.e. Access, Excel, ASCII, etc), a dBASE expression
should be used. MatchUp "corrects" the statement appropriately.

238 MatchUp
Keyboard Shortcuts:
LEFT, RIGHT Move the highlight left/right one column.
CTRL+LEFT, CTRL+RIGHT Pan the browser left/right one column.
HOME, END Move the highlight to the first/last column.
UP, DOWN Move the highlight up/down one row.
CTRL+UP, CTRL+DOWN Pan the browser up/down one row.
CTRL+HOME, CTRL+END Move the highlight to the first/last row.
PAGEUP, PAGEDOWN Move the highlight up/down one screen.
CTRL+PAGEUP,CTRL+PAGEDOWN Pand the browser up/down one screen.
Editing Records:
In the browser, double-click (or hit SPACE or ENTER) the desired field to edit. Hit TAB,
ENTER or an arrow key to save changes, ESC to cancel. ENTER or TAB will keep you in a
continuous editing state (see Options | Settings to modify this behavior).

Resizing a Column:
Move the pointer between the gray column headings until it becomes a , then drag the
column's edge to the desired size.

Moving a Column:
Click the column's gray heading area, then drag it to a new location.

Changing what a Column Displays:

You can change what data is displayed in a column by double-clicking its gray heading area.
You will then have the option to display a specific field or the result of a dBASE expression
(see Column | Change Column). Note that the same field can appear as many times as you
like!

Adding Records:
Select File Control | New Record or hit CTRL+N. You don't have to be at the bottom of the
table to add records (but the record will always be added to the end). This option is not
available for all database types.

Deleting Records:
To mark (or unmark) a dBASE record for deletion, either select File Control | Toggle
Delete or hit CTRL+T. Records marked for deletion can be permanently removed with File
Control | Pack Table.

Tools 239
File Control:
New Record Add a new empty record to the end of the table (dBASE files only).
Toggle Delete Mark (or unmark) the current record for deletion (dBASE files only).
Goto Record Jump to a user specified record number.
Find Record Locate a record fitting a specified criteria.
Find Again Locate the next record fitting the previously specified search criteria.
Set Filter View only records that meet a specified criteria.
Set Index View records in a specified sequence.
Append Records Add records from one table into the current table.
Concatenate Fields Join up to eight fields into a single field.
Copy Records Copy records fitting a specified criteria to a new file.
Count Records Count the number of records that meet a specified criteria.
Delete Records Mark for deletion records that meet a specified criteria.
Print Records Print records fitting a specified criteria.
Recall Records Unmark for deletion records that meet a specified criteria.
Replace/With Replace the contents of a field with the result of a specified expression.
Search & Replace Search for text and replace it with something else.
Sort Records Sort the table by a specified criteria (dBASE files only).
Pack Table Physically and permanently remove records that have been marked for
deletion (dBASE files only).
Split Table Split the table in a variety of ways.
Vertical Display Display the columns of the currently highlighted record in a vertical
window.

Column:
Add Column Display a new column in the browser.
Remove Column Remove an existing column from the browser.
Change Column Change what data is displayed in a column.
Auto-Size Column Change the width of a column to fit the longest string.
Auto-Size All Columns Change the width of each column to fit its longest string.
All Columns This Size Change the width of each column to the size of a column.
Reset Columns Reset all column widths to their original sizes. Filter and Index settings
are also reset.

Options:
Settings Modify the browser's display settings.
No Toolbar Hide the browser's toolbar.
Small Toolbar Display the small toolbar.
Medium Toolbar Display the medium toolbar.
Large Toolbar Display the large toolbar.

240 MatchUp
7.1.1 Find Record

Search for Text The word or phrase that you want to locate.

Ignore Case Check if you want to locate the text regardless of the case.

Search Select how you would like to search:

· Any Field Every field will be searched for the specified text.
· Selected Field(s) The checked fields will be searched for the specified text.
· By Expression Search until a record matches the specified criteria. The for
expression must be entered in dBASE syntax. You can use the Expression Builder if
you need help.

Search from Start Start searching from the first record.

Search from Here Start searching from the current record.

Note:
Only rows that are currently being displayed are searched. If you have used Set Filter to hide
some of the rows, you may not find what you're looking for.

Tools 241
7.1.2 Set Filter

Allows you to view records that meet certain criteria. A filter works just like a "for" condition
except that it remains active until you either change or remove the filter. If you Copy, Sort,
etc. once a filter is been set, the program will perform those actions only on the records that
meet your filter criteria. You'll notice that subsequent File Control options will have the
current filter condition listed in the "for" condition box. When a filter is set, the Set Filter tool
will have a check next to its icon on the toolbar:

Show All Records (no filter) Removes the current filter.

Filter by Expression Sets the filter to the expression specified to the right. The filter
expression must be entered in dBASE syntax. You can use the Expression Builder if you
need help.

Use currently activated index If an index has been set, you can check this box to keep that
index in effect. If not checked, the index will be reset to the table's natural order.

Display Filtered Records In what manner filtered records should be shown:

· Don't Display Filtered Records When browsing, only records meeting the filter
expression will be displayed.
· Show Filtered Records in an Alternate Color When browsing, records not meeting
the filter expression will be displayed in the selected color.

242 MatchUp
7.1.3 Set Index

Allows you to view records in a certain sequence. This only effects the order that records are
displayed and not the physical record order. If you want to change the physical order of a
table, use File Control | Sort Records. When an index is set, the Set Index tool will have a
check next to its icon on the toolbar:

Index Records How you would like to order the records:

· 'Natural' Order (no index) Removes the active index.
· Index by Field(s) Indexes by the field(s) you select.
· Index by Expression Indexes the file according to the expression specified to the
right. The index expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Order Sequencing order:

· Ascending "A" before "B", "1" before "2", "03/10/68" before "04/10/68".
· Descending "C" before "B", "3" before "2", "05/10/68" before "04/10/68".

Use currently activated index If a filter has been set, you can check this box to keep that
filter in effect. If not checked, the filter will be reset to show all records.

Tools 243
7.1.4 Append Records

Append records from another table onto the end of the current table. This option is not
available for all table types.

When you first select this option, you are prompted for an input table. Once you have
selected an input table, you will see the Append Records dialog:

Note: Unlike any other File Control tool, the Record Range settings apply to the input table,
not the current table.

Note Field names and lengths should match between the two tables. Data in fields from the
input table that don't have a matching field (ie, a field with the same name) in the current
table will not be added.

All Records in the Table All records in the input table will be appended.

244 MatchUp
Records meeting these conditions Only append records meeting the conditions:
· Record Range Append records between the From and To range. To process from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict appending to only records that meet the specified criteria.
The for expression must be entered in dBASE syntax. You can use the Expression
Builder if you need help.
· While (Expression) Stop appending when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 245
7.1.5 Concatenate Fields

'Concatenate' is just a fancy word for 'join'. It makes us feel superior when we use it.

Input Fields Select the fields that you want to join together, in the order you want them
joined.

Output Fields Select the field that will receive the joined result.

246 MatchUp
If data exists in the output field Select what should happen if there's already something in
the Output Field:
· Append new data to the existing data The joined result will be tacked onto the end
of the existing data.
· Overwrite the existing data The joined result will replace any existing data.

Separate fields with Select what should be inserted between the joined data:
· Single space "John" and "Smith" will become "John Smith".
· No separator "John" and "Smith" will become "JohnSmith".
· This separator "John" and "Smith" will become "John<your connector>Smith".

Remove leading/trailing spaces Select whether spaces should be trimmed:

· Trim leading spaces Remove spaces from the beginning of the field.
· Trim trailing spaces Remove spaces from the end of the field.

Records You can process all or only part of the table:

· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To process
from a specific record all the way to the end of the file, specify "0" as the To
record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 247
7.1.6 Copy Records

Copy records fitting a specified criteria to a new table.

Copy Records You can copy all or only part of the table:
· All Records in the Table To copy all records.
· Records meeting these conditions To copy only certain records:
· Record Range Copy records between the From and To range. To copy from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict copying to only records that meet the specified criteria.
The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop copying when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

When you have completed this dialog and hit OK, you are prompted for the name of an
output table.

7.1.7 Count Records

Count the number of records meeting a specified criteria.

248 MatchUp
Count Records You can count all or only some of the records:
· All Records in the Table To count all records. Not particularly useful when counting.
· Records meeting these conditions To count only certain records:
· Record Range Count records between the From and To range. To count from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict counting to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop counting when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 249
7.1.8 Delete Records

Mark for deletion records meeting a specified criteria. This option is only available for dBASE
files.

Count Records You can delete all or only some of the records:
· All Records in the Table To delete all records.
· Records meeting these conditions To delete only certain records:
· Record Range Delete records between the From and To range. To count from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict deleting to only records that meet the specified criteria.
The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop deleting when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

250 MatchUp
7.1.9 Print Records

Print records meeting a specified criteria. The printed output will match the current browser's
appearance: column contents, positions, sizes, filtering and ordering.

Printer Select the printer to output to.

Font/Size The font to be used for printing.

Properties Displays the printer's Printer Setup dialog. You can usually change settings
such as Paper Orientation, Paper Size and Print Quality. This dialog is dependant on your
printer driver, so your mileage may vary.

Page Setup Displays the WIndows Page Setup dialog. You can change the Paper Size,
Orientation and Margins.

Field Structure Check this box to print the table's structure.

Records Check to print the table's contents.

Tools 251
Page headings and page numbers Check to print each page with a title indicating the
name of the table, as well as a page number.

Field headings Check this box to print field names at the top of each page.

Records You can print all or only some of the records:

· All Records in the Table To print all records.
· Records meeting these conditions To print only certain records:
· Record Range Print records between the From and To range. To count from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict printing to only records that meet the specified criteria.
The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop printing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

252 MatchUp
7.1.10 Recall Records

Remove deletion marks from records meeting a specified criteria. This option is only
available for dBASE files.

Un-Mark Records for Deletion You can unmark all or only some of the records:
· All Records in the Table To unmark all records.
· Records meeting these conditions To unmark only certain records:
· Record Range Unmark records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict unmarking to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop unmarking when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 253
7.1.11 Replace/With

Replace the contents of a field with the result of a specified expression.

Replace Field Select the field that will receive the result of the With Expression.

With Expression Specify the dBASE expression that will determine the new contents of the
Replace Field. The contents must be of the same type of data as the Replace Field. For
example, if the Replace Field is a Numeric field, then this expression must result in a
number. You can use the Expression Builder if you need help.

Replace Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

254 MatchUp
When you have completed this dialog and hit OK, you are prompted for a Processing Mode.

Tools 255
7.1.12 Search & Replace

Search for specific text and replace it with another.

Source How the Search & Replace should be directed:

· Screen The search and replace text will be entered in the Search for and Replace
with prompts.
· File The search and replace text will be retrieved from the specified file. This file must
be a dBASE database with two character fields. The first field is used as the search
text, the second as the replace text.

Search & Replace all fields Every field in the table will be searched.

256 MatchUp
Search Field The field to search.

Replace Field The field to replace into. The default is to use the Search Field, which is
usually what you want.

Delete 'Search for' string from Search Field If you chose a Replace Field that is different
than the Search Field, you can remove the Search for text from the original by checking this
box.

Search for The text to search for. If you don't elect to ignore case, then be sure to enter the
data exactly as desired. Checking Blank Field searches for a blank Search Field.

Replace with The text to replace the Search for text with. It will be used exactly as you
enter it (i.e., it is case dependant). Checking Blank Field will clear the Replace Field when
the Search for text is found.

Ignore Case Check if you want to search regardless of the case.

Whole word/Any occurrence If Whole word is selected, "IT" will be found in "SMITH",
"SMIT" or "ITH". In order to be considered a whole word, the search text must occur as a
word or a series of words. If Any occurrence is selected, the text may appear anywhere in a
field.

Multiple occurrences/First occurrence If Multiple occurrences is selected, each and

every occurrence of the search text in a field is replaced. If First occurrence is select, only
the first one is replaced.

Search & Replace Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

When you have completed this dialog and hit OK, you are prompted for a Processing Mode.

Melissa Data has gone through several different iterations of methodology to determine
leading and trailing spaces (DOS users will remember the degree symbol and "plaiditer"
systems). With our Windows programs, you just need to type in the phrase just as you would
like it to be searched or replaced. For example, if you want to search for "and", with trailing
and leading spaces, enter it as " and " (note the leading and trailing spaces; don't type the

Tools 257
double quotes). The trailing space(s) won't appear unless you highlight the contents of the
entry (in the field, hit CTRL+A).

If you are using a Search and Replace file, you can use the original method to denote trailing
spaces, the degree symbol (ASCII character 248).

258 MatchUp
7.1.13 Sort Records

Physically sorts a table by a specified criteria. This option is only available for dBASE files.

Sort Records How you would like to order the records:

· Sort by Field(s) Indexes by the field(s) you select.
· Sort by Expression Sorts the file according to the expression specified to the right.
The for expression must be entered in dBASE syntax. You can use the Expression
Builder if you need help.

Tools 259
Order Sequencing order:
· Ascending "A" before "B", "1" before "2", "03/10/68" before "04/10/68".
· Descending "C" before "B", "3" before "2", "05/10/68" before "04/10/68".

Records You can sort all or only some of the records:

· All Records in the Table To sort all records.
· Records meeting these conditions To sort only certain records:
· Record Range Sort records between the From and To range. To count from a
specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict sorting to only records that meet the specified criteria.
The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop sorting when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Records that don't meet the above conditions If Record Range, For or While conditions
are specified, you need to indicate what to do with records that don't satisfy the conditions:
· Should precede sorted records Records not matching the criteria will be placed (in
'natural' order) before the sorted records.
· Should follow sorted records Records not matching the criteria will be placed after
the sorted records.
· Should be removed from the table Records not matching the criteria will be
permanently removed.

260 MatchUp
7.1.14 Split Table

Splits the table into one or more sub-tables using the method you select (indicated on the
tabs at the top of the dialog):
· <n> Records/Bytes/Files Split the table so that the resultant tables are no larger than
the specified size.
· <n> th Select Split the table so that every nth record is taken.
· Random Select Randomly select a specified number of records from the table.
· Contents of Field Use a source field to determine which resultant table a record
should be copied. Commonly used for undoing a Merge/Purge or dividing a large table
up by Zip Code or SCF.

Tools 261
Split Table: <n> Records/Bytes/Files

Table will be split into <n> Records, Bytes, or Files Select the splitting method:
· Make files having no more than...records each Each output table will have the
specified number of records (the last table may have fewer).
· Make files no larger than...bytes each Each output table will be the same size (the
last table may be smaller).
· Make exactly...files of (nearly) equal size Records will be evenly divided into a
specified number of tables.

262 MatchUp
Output Files Select how output tables will be named:
· Ask me for a file name for each file Each time a file is created, you will be prompted
for a name.
· Assign files sequential numbers; put them in this folder Files will be named
0000001.dbf, 00000002.dbf, etc, and stored in the specified folder.

Split Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 263
Split Table: <n>th Select

Take every...th record from the table Specify the gap between each record selected. For
example, "4" will select every fourth record or 25% of the table.

Also mark selected records for deletion Selected records are marked for deletion. This
option is provided so that you can provide incremental selections. For example, if you later
wanted to process only the reocords that were previously selected, you could specify a For
Expression of deleted(). This option is only available for dBASE files.

264 MatchUp
Split Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 265
Split Table: Random Select

Randomly select...records from the table Specify how many records should be selected.

Also mark selected records for deletion Selected records are marked for deletion. This
option is provided so that you can provide incremental selections. For example, if you later
wanted to process only the reocords that were previously selected, you could specify a For
Expression of deleted(). This option is only available for dBASE files.

266 MatchUp
Split Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 267
Split Table: Contents of Field

Base split on contents of field Select the field whose contents will determine to which table
a record will be copied.

Start at character position Specify the point in the field at which data should be extracted
for the split.

Number of characters to use Specify how many characters should be used to determine
the split (counting starts at the Start Position, not the start of the field).

268 MatchUp
Output Files Select how output tables will be named:
· Ask me for a file name for each file Each time a file is created, you will be prompted
for a name.
· Use the contents of the field as the file name; put them in this folder The contents
of the field will be used as the file name (illegal characters are stripped) and stored in
the specified folder.
· Assign files sequential numbers; put them in this folder Files will be named
0000001.dbf, 00000002.dbf, etc, and stored in the specified folder.

Split Records You can process all or only some of the records:
· All Records in the Table To process all records.
· Records meeting these conditions To process only certain records:
· Record Range Process records between the From and To range. To count from
a specific record all the way to the end of the file, specify "0" as the To record.
· For (Expression) Restrict processing to only records that meet the specified
criteria. The for expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.
· While (Expression) Stop processing when records no longer meet the specified
criteria. The while expression must be entered in dBASE syntax. You can use the
Expression Builder if you need help.

Tools 269
7.1.15 Add Column

A column will be added to the browser at the highlighted column position. Its contents will be
what you specify below:

Heading/Size The name and size of the column you want to create.

Column Contents What should be displayed in the column:

· Expression Display the result of the specified expression as the column contents. The
expression must be entered in dBASE syntax. You can use the Expression Builder if
you need help.
· Field Displays the contents of an existing field.

270 MatchUp
7.1.16 Change Column

A column will be added to the browser at the highlighted column position. Its contents will be
what you specify below:

Heading/Size The name and size of the column.

Column Contents What should be displayed in the column:

· Expression Display the result of the specified expression as the column contents. The
expression must be entered in dBASE syntax. You can use the Expression Builder if
you need help.
· Field Displays the contents of an existing field.

Tools 271
7.1.17 Settings

Display Font/Size The browser's font.

Printer Font/Size The font used when printing.

Filtered Record Color The default color used for records that are being filtered.

When editing... Control the behavior of the ENTER key:

· ENTER moves to the next field to the right When editing a field, the ENTER key will
move the cursor to the next field to the right.
· ENTER moves to the next row down When editing a field, the ENTER key will move
the cursor to the same field, next row down.

When not editing, at the rightmost column... Control the behavior of the TAB key:
· TAB moves to the leftmost column, one row down When in the rightmost column,
the TAB key will move the cursor one row down and circle to the first column.
· TAB moves to the leftmost column, same row When in the rightmost column, the
cursor will circle back to the first column of the same row.

272 MatchUp
7.1.18 Processing Mode

The Processing Mode dialog appears on the Search & Replace, Replace/With and
Concatentate Fields commands:

Processing Mode Specify what records should be brought to your attention for review:
· Verify each replacement The screen will display the current contents of the field and
what it will look like after replacement. The processor will ask if you want to process
the record every time.
· Verify only when replacement that would result in a truncation The field is
displayed for review only when processing will result in a truncation. If the replacement
string is longer than the search string (so a truncation would result), the processor will
pause to verify whether or not to proceed with the replacement.
· Skip replacements that would result in a truncation In all cases the search string is
automatically replaced with the replacement string unless the replacement will cause a
truncation. No replacements are displayed to the screen.
· Replace all records, regardless of truncation In all cases, the field is automatically
replaced with the replacement string or as much of the replacement string as will fit -
no exceptions. These replacements are not displayed to the screen.

Tools 273
274 MatchUp
7.2 Modify Structure
Modify Structure allows you to change the structure of a dBASE table: to add new fields,
delete fields, move fields, resize fields, and rename fields.

With the exception of renaming fields, all structure changes require enough disk space to
make a copy of the current database.

Add Add a new field at the highlight's position.

Remove Remove the currently highlighted field.

Change Modify the currently highlighted field. You can also modify fields by double-clicking
the Field Name, Type, Size or Decimals entry that you want to change.

Tools 275
Adding Fields:

Field Name Field names must be between 1 and 10 characters. The first character must be
alphabetic, but the others may be letters, numbers, or the underline (_). Embedded spaces
are not allowed.

Field Type There are four options: Character, Date, Logical, and Numeric.

Field Size Date fields are automatically sized to 8 and Logical fields to 1. Character field
sizes must be between 1 and 254. Numeric field sizes must be between 1 and 20.
Remember that if you shorten a field, you will lose any data that exceeds the new field width.

Decimals (Numeric fields only) Decimal places must be between 0 and the field's size minus
2 (i.e., a numeric field of size 10 can have between 0 and 8 decimal places).

276 MatchUp
Tools 277
7.3 ASCII Conversion
7.3.1 Import from ASCII

Most database applications can convert to and from ASCII, since this is the most universal
data format available. MatchUp can process files in ASCII delimited, fixed field and flat files if
you specify a structure. If you select an ASCII file when you are in a setup, you will specify
the structure, but you won't actually import the file.

Upon selecting this option, you are first prompted for the name and location of the ASCII file.
Then, the following screen is presented:

Use the Interactive Importer Use the Interactive Importer to determine the ASCII file's
structure.

Manually Enter Manually type in the ASCII file's structure.

Inherit currently selected table's structure The structure of the currently open table will be
used as a starting point for the ASCII file's structure. The Interactive Importer is then
displayed for you to fine-tune the results.

Append to the currently selected table The ASCII file is appended directly to the end of
the current table. The ASCII file's structure must match the current or the ASCII file's records
will not be imported correctly.

Copy another table's structure Use another table (that you will select) as a starting point
for the ASCII file's structure. The Interactive Importer is then displayed for you to fine-tune
the results.

Once the ASCII file's structure has been specified and/or reviewed (using one of the above
methods), you are prompted for an output table name. The file is then imported.

278 MatchUp
Import from ASCII: Interactive Import

Split Field (SDF and Flat File formats only) You can split the highlighted field into two
smaller fields by clicking this button. The field will be split in half and then you can adjust the
fields' sizes by dragging the dividing line to the left or right. Note that this is a positional split,
not an intelligent split. You cannot use Split Field to break up a name field ("Mr. John Smith")
into its parts. For this type of work, you should use the Name Splitter in Personator or
MatchUp.

Join Field (SDF and Flat File formats only) You can join two adjacent fields into a single
larger field by clicking this button. The highlighted field and its neighbor to the left will be
joined.

Exclude Field You can intentionally exclude a field during importing by clicking this button.
Of course, this field won't be removed from the original ASCII file, just the imported
database. Excluded fields are shown in gray.

Clear Fields You can wipe out all of the fields (so you can start all over) by clicking this

Tools 279
button.

Get field names from first record If the ASCII file's first record contains field names (as
opposed to data), then check this box to use that data as field names (you can still change
them to whatever you like). During the import, the first record will not be imported as data.

File Info Click this button if you need to change the ASCII file's type. Occasionally, the
program will incorrectly guess the file's type and/or record size. Changing an ASCII file's type
will cause the importer to reassign fields, so any changes made to field names, positions,
etc., will be lost.

Field Info Click this button or double-click on the field's heading to change the information
about the highlighted field. You can change the field's name, type, size and decimal places.

280 MatchUp
Interactive Import: File Information

ASCII File Type The type of ASCII file being imported:

· SDF Often called "fixed field" or "standard data format". Fields are of fixed length
(padded with spaces) and each record is delimited with a carriage return and a line
feed.
· Ragged SDF A spin-off of the SDF format, usually made by custom or in-house
programs. The difference is that all trailing blanks at the end of each record are
trimmed off.
· Flat File A common mainframe format. Fields are of fixed length, just like SDF, but
there are no record delimiters of any kind.
· Delimited A common ASCII format. Fields are of varied length and delimited with a
single character, usually a comma or a tab. Sometimes (particularly with character
fields), fields are surrounded with double quotes. Each record ends with a carriage
return and a line feed.

Delimiter (Delimited format only) The character that separates fields. Everybody's favorites,
<Comma> and <Tab>, are listed first.

Record Size (SDF and Flat File formats only) The number of bytes used to store each
record. Remember to count the carriage return and/or line feed in this number.

Tools 281
Interactive Import: Field Information

Field Name The field name must be between 1 and 10 characters. The first character must
be alphabetic, but the others may be letters, numbers, or the underscore (_). Embedded
spaces are not allowed.

Field Type There are seven options: Fixed Character, Variable Character, Integer, Float,
Decimal, Logical, and DateTime. The program will automatically change the size of some of
these types.

Field Size Some fields have fixed sizes, so you cannot edit these sizes. Character field sizes
must be between 1 and 254.

Decimals (Numeric fields only) Decimal places must be between 0 and the field's size minus
2 (i.e., a numeric field of size 10 can have between 0 and 8 decimal places).

282 MatchUp
Import from ASCII: Manually Enter

Add (or CTRL+INSERT) Add a new field to the bottom of the list.

Remove (or CTRL+DELETE) Delete the currently highlighted field.

Change Modify the currently highlighted field. You can also modify fields by double-clicking
the Field Name, Type, Size or Decimals entry that you want to change.

Moving Fields:
To move a field, just click and drag it to the desired location. Alternately, you can press
CTRL+UP or CTRL+DOWN to accomplish this.

Tools 283
Adding Fields:

Field Name Field names must be between 1 and 10 characters. The first character must be
alphabetic, but the others may be letters, numbers, or the underline (_). Embedded spaces
are not allowed.

Field Type There are seven options: Fixed Character, Variable Character, Integer, Float,
Decimal, Logical, and DateTime. The program will automatically change the size of some of
these types.

Field Size Date fields are automatically sized to 8 and Logical fields to 1. Character field
sizes must be between 1 and 254. Numeric field sizes must be between 1 and 20.
Remember that if you shorten a field, you will lose any data that exceeds the new field width.

Decimals (Numeric fields only) Decimal places must be between 0 and the field's size minus
2 (i.e., a numeric field of size 10 can have between 0 and 8 decimal places).

284 MatchUp
7.3.2 Export to ASCII

Source Table This lists each field in the table. By default, all fields will be exported. If you
wish to export just a few fields, first click None, then click the field(s) to export and click Add.

Output File This lists each field to be exported. To remove field(s) from the list, click the
field(s) to remove and click Remove. Note that the output field order of the Output File will
always follow the same order as the Source Table. The only way to change the order is to
physically modify the table's structure before exporting.

File Type The type of ASCII file to be created:

· SDF Often called "fixed field" or "standard data format". Fields are of fixed length
(padded with spaces) and each record is delimited with a carriage return and a line
feed.
· Flat File A common mainframe format. Fields are of fixed length, just like SDF, but
there are no record delimiters of any kind.
· Delimited A common ASCII format. Fields are of varied length and delimited with a
single character, usually a comma or a tab.

Delimiter (Delimited format only) The character that separates fields. Everybody's favorites,
<Comma> and <Tab>, are listed first.

Add first record with field names Check this box to create an initial record that contains
the field names (your ASCII file will be one record longer).

Tools 285
Macintosh format (CRs, no LFs) (SDF and Delimited formats only) Check this box to
terminate each record with only a carriage return. Some Apple applications require this
format.

Export records marked for deletion (dBASE tables only) Check this box to export records,
regardless of whether or not the record's been marked for deletion. Deletion marks don't
exist in ASCII files, so all records will appear as 'normal' in the ASCII file.

Strip embedded CRs and LFs Sometimes you will come across tables where field(s) have
carriage returns (CRs) and/or line feeds (LFs) embedded (i.e., the characters aren't being
used to delimit records - Excel seems to be the most common source of this problem.).
These embedded CR/LFs can cause big problems with some database applications, so we
offer the option of stripping them here.

Delimit data with double quotes There are several different standards as to how fields
should be delimited:
· Always Every field gets delimited with double quotes.
· Only Character Fields Only character fields get double quotes - this is the most
popular standard.
· Only when there is a comma The field is delimited with double quotes if a comma is
embedded in the field.
· Never No double quote delimiters ever.

286 MatchUp
Tools 287
7.4 Matchcode Editor

Very Important Note:

A common source of confusion is the relationship between matchcode components and field
names. Note that nowhere in the Matchcode Editor do you specify an actual table or field
name. You only supply component field types. So how does the program know to use a field
called "NAME" as the source for first name? You specify it on the Matchcode Mapping setup
tab. This may seem like an extra step, but it gives you great flexibility. First, you create
matchcodes that work for all of your databases, regardless of their structure. Then you "map"
the fields in the Matchcode Mapping tab of your Merge/Purge or Purge setup.

The Matchcode Editor lets you maintain your personal matchcode database. Once you've
used MatchUp for a while, you will probably have developed custom matchcodes that work
well with your data. When you've developed the "Killer Matchcode", let us know. We're
always looking for hot new prospects.

288 MatchUp
7.4.1 Matchcode List

This is a list of the matchcodes in your matchcode database. The middle part of the screen
will list each component in the highlighted matchcode.

Our "canned" matchcodes are shown in red and are marked as "Read-only". This was done
so that our technical support staff can rely on a consistent benchmark when troubleshooting
users' problems. You can make your own matchcodes read-only by highlighting the
matchcode and pressing CTRL+R to toggle the read-only status. This can come in handy in
maintaining control over operating procedures.

The order of the matchcodes can be changed by dragging and dropping matchcodes. Note
that the order of this table has no impact on any processing, it's just for your benefit.

Add Creates a new empty matchcode. If there is a matchcode that is close to what you
need, it is usually easier to highlight that matchcode and click Copy.

Remove Delete the highlighted matchcode.

Copy Clone the highlighted matched with a new name that your specify. Now you can edit
this new matchcode to your heart's delight.

Rename Change the name of the highlighted matchcode. You can also rename a
matchcode by double-clicking it.

Tools 289
7.4.2 Matchcode Components

This table lists the components in the currently selected matchcode. As you move from
component to component, the properties in the lower left corner of your screen, will change
to reflect the currently selected component.

Add Create a new component to the list. A dialog box pops up prompting you for this new
component's properties. New components are always added to the bottom of the list; you
can change a component's position by dragging and dropping it in the list.

Remove Delete the highlighted component from the matchcode.

I don't understand these X's Well, you've come to the right place! The Matchcode
Combinations topic is probably a good place to start.

Component Order:
The order of the components can be changed by dragging and dropping them or CTRL+UP
or CTRL+DOWN. The order of this table can have some impact on processing in two ways:
· The first component has special restrictions that the others don't (see below).
· The order of components can effect processing speed. See Optimizing Matchcodes for
more information.

290 MatchUp
7.4.3 Component

You can change properties by double-clicking or right-clicking a property or by highlighting

the component and changing the properties below.

Data Type The type of data represented by this component.

Size The number of characters that will be used from this component. See Component
Properties for more information.

Label A label that can be attached to this component. MatchUp does not itself use this label,
but it can be helpful in remembering for what a particular General component was intended.
If a label is present, it will be displayed on the Matchcode Mapping screen.

Maximum number of words The maximum number of words that will be used. See
Component Properties for more information.

Start Where character extraction should begin:

· Left From the left end of the data (most common choice).
· Right From the right end of the data.
· Position From the specified character position in the data.
· Word From the specified word in the data.

Trim Whether leading and/or trailing white space should be stripped from the data. See
Component Properties for more information.

Custom When processing custom data types, a custom lookup table is specified here. See
Matchcode Components for more information.

Tools 291
First Component Requirements:
For accurate clustering, MatchUp places some special restrictions on the first component in
a given matchcode (clustering is discussed in Concepts).

The following must be true for a matchcode's first component:

· It must be used in every condition.
· It cannot use certain Matching Strategies: Containment, Frequency, Fast Near,
Frequency Near or Accurate Near (other types are okay, though).
· It cannot use "Match if one field is blank".
· It cannot use "Match initial to full field".
· It cannot use Swap Matching.

MatchUp will display all components that meet these conditions with a pale green
background - until a compent not meeting these conditions is encountered. Having additional
components that meet these criteria will usually make the matchcode process faster (see
Optimizing Matchcodes for more information).

292 MatchUp
7.4.4 Matching Strategy

Matching Strategies allow for matching of non-exact components. These options are
mutually exclusive (i.e., you can only select one at a time). The following settings are
available:
· Exact Match Data must be identical to be seen as a match.
· Soundex An auditory matching algorithm originally developed by the INS and later
adopted by the USPS.
· Phonetex An auditory matching algorithm.
· Containment Match when one record's component is contained in another record's.
· Frequency Match the characters in one record's component to the characters in
another, without any regard to the sequence.
· Fast Near A typographical matching algorithm. Faster than Accurate Near.
· Accurate Near A typographical matching algorithm. More accurate than Fast Near.
· Frequency Near A typographical matching algorithm. A combination of Frequency
and Accurate Near.
· Vowels Only Only vowels will be compared. Consonants will be ignored.
· Consonants Only Only consonants will be compared. Vowels will be ignored.
· Alphas Only Only alphabetic characters will be compared.
· Numerics Only Only numeric characters will be compared. Decimals and signs are
considered numeric.

Tools 293
Matching Strategy: Soundex

Soundex is an algorithm developed in 1917 by the Immigration Service and later adopted by
the USPS. It is identical to the Soundex() function found in any of the dBASE languages. Its
basic operations are to (1) keep the first letter of the string, (2) remove any vowels, (3) ignore
all H's, W's and Y's, (4) change all double letters to singles, and (5) substitute numbers for
the next 3 consonants using the following table (if there are no more consonants, 0's are
used):

Character(s) Code
FPV 1
CSKQXZ 2
DT 3
L 4
MN 5
R 6
B 7
GJ 8

For example, "Peoplesmith" would be Soundexed to "P142".

Although Soundex has been in use over 80 years and has a wide following, we cannot
understand how "S", "K", and "Q" can be grouped together as the same sound.
Consequently, we greatly prefer Phonetex.

Matching Strategy: Phonetex

Phonetex (pronounced "Fo-net-icks") is similar to Soundex, but there are some key
differences in its operation: (1) scan the string for certain letter combinations, and substitute
alternate representations. For example, hard C's are converted to K's and soft C's are
converted to S's. Leading GN's, PN's, and MN's are converted to N's, and so on. (2) certain
phonemes like "PH" and "IGHT" are recognized. (3) Vowels are then removed from the
string, and then numbers are substituted for the remaining consonants using the following
table:

294 MatchUp
Character(s) Code
X 1
KQ 2
L 3
R 4
MN 5
CSZ 6
FVW 7
BDPT 8
GJ 9

For example, "Peoplesmith" would be Phonetexed to "P83658".

Note that Phonetex doesn't have the 4 character restriction that Soundex has. Additionally,
consonant combinations are specially treated, and letter groupings are more consistent with
English speech patterns.

Matching Strategy: Containment

A match is found when one record's component is contained in another record. For example,
"no" is contained in "innovation".

Matching Strategy: Frequency

The characters in one record's component are matched against the characters in another
without any regard to the sequence. For example "abcdef" would match "badcfe".

Matching Strategy: Fast Near

A typographical matching algorithm. It works best in matching words that don't match
because of a few typographical errors. Exactly how many errors is specified on a scale from
1 to 4 (1 being the tightest). The Fast Near algorithm is a speedy approximation of the
Accurate Near algorithm. The tradeoff for speed is accuracy; sometimes Fast Near will find
false matches or miss true matches.

Matching Strategy: Accurate Near

A typographical matching algorithm. The Accurate Near algorithm produces more accurate
results than Fast Near, but at the cost of speed.

Tools 295
Matching Strategy: Frequency Near

The characters in one record's component are matched against the characters in another
without any regard to the sequence. For example "abcdeg" would match "badcfe". You can
specify the number of allowed errors with the slider.

296 MatchUp
7.4.5 Short/Empty Settings

Match if both fields are blank If two records have the same empty component, that
component will be counted as matching. So, two records with the first name missing will
match. See Component Properties for more information.

Match if one field is blank Allows MatchUp to match missing data with the full data. For
example, "Smith" matches "John Smith". However, two records with the same component
missing will not match. See Component Properties for more information.

Match initial to full field Allows MatchUp to match abbreviated data with the full data. For
example, "J Smith" matches "John Smith". See Component Properties for more information.

Tools 297
7.4.6 Combinations

The 1 through 16 check boxes define component combinations that should be considered a
match. It is easier to visualize the effects of these boxes if you look at the Component Table
part of the screen:

It is important to note that each vertical column of checkmarks designates one acceptable
matchcode. For example, the illustration above shows a combination that is made up of 4
matchcodes:
1. Zip5, Last Name, Street #, and Street Name.
2. Zip5, Last Name, and PO Box
3. Zip5, Company, Street#, and Street Name.
4. Zip5, Company, and PO Box

Since Street Number is highlighted, the Combinations box displays check marks for the
component combination Street Number appears in: 1 and 3.

One final comment: due to a requirement of Clustering, at least the first component must be
used in every combination (which is why we provide the check marks for the first row.

298 MatchUp
Combinations: Advanced Combination Settings

Warning: Advanced Topic!

The Advanced Combination Settings allow you to specify whether or not a particular
combination should be evaluated in every situation. In most cases, you will want to evaluate
all combinations in all situations (the default).

This powerful option can be used in situations where an awkward workaround would
otherwise be needed.

Tools 299
Used Advanced Combination Settings Check if you need to set advanced combination
settings for any combination in your matchcode.

Combination Select the combination that you would like to configure.

Reset this combination Reset the highlighted combination to its default (non-advanced)
configuration.

Reset all combinations Reset all combinations to the default (non-advanced) configuration.

Use in comparisons with records from Regular Files This combination should be
evaluated when dealing with regular records.

Use in comparisons with records from Suppression Files This combination should be
evaluated when dealing with suppression records.

Use in comparisons with records from Intersection Files This combination should be
evaluated when dealing with intersection records.

AND this is true If the above selections are satisfied and the specified expression is true.
The expression is a dBASE expression, but instead of field names, component names
should be used when needed (see below).

OR this is true If the above selections are satisified or the specified expression is true. The
expression is a dBASE expression, but instead of field names, component names should be
used when needed (see below).

(no additional conditions) Only the above selection(s) needs to be satisified.

When MatchUp evaluates the above AND or OR conditions, it performs this check on both
records being compared. You must tell MatchUp if both records need to satisfy the condition
or if only one needs to:
· Use this combination if either record satisfies the condition
· Use this combination only if both records satisfy the condition

dBASE Expressions:
The dBASE expressions used for the AND and OR conditions are nearly the same as the
type that you would use for an Input Filter condition. The one difference is that these
expressions can't use field names (because we don't know what tables you are planning on
using with this matchcode). Instead, you can use the matchcode's component names. For
component names with embedded spaces, the spaces should be removed (ie, the First
Name component should be specified FirstName). If you have several components that are
the same type, or you would like more specific names, assign the component(s) a Label in
the Matchcode Editor, and refer to that label in the expression.

In some instances, you will want to refer to a field in your table that isn't in the matchcode.

300 MatchUp
You can accomplish this by adding a component to the matchcode (usually a General will
do), but not assigning it any X's (so it will not be directly required by any combination). You
can then refer to this field in your expression.

When you use a component in an expression, the following applies:

· It will always be a character data type.
· It will always be of whatever size you specified for the component.
· It will be padded with spaces so that it is the specified size (ie, you may need to
RTrim() when doing string comparisions.

It will receive any early matching treatments which are described in Optimizing Matchcodes.

Tools 301
Combinations: Advanced Combination Example

Advanced Combinations may seem like such an abstract concept that you couldn't imagine
why someone would ever use them.

Say you would like to do a mailing to potential MatchUp customers. The target would consist
of Data Processing workers in large and small companies. Now, the logic may be this:
· For small companies (less than 20 employees), a loose matchcode (more matches)
could be used, because chances are that if the mail piece landed in the hands of the
inappropriate individual, it would be passed on
· For large companies (over 20 employees), a tighter matchcode (less matches) could
be used, because information doesn't exchange hands so readily in Cubical City.

A matchcode like the following would be created:

Component Label Size Fuzzy Short/Empty 1 2 3 4

Zip5 5 Exact None X X X X
Last Name 5 Exact Both Fields X X
Company 10 Exact Both Fields X X
Street Number 5 Exact Both Fields X X
Street Name 10 Exact None X X
PO Box 10 Exact None X X
General Employees 5 Exact None
This sets up 4 matching conditions:

Large Companies:
1. Zip5 + Last Name + Street Number + Street Name
2. Zip5 + Last Name + PO Box

Small Companies:
3. Zip5 + Company + Street Number + Street Name
4. Zip5 + Company + PO Box

However, as it stands right now, all conditions will be evaluated in all situations, large or
small company. That's where the Advanced settings come in…

302 MatchUp
Tools 303
Here, we've set a rule for condition one that states that when Employees is greater than 20,
Condition 1 should be used. Condition 2 gets the same rule. Conditions 3 and 4 get a
different rule:

Which states the exact opposite of the rule for Conditions 1 and 2.

By now, you're probably wondering about this General field with the "Employees" label.
Unlike most dBASE expressions, the Matchcode Editor cannot use field names when
evaluating an expression. The reason is simple: there are no tables! But you can use the
contents of components. So we throw a General component in to use for our evaluation. We
give it a label ("Employees") so that we know what we're talking about in our expressions
and also later when we're on the Matchcode Mapping tab. So when we use this matchcode
in a setup, we will need to map an Employee Count field into the "Employee" component.

304 MatchUp
When you use a component in an expression, the following applies:
· It will always be a character data type.
· It will always be of whatever size you specified for the component.
· It will be padded with spaces so that it is always the specified size (ie, you will want to
use RTrim() when doing string comparisons).

One other thing to note is that the General field isn't included in any combinations. The
Matchcode Editor is okay with this and will not remove this field during any optimizations.

Tools 305
7.4.7 Swap Match Pairs

Swap Matching allows you to match "John Smith" with "Smith John". Essentially, you indicate
in the Matchcode Editor two components whose contents can be swapped (in this case, First
Name and Last Name). The components must be of the same size and should have the
same set of matching options (i.e. don't Phonetex one and SoundEx the other). You can
define up to eight pairs.

Both components must match When a pair of components is flipped, both components
306 MatchUp
must match in the reversed configuration. For example, "John Smith" would match "Smith
John", but not "Smith <No FN>" nor "Smith Bob". We call this option a "full swap".

Either component can match When a pair of components is flipped, either component can
match in the reversed configuration. For example, "John Smith" would match "Smith John",
"Smith <No FN>", and "Smith Bob". This option is sometimes referred to as a "half swap".

Full Swap or Half Swap?

It depends on the situation. If you're matching two components that have the potential for
being reversed (First and Last Name, for example), a full swap is probably what you want.

However, if you're matching several items appearing on a list, a half swap is the most likely
option. For example, if you have a list that contains several household members (ie, FIRST1,
FIRST2, FIRST3), you would like to catch matches where FIRST1 matches FIRST1, FIRST2
or FIRST3. In this situation, a Full Swap may not always work, as the ordering of names is
inconsistent. Half Swap is the way to go.

7.4.8 Print Matchcode

You can print any or all of your matchcodes using this dialog box.

Tools 307
7.4.9 Optimize

The Optimize option tries to re-arrange your Matchcode's components in the most efficient
order. It will also remove redundant combinations and arrange them in an optimal order.

In some (rare) cases, the existing component order may be critical to your process (for
example, if you're using the Output Matchcode in another program). Sometimes the existing
combination numbering may be critical (for example, if you're using the Output Matchcode in
another program). In both of these cases, it's best to not use the Optimize option, or at least
double-check the changes that MatchUp has made to your matchcode.

See Optimizing Matchcodes for more information on some of the more advanced
optimizations that this option does and does not perform.

308 MatchUp
Tools 309
7.5 User Settings
If you invest three minutes now to set up MatchUp with your personal taste and style, it will
pay off handsomely in the hours you will save in the future!

7.5.1 General
General: Interface

310 MatchUp
ENTER key Control the behavior of the ENTER key in dialog boxes:
· ENTER moves to the next dialog prompt, CTRL+ENTER hits OK When navigating
a dialog box, hitting ENTER will proceed to the next prompt. Technically, ENTER is
supposed to hit the OK button, but many people find this an unnatural behavior, so we
give you a more conventional interpretation.
· TAB moves to the next dialog prompt, ENTER hits OK When navigating a dialog
box, hitting ENTER will immediately hit the OK button. This is standard Windows
behavior.

Screen Saver Control the screen saver's behavior:

· Don't suppress screen saver The Windows screen saver's behavior will not be
changed.
· Suppress screen saver when processing When processing, MatchUp will
temporarily disable the screen saver. This ensures that all processing power is given
to the program rather than rendering flying toasters.

Tip of the Day Control how often you want to see the Tip of the Day dialog:
· Always show tips at startup You'll get them every time.
· Show tip once per day Only the first time, so pay attention.
· Sometimes show tip Only when you least expect it.
· Never show tip Don't like them?

Advanced: dBASE Emulation:

SET EXACT OFF This advanced option allows you to control how the = operator behaves in
dBASE expressions. If this is checked, "Shortcutted" string comparisons are performed. For
example, "BLACKWELL"="BLACK" will return true (but "Black" will not equal "Blackwell".).
This behavior is the opposite of SET EXACT ON. See = Operator, and == Operator in the
help file for more information.

Advanced: ADO/DAO Usage

This advanced option allows you to select which Microsoft technology to use when accessing
Access and Excel files. You may find that one technology performs better than the other in
certain situations. Also, see the Frequently Asked Questions if you are having difficulties with
Access or Excel files.

Tools 311
General: Main Window

ToolBar Change the toolbar's appearance:

· No Toolbar Hide the browser's toolbar.
· Small Toolbar Display the small toolbar.
· Medium Toolbar Display the medium toolbar.
· Large Toolbar Display the large toolbar.

Background Change the main window's background:

· No Background (use standard window colors) If your computer was made during
the cold war, this is for you.
· Single Color Nice and plain, nothing to see here.
· Gradient Good Lord, the colors!

312 MatchUp
General: User File Locations

Default location of setup (.dt) files: Lets MatchUp know where to start looking for your .dt
setup files.

Default location of data files: Lets MatchUp know where to look for your data files.

The above two options can make work a bit easier for you. If you always store your setup
files and/or data files in the same location, it is worth your while to specify the folder(s)
above. If specified, Add Table and File | Open Setup will always start in the specified
location. If no folder is specified, these commands will start in the last file location (which is
very useful for some people).

Tools 313
General: Auto-Update

Auto-Update Select how often MatchUp should check for updates:

· Don't ever check for updates You can still download updates from our web site and
install them manually.
· Periodically check for updates, but ask first Sometimes you really need to get the
job done without waiting for an update.
· Periodically check for updates without asking Updates are pretty fast over
broadband connections and don't take more than a minute.

314 MatchUp
7.5.2 Setup
Setup: Field Naming

This screen lets you customize how MatchUp determines what kind of data is contained in a
field. By entering your field naming habits, you can set up a new job significantly faster.
These descriptions are used in Merge, Purge, Merge/Purge, and Update setups. Similar
descriptions are used for CASS setups.

Edit Descriptions for Select the type of field you want to add naming description(s) to. For
example, if you always name Prefix fields "MR_MS", you'll want to select "Prefix" in this drop
down box.

Add Description Enter a field name for this type of field. Wildcards are allowed. For
example, if you always name Prefix fields "MR_MS", you'll want to put "MR*" or "MR_MS" in
this box. When you have typed the information, click Add to put it on the Existing
Descriptions list. If your database structures are fairly consistent, it is usually better to not
use wildcards and specify the entire field name. For example, if you specified "CO*" to catch
a field you usually call "COMPANY", MatchUp can mistake "CONTACT" as a Company field,
as it fits the wildcard description.

Existing Descriptions Lists field names for the currently selected type of field.

Remove Removes the highlighted description from the list.

Move Up/Move Down The position of a field in the description list is important, as MatchUp
evaluates the Description List from the top down. Therefore, you can improve how well
MatchUp determines field names by putting the most common field descriptions at the top of
the list.

Tools 315
Setup: Field Mapping

Automatically map fields on "Matchcode Mapping" tab Check this box if you want
MatchUp to automatically determine field names on the Matchcode Mapping setup tab.

Automatically map fields on "Output Field Mapping" tab Check this box if you want
MatchUp to automatically determine field names on the Output Field Mapping setup tab.

Automatically map fields on "Update" tab Check this box if you want MatchUp to
automatically determine field names on the Update setup tab.

316 MatchUp
Setup: Default Output Table

Default Table Type When you create a new Merge/Purge or Merge setup, certain
restrictions are made on your Output Fields (naming, sizes, etc.). These restrictions are
based on what database type you choose for your output tables. The catch is that MatchUp
usually has no idea what database type you have in mind (as the Output Tables often have
yet to be specified). In cases like these MatchUp will usually make a guess based on the
database type(s) of your input tables. However, if you've specified a Default Table Type
here, it will be used instead.

When creating a Merge or Merge/Purge setup, the default output structure:

· is empty No output structure will be entered. You will have to enter a structure for the
output table.
· is inherited from the first added input table Inherit the structure from the first source
file given. When you do a Merge or a Merge/Purge, the default output field structure
will be the structure of the first input file you select.
· is the following The default output structure is specified in the table below. Use the
Add, Remove and Change buttons to add, remove and change fields. The Inherit
from Table button allows you to aquire a structure from an existing database that you
specify (you can modify the structure afterwards if you need to).

Tools 317
Remember that no matter what output structure you decide to go with, you will have the
option to change the output table structure while you are creating the setup.

318 MatchUp
Setup: CASS
CASS: Field Naming

This screen lets you customize how MatchUp determines what kind of data is contained in a
field. By entering your field naming habits, you can set up a new job significantly faster.
These descriptions are used in CASS setups. Similar descriptions are used for Merge,
Purge, Merge/Purge and Update Setups.

Edit Descriptions for Select the type of field you want to add naming description(s) to. For
example, if you always name Prefix fields "MR_MS", you'll want to select "Prefix" in this drop
down box.

Add Description Enter a field name for this type of field. Wildcards are allowed. For
example, if you always name Prefix fields "MR_MS", you'll want to put "MR*" or "MR_MS" in
this box. When you have typed the information, click Add to put it on the Existing
Descriptions list. If your database structures are fairly consistent, it is usually better to not
use wildcards and specify the entire field name. For example, if you specified "CO*" to catch
a field you usually call "COMPANY", MatchUp can mistake "CONTACT" as a Company field,
as it fits the wildcard description.

Existing Descriptions Lists field names for the currently selected type of field.

Remove Removes the highlighted description from the list.

Move Up/Move Down The position of a field in the description list is important, as MatchUp
evaluates the Description List from the top down. Therefore, you can improve how well
MatchUp determines field names by putting the most common field descriptions at the top of
the list.

Tools 319
CASS: Output fields

When a CASS setup is created, these field names (and sizes if the fields don't exist in the
table) will be automatically specified. If you have developed a field naming convention,
entering it here will save you from ever entering it again in CASS setups.

320 MatchUp
CASS: Options

The following settings will automatically be selected whenever you start a new CASS setup.

If there's a coding error What should MatchUp do if it comes across a coding error (you
can choose one or both):
· Clear the Input Plus 4 field
· Clear the Input Carrier Route field

City/State Delimiter If you elect to put city, state, and zip data into a single field, how would
you like them delimited?
· Delimit City & State with a space
· Delimit City & State with a comma

Tools 321
Zip/Plus 4 Delimiter Same deal, but seperating the Zip and Plus 4:
· Delimit Zip & Plus 4 with a dash
· Delimit Zip & Plus 4 with a space
· Delimit Zip & Plus 4 with no delimiters

Perform Delivery Point Validation (DPV) Check if you want to verify each address's
delivery point during CASS verification.

Use CASSmate enhanced processing Whether or not to use the enhanced power of
CASSmate to get more CASS matches. Enhanced CASSmate processing is only used when
normal CASSing attempts have yielded no results.

Process via Zip Index CASSmate generally processes faster in Zip order. But, often your
table is not in Zip order (and you don't want it to be). With this option checked, MatchUp will
create a temporary index that will cause your table to be CASSed in Zip order (but it's
physical order will remain untouched). This option is disabled for all table types exact dBASE
(the speed gain is countered by random access slowdowns).

Form 3553 information This is the general information necessary to fill out the USPS Form
3553 for the CASS certification. Be sure to fill in this information or your Form 3553 will not
print out complete.

322 MatchUp
Setup: Warnings

Before processing, MatchUp checks for possible error and warning situations. You can
suppress any or all of the warnings by unchecking it here.

Note: Only warnings can be shut off. Errors (as opposed to warnings) are not shown here
because you can not turn them off.

See also Setup Inspector.

Tools 323
7.5.3 Processing
Processing: General

Ratio: Screen Updates vs. Processing Speed:

Adjust the slider to your preference. When you set for more frequent screen updates, it
appears that records are being processed faster, but it's an optical illusion. In reality, fewer
records are being processed (because more time is being wasted updating the screen). Test
it yourself!

Because Windows is a multitasking environment, it is quite possible that you may be running
another Windows program while processing (which is completely legal). You will notice that
faster slider settings will slow down other windows programs running, while slower slider
settings will speed them up. You should choose the slider setting that works best for what's
running - some programs demand more processor time than others.

The processing screen has a pause button, so if you need to have your computer's
"undivided attention", you can pause MatchUp to do your things and then resume
processing.

Work Folder: The work folder stores the Purge and Merge/Purge Key and Index files. If you
are often getting "Out of disk space" warnings, you should confirm that the work folder is a
Drive/folder with lots of disk space. Also, in network situations, it is desirable to use a fast
local drive instead of a network drive as speed is greatly degraded by network traffic.

324 MatchUp
Processing: Translation Table

Translation Table These tables are used to exchange foreign characters for an English
equivalent when building matchcodes:
· CP 1252 Windows Default
· CP 437 English Old DOS days
· CP 850 Multilingual Slightly less old DOS days
· CP 852 Slavic
· CP 860 Portuguese
· CP 10000 Macintosh They just have to be different, don't they?
· CP 863 Canadian
· CP 865 Nordic

For data that's been keyed in North America in the past 5 years or so, CP 1252 Windows is
usually the Code Page that was used. If it was keyed overseas or a while ago, anything
goes.

Import a Translation Table from Dirty Harry Allows you to import tables from our Dirty
Harry's Character Assassin program. This also means that you can use Dirty Harry to build
custom translation tables. Note that this doesn't change your data in any way, just how
MatchUp compares data.

If you have created your own translation tables (the code pages we didn't cover), please
share these with us so we can supply them to all our users.

Tools 325
Processing: Counting Method

This determines how your Count and Multi-Buyer Count fields are populated. The best way
to understand your options is to carefully examine the examples on the screen.

326 MatchUp
Processing: Gathering

Gather information to both the Output and Duplicate tables MatchUp will gather
information from source tables and copy it to the output tables and any dupe tables.

Gather information to only the Output table MatchUp will copy the information only to the
output tables.

Tools 327
Processing: Stacking

This determines how your Stacking data is populated. The best way to understand your
options is to carefully examine the examples.

328 MatchUp
Processing: Status Codes

In Merge/Purging and Deduplicating, you are given the opportunity to specify an output
status field. This field may be formatted as a Binary or as ASCII characters.

Binary saves space, but it is difficult to read. Basically, each status code (a number) is stored
in a four byte binary representation. For example, the number "12345" is stored as "90".

Alternately, ASCII Characters can used.

Purge & Merge/Purge Status Codes:

ASCII Hexadecimal Decimal Description
A 0X20 32 Record is a Dupe.
B 0X10 16 Output record has a Dupe.
C 0X00 0 Output record has no Dupes (it is unique).
D 0X02 2 Record was suppressed.
E 0X08 8 Record was not intersected.
F 0X04 4 Intersection record was used to intersect.
G 0X01 1 Suppression record caused a suppression.

In addition to this code, MatchUp can tell you which matchcode was used in a match.
Immediately following the status code will be the number(s) of the matchcode combination(s)
that caused the hit. If the binary format is used, this will be a 16-bit mask of the OR'd values.
If you don't know what a "16-bit mask" is, stick with ASCII.

Tools 329
Processing: Multi-Threading

If you have purchased the multi-processor add on, this option will be enabled. This gives you
control over how many simultaneous threads can be used during the two dedupe passes.
This option requires a bit of trial and error as the best settings may not necessarily be the
number of processors. Other factors include disk access speed, available memory, and other
running process.
Multi-threading works in this way:

Pass 1 - Building Keys:

A thread is launched for the first source table. If the maximum thread count is more than one,
a second thread is launched for the second source table. This continues until the maximum
thread count is reached. Once this happens, new threads (for additional source files) are
only launched when one becomes available (ie a table's keys have all been built).

Obviously, for a single input table, multi-threading this pass has no advantage.

Threads are assigned to files in the order they are listed in the Input Tables tab. It is usually
best to put your larger tables at the top of the list. For example, say you have 3 input tables,
2 medium-sized, and 1 large, and are processing with 2 threads:

If they are ordered 1, 2, 3 in the setup, thread allocation would look like this:

330 MatchUp
If they are ordered 3, 1, 2 in the setup, thread allocation is more like this:

The second scenario is quite a bit faster. MatchUp is not able to figure out which table is
going to take the longest (it is not necessarily the table with the most rows), so you need to
determine this yourself and perform the appropriate action.

In most cases, the optimal maximum thread setting is equal to the number of processors in
your machine.

Pass 2 - Deduping:
A thread is launched for each cluster. MatchUp typically processes several thousand clusters
in a single run, so one would think multi-threading would really speed this process up.
Unfortunately, this isn't usually the case, as each one of these threads are vying for the same
resources, primarily the key file, source tables, and output table(s). When one thread is
accessing any a resource, any other threads must wait until the first thread is done with the
resource before they can access it (known as 'resource locking'). Although you may find
otherwise (experimentation with your data is the key), usually a setting of 0 or 1 work best
here.

0 Threads versus 1 Thread:

MatchUp without the multi-threading option uses a setting of 0 threads. What this means is
that it never launches threads while processing. For example, during pass 2 (0 threads), the
calling program (aka, the 'thread master') actually pauses while a cluster is processed and
waits until it is complete before figuring out the makeup of the next cluster. If the maximum
thread setting is 1 or more, the thread master keeps itself busy by looking for the next cluster
(ie, it builds up a 'thread queue'). The thread master can keep itself pretty busy with this
work, to the point that it steals processing cycles from the launched threads. For this reason,
usually a pass 2 setting of 1 is better than 2 on a 2 processor machine.

Tools 331
7.5.4 Analyzer

These settings are used as a default when you analyze a setup for the first time as well as
when you reset the Analyzer display.

Display Font/Size The font used for the analyzer display.

Printer Font/Size The defaul font used when printing.

Records from Regular, Self-Purge and No-Purge List(s):

· Unique Records Show records with no matches.
· Records having Duplicates Show all records that have matched other records.
· Duplicate Records Show duplicate records (this doesn't include suppressed or non-
intersected records).
· Suppressed Records Show records that were matched against a suppression list(s).
· Non-Intersected Records Show records that didn't match against any intersection
list(s).

Records from Suppression List(s):

· None Don't show any records from the suppression list(s).
· Only Hits Only show records that were actually used to suppress other records.
· All Records Show all records from the suppresion list(s).

332 MatchUp
Records from Intersection List(s):
· None Don't show any records from the intersection list(s).
· Only Hits Only show records that were actually used to intersect other records.
· All Records Show all records from the intersection list(s).

Tools 333
7.5.5 Reporting

Allows you to change the default display and printer fonts for Reporting.

Display Font/Size The font used for the analyzer display.

Printer Font/Size The defaul font used when printing.

Logo Bitmap Template Reports usually print a bitmap at the top of the first page. By default,
this is the Melissa Data Logo. You probably have a better one in mind. To use it, create a
Windows bitmap, sized approximately 200 x 200 pixels and specify it here.

Multi-Buyer Statistic The Multi-Buyer count that appears on the File Summary and Source
Code File Summary can be reported in a variety of ways, as shown in the examples on the
screen.

334 MatchUp
Tools 335
8 Additional Tools
8.1 View File
Allows you to view the an open table.

Right-clicking gives the options:

· Auto-Size Column Change the width of a column to fit the longest string.
· Auto-Size All Columns Change the width of each column to fit its longest string.
· All Columns This Size Change the width of each column to the size of a column.
· Reset Columns Reset all column widths to their original sizes.

336 MatchUp
Additional Tools 337
8.2 Field List
Lists the fields in each source table.

Fields shown here with a cyan background have already been used in the setup, ones with a
white background have not.

When you place the mouse over a field, general information about that field is displayed at
the bottom of the window:

338 MatchUp
The Field List isn't just for show, however. You can click and drag fields from the field list into
input and output fields in your setup:

Additional Tools 339

8.3 Expression Builder

The Expression Builder is a tool to aid you in creating dBASE expressions. Basically, you
build your expression step by step. See dBASE syntax for more information on what makes
up a dBASE expression. There is some overlap between some of the different data types as
some functions are useful with different types.

String Available string operators and functions. Double-click a selection to insert it into
Expression.

Numeric Available numeric operators and functions..

Logical Available logical operators and functions.

Date/Time Available date operators and functions.

340 MatchUp
Misc/Table Functions that are specific to table operations and/or dont really fit into any other
category.

Fields This list allows you to select the fields in the current database.

Expression This area shows the expression as you build it.

Test Click this button to ensure that the dBASE expression shown is valid in syntax. This
won't verify whether the expression will do what you want - only you can tell.

Additional Tools 341

342 MatchUp
8.4 Check Mapping

When merging files, it is important that all input fields line up and that no data is lost.
MatchUp makes this chore a lot easier by telling you which fields have not been used. You
can use Check Mapping on each file used in the merge or have been used more than once.

Mapped Fields These fields are currently being used in the merge.

Doubly Mapped These fields have been mapped more than once, possibly by mistake.

Unmapped Fields These fields are not being used in the merge, but there is an output field
that has the same name. When auto mapping is on, you won't see fields in this box unless
you have unmapped the field yourself. With the auto-mapping off, you could end up with
several fields in this box. If you would like to map the field to the output file's structure, check
the box next to that field.

Missing Fields These fields are not being used in the merge and there isn't an output field
having the same name. If you would like to add a field to the output file's structure, check the
box next to that field.

Additional Tools 343

8.5 Check Truncations

Note:
There is absolutely no easy way that MatchUp can find when truncations have actually
occurred. This option tells you when the potential for truncations exists, not if they do exist.
All output fields are listed, but the ones that don't have a potential for truncation are greyed
and can not be selected.

Say you have a City/State/Zip field of 30 characters, and you elect to split the field into a City
field of 20 characters. MatchUp has no way of knowing if you have any cities that are longer
than 20 characters. To be on the safe side, MatchUp warns you about the possibility of
truncation here.

344 MatchUp
Additional Tools 345
8.6 Global Scatter/Gather
When you need to specify gathering and/or scattering for several fields or files, it can get
quite tedious to setup each file. To make this easier, use the Global Gather/Scatter dialog:
· Output Gathering
· Input Gathering
· Input Scattering

8.6.1 Output Gathering

Once you select the field(s) you want to gather to, check Gather and then select the Method
you want to use. To execute your selections, click Change!

346 MatchUp
8.6.2 Input Gathering

Select the source fields you want to use while Gathering. Left-click the fields that you want to
change. Fields shown in green have their gather/scatter setting on, fields in white have their
setting off. Left-clicking a row or column heading will toggle the entire row or column.

Note:
You are able to select any field here, even if Gathering has not been enabled for its Output
Field. The Gather status on the main screen will be grayed until Gathering has be enabled
for its output field.

Additional Tools 347

8.6.3 Input Scattering

Select the source fields you want to use while Scattering. Left-click the fields that you want to
change. Fields shown in green have their gather/scatter setting on, fields in white have their
setting off. Left-clicking a row or column heading will toggle the entire row or column.

Unlike Gathering, the Output Field does not need to have Gathering turned on.

348 MatchUp
Additional Tools 349
9 Reference
9.1 Command Line
Several of MatchUp's options are available only from the command line. Most of these
options are used in "hands off" processing (what we call batch processing), while others can
save you the time of selecting the same files every time you run MatchUp. The syntax for
MatchUp's command line is:

dt3 [setupfile] [options]

[setupfile] The name of the setup file to use in processing. If the setup file is not in the
working directory, you must specify the file's path.

[options] Processing options:

· /T Used in Batch Processing to test a batch run.
· /R Used in Batch Processing to execute a batch run.

Note that with no options, specifying [setupfile] will only open this file, but will not begin
processing until you choose to.

The switches are not case-sensitive and can occur in any order. "-" can be used instead of
"/". Each switch must be preceded by a space, however.

You can specify command line options in one of two ways:

· Creating a new shortcut: Right click on the Desktop, select New | Shortcut. For the
command line, enter the desired command line. It is usually a good idea to specify
the full path of the MatchUp executable, as well as the database and setup paths. For
example,

c:\dt\dt3.exe c:\job285\fallmail.dt

would make MatchUp automatically open the fallmail.dt setup. Since the setup file
contains the specifications for both source and output tables, it is unnecessary to
specify them.
· Running MatchUp from a DOS session. From the Start Menu, select Command
Prompt. At the DOS prompt, change to the desired working directory and simply type
the desired command line. MatchUp will be opened in a new window.

350 MatchUp
Long file names with spaces?
For example, your setup file is called Merge Purge.Dt. Just surround it with double quotes.
When using long file names having spaces, applications can get confused as to where one
parameter stops and the next one begins. Surrounding a parameter with quotes solves this
problem. This works on any 32-bit Windows program. Note that you do this for each
parameter, not the whole command line (a common mistake). For example:

c:\dt\dt3.exe "c:\Job Folder\Merge Purge.dt" /r

Or another, more complicated example (notice the careful use of double quotes here):

"c:\Program Files\Dt3\Dt3.exe" "c:\Job Folder\Merge Purge.dt" /r

Reference 351
352 MatchUp
9.2 Batch Processing
You can run MatchUp completely uninterrupted from a shortcut or DOS command line. To let
MatchUp know that you want this special capability, add one of these switches onto the end
of the command line:
· /T To test that the setup file will run when the time comes.
· /R To run the setup file.

As outlined in the command line section, there are one or two ways to use these switches.
You can exploit the DOS session method even further by using a batch file (just like in
regular old DOS!). The only caveat to this method is that you should check the Always
Suspend option in the shortcut's property dialog box. Why? Take the following sample
batch:

c:\dt\dt3 c:\work\file.dt /R
c:\sl\sl32 c:\work\file.dbf c:\work\file.sl /R

With Always Suspend checked (nowadays more often than not), the second line (the
StyleList line) will not be executed until the first line is complete. Without "Always Suspend"
checked, the second line executes moments after the first line is started. Not a desirable
situation when the input of the second line depends entirely on the successful completion of
the first line!

On some systems, the "Always Suspend" setting is not enough. There are many script
processing utilities that have addressed this problem and provide a solution. Also, the DOS
START command will often do the trick if you use the /WAIT switch. For example:

start /wait c:\dt\dt3 c:\work\file.dt /R

c:\sl\sl32 c:\work\file.dbf c:\work\file.sl /R

For programmers, there's an alternate solution: When MatchUp is done processing in batch
mode, it creates a file called dt.flg. You can use the presence of this file as an indication that
it is okay to continue. For example, in FoxPro:

erase d:\dt\dt.flg && erase the flag file

run d:\dt\dt3.exe "d:\dt\merge purge.dt" /R && launch MatchUp

&& Loop until the flag file appears:

do while !file("d:\dt\dt.flg")
=inkey(1000) && sleep for a second
enddo

Note the use of double quotes around the setup file. When using long file names having
spaces, applications can get confused as to where one parameter stops and the next one
begins. Surrounding a parameter with quotes solves this problem.

Reference 353
354 MatchUp
9.3 Errors

General File Errors:

General file errors are situations that may occur while attempting to access a database,
setup, or ASCII file:
· Open Database fail The specified database couldn't be opened, or perhaps the
specified folder doesn't exist.
· Open Table fail The specified table couldn't be opened. It may not exist or you may
not have the necessary permission to open it.
· Create Table fail The specified table couldn't be created. You may not have the
necessary permission needed to create it.

Setup File Errors:

Setup file errors can occur while attempting to open a setup file:
· File doesn't appear to be a MatchUp Setup file You've specified a file that is not a
MatchUp Setup File or the Setup File has become corrupt.

DBMS File Errors:

DBMS file errors can occur while processing, but shouldn't:
· Read error An unexpected error occurred while reading the specified file. Try the
operation again. If the situation still persists, call Melissa Data Technical Support.
· Write error An unexpected error occurred while writing to the specified file. Try the
operation again. If the situation still persists, call Melissa Data Technical Support.

dBASE Expression Errors:

dBASE errors occur when you are entering a dBASE expression as a While, For, Filter or
Sort expression.
· Syntax error The expression is syntactically incorrect. Check for a misspelled function
or field name.
· Mismatched Parenthesis The expression is missing a right or left parenthesis.
Remember that you should have a matching number of left and right parenthesis.
· Mismatched Delimiters A literal character string was specified with non-matching (or
missing) delimiters.
· Expression does not evaluate to data type The expression does not result in the
specified data type. For example, "LTrim(First)" (a character expression) is not an
appropriate expression for a Filter condition, which requires a logical expression.
· Field does not exist Specified field does not exist. Check for a misspelling.
· Function does not exist Function is not valid. Check for a misspelling.
· Empty Expression No expression was specified, although one is required.
· Invalid number of arguments Too many or too few parameters were given for the
specified function. This is usually the result of a misplaced or missing parenthesis,
comma, or misspelling.

Reference 355
 · Type error One of the parameters given for the specified function is of the wrong type.
For example, "SubStr(RecNo(),12,3)" will result in this error because SubStr requires a
Character expression and RecNo() is Numeric.

Other Errors:
These errors may occur anytime while running MatchUp and, although you may find this
hard to believe, may indicate a bug in the program:
· Unknown error
· Exception Error
· GPF

356 MatchUp
Reference 357
9.4 dBASE
9.4.1 dBASE Syntax

While it is beyond the scope of this documentation to explain the full use of dBASE syntax
(you should see a book or manual on dBASE, Clipper, or FoxPro for that), some rudiments
are explained below.

dBASE syntax involves using quasi-English expressions to tell the program what you want to
do. These expressions use Operators, Functions and Literals in conjunction with the Field
Names in your database.

Definitions:
· FOR Condition A True/False expression determining which records will be affected
during a database operation. Each record in the table is evaluated "for" this condition.
· WHILE Condition An operation that will continue only "while" the expression is True,
then it will stop.
· SORT Expression An expression that will be evaluated for each record to be sorted. It
can result in a Character, Number, Date, or logical value.
· FILTER Condition Similar to a "for" condition, except in its context. "For" and "while"
expressions are always performed in conjunction with another operation, whereas
filter conditions occur all by themselves and affect any subsequent operations like a
"for" would.

Types of Expressions:
There are four types of expressions: Character, Date, Numeric, and Logical. In this
document, they look like this:
· <Cexp> Character expression.
· <Dexp> Date expression.
· <Nexp> Numeric expression.
· <Lexp> Logical expression.
· <Exp> Any type of expression

Expressions can consist of…

a literal "Melissa Data"

an expression SubStr("Melissa Data", 5, 3)
a field Name

358 MatchUp
Specifying Literals:
Literals must be specified in a special way, so the program can distinguish between
character, date, numeric, or logical literals.
· Character Literals Delimit with double quotes (") or single quotes ('). For example,
"Melissa Data" and 'Melissa Data' specify the same character literal. You
cannot embed the delimiters. For example 'MacDonald's' is not valid (so use one of
the other delimiters, like "MacDonald's"). Note that the old square bracket delimiters
are no longer allowed. They are now used to delimit field names.
· Date Literals Delimit with curly braces ({ and }), and use the format {mm/dd/yyyy}. For
example, {3/10/1968} represents March 10, 1968. In the old days, there was no
such thing as a date literal, and people usually specified them by converting a
character literal (for example CToD("3/10/1968")). If you omit the century (i.e.
{03/10/68}) the year will be assumed to be between 1950 and 2050.
· Numeric Literals There are no delimiters, just specify the number. For example, 3.14
is a numeric literal.
· Logical Literals There's only two, so they're special: .T. for true, and .F. for false.
Yes, those periods are required.

Specifying Functions:
Functions are fairly simple. Type the function's name, a left parenthesis, any parameters
(separated by commas), and a right parenthesis. If you nest functions, be sure that you
consistently match parenthesis. Following the dBASE tradition, casing is not significant and
only the first four letters of a function's name are needed, so…

ALLT(SUBS(NAME,5,4))

is the same as

AllTrim(SubStr(Name,5,4))

Specifying Fields:
In most cases, simply specifying the field name will work. However, if the field contains
spaces, you must delimit the field with square brackets, as in [First Name]. dBASE tables
don't allow embedded spaces, but many other DBMS's do.

See dBASE Functions (Alphabetic listing), dBASE Functions (Categorical listing) in the Help
Files for more information.

Reference 359
360 MatchUp
9.5 About Melissa Data Corporation
Thank you for using at our problem-solving software for mailing lists. Our programs are
designed to help people to communicate more personally (and more effectively) in this
increasingly computerized (and impersonal) age:

We offer these solutions to today's mailing list problems:

· MatchUp™ combines many powerful and convenient features like letting you process
files in different formats (Xbase/ASCII) and different field structures directly, saving
you the time to make them uniform. Matchcodes can use any parts of any fields, and
you can process up to 16 matchcodes simultaneously. Does suppression and
intersection runs, parses street addresses, and offers complete, editable reports.
· StyleList™ makes your labels and envelopes more attractive, so they stand out and
get opened more often. It converts All-upper to Upper/Lower case (or vice versa),
expands user-definable abbreviations (or creates them), corrects punctuation, and
controls field lengths.
· Personator™ increases response rates and profit from direct mail by (1) splitting Full
or Inverse name fields into Prefix, FN, MN, LN, and Suffix/Title, (2) adding correct
prefix (genderizing), and (3) making perfect salutations. It simplifies
merge/sort/deduplication. Pro Series splits addresses.
· GenderBase 100™ is an optional Personator add-on with over 100,000 single word,
un-hyphenated first names (the genderizing table supplied with Personator has the
8,500 most common). When used with Personator pro series it can correct common
abbreviations and typos.
· Right Fielder™ "re-fields" badly fielded lists: distinguishes names from companies,
primary from secondary address, domestic from foreign, etc. Imports "print image" and
"label format" files into true databases
· Dirty Harry's Character Assassin™ is a power user tool that strips or replaces any
ASCII character with any other at a blazing speed.

For programmers, we offer several products as programming APIs:

· Personator API™ Name splitting, genderizing, salutation building, and street splitting.
· Right Fielder API™ "re-fielding" data.
· StyleList API™ Casing, abbreviating, abbreviation expansion, e-mail, phone, and
zip/postal code reformatting.
· MatchUp API™ Matches database records.

Reference 361
APIs are provided as 32-bit Windows DLLs. These are callable from C/C++, Visual Basic,
Foxpro, and other languages that support DLL function calls.

Typical uses are:

· Real-time data entry validation.
· Enhancement of databases with custom file structures (i.e., files that cannot be
exported into a format that our Windows products can readily use).
· Providing functionality for a custom written application seamlessly (i.e., without shelling
to another application and with the desired look and feel).

362 MatchUp
Reference 363
9.6 Copyright & License Agreement
Copyright 1995-2007 Melissa Data, Inc.

This Melissa Data Corporation program ("the Software") is protected by both United States
Copyright Law and International Treaty provisions.

By installing this software you accept all of the terms contained in this license
agreement, so read this license agreement carefully before installing this Software.

You may either (a) make one copy of the Software for backup or archival purposes only, or
(b) transfer the Software to a single hard disk provided you keep the original disks for backup
or archival purposes only.

You may use the Software (a) on any single-user computer provided you have physical
possession of the CD at all times during use of the Software, or (b) on a network provided
each person using the Software has purchased a separate copy of the Software and each
such person has physical possession of the CD provided with it. You may not temporarily
transfer possession of the right to use the Software to another individual.

To permanently transfer the Software to another party, you must at the same time either
transfer all copies of the Software (regardless of form) to the same party or destroy any
copies not transferred. You may terminate this license at any time by destroying all copies of
the Software (in any form). Your license terminates automatically if you fail to comply with
any terms or conditions of the Melissa Data Corporation License.

You may not copy, reproduce, or distribute the Software nor alter, modify, reverse engineer,
decompile, disassemble, or otherwise attempt to render source code from the Software. You
may not copy or otherwise reproduce the Manual or Help Files, in part or in whole, without
the prior written consent of Melissa Data, Inc.

MatchUp, MatchUp API, Street Smart, CASSmate, StyleList, StyleList API, Personator,
Personator API, GenderBase 100, Right Fielder, Right Fielder API, Dirty Harry's Character
Assassin, and Melissa Data are registered trademarks of Melissa Data, Inc.

Borland C++ is a registered trademark of Borland International, Incorporated.

Microsoft Visual Basic, Microsoft Visual C++, Microsoft FoxPro, Microsoft Visual FoxPro,
Microsoft Access, Microsoft Excel, Microsoft Word, SQL Server 7, Microsoft SQL Server
2000, Microsoft Windows 95/98/ME, and Microsoft Windows NT/2000/XP are registered
trademarks of Microsoft Corporation.

Oracle is a registered trademark of Oracle Corporation.

DB/2 is a registered trademark of IBM Corporation.

364 MatchUp
AddressObject and Melissa Data are registered trademarks of Melissa Data Corporation.

ZIP Code, ZIP + 4, and CASS are registered trademarks of the United States Postal Service
(USPS).

All other brands and products are trademarks of their respective holder(s).

Reference 365
366 MatchUp
9.7 Limited Warranty
Melissa Data warrants that the Software will perform substantially in accordance with Melissa
Data's current published specifications, documentation, and authorized advertising; that the
Help File(s) and Manual (if provided) contains the necessary information to use the Software;
and that the media (if provided) on which the Software is furnished will be free from defects
in materials and workmanship for a period of ninety (90) days from the date of purchase. The
remedy for breach of this warranty is limited to replacement or refund at your discretion and
shall not encompass any other damages, including but not limited to loss of profit, special,
incidental, consequential, or other similar claims.

EXCEPT FOR THE LIMITED WARRANTY ABOVE, THE SOFTWARE, MANUAL, AND
HELP FILE(S) ARE PROVIDED "AS IS". Melissa Data MAKES NO OTHER WARRANTY,
EXPRESS OR IMPLIED WITH RESPECT TO THE SOFTWARE AND/OR HELP FILE(S)
AND SPECIFICALLY DISCLAIMS THE IMPLIED WARRANTIES OF THE
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. PEOPLESMITH
DOES NOT WARRANT THAT THE SOFTWARE AND/OR HELP FILE(S) WILL MEET
YOUR REQUIREMENTS OR EXPECTATIONS OR THAT THE OPERATION OF THE
SOFTWARE WILL BE UNINTERRUPTED AND/OR ERROR FREE. YOU ARE SOLELY
RESPONSIBLE FOR THE SELECTION OF THE SOFTWARE TO ACHIEVE YOUR
INTENDED RESULTS AND FOR THE RESULTS ACTUALLY OBTAINED. IN NO EVENT
SHALL PEOPLESMITH BE LIABLE FOR ANY LOSS OF PROFIT OR ANY OTHER
COMMERCIAL DAMAGE, INCLUDING BUT NOT LIMITED TO SPECIAL, INCIDENTAL,
CONSEQUENTIAL OR OTHER DAMAGES.

This Agreement shall be construed, interpreted, and governed by the laws of the State of
California.

See also: Copyright & License Agreement

Reference 367
 AND Values 54

Index Gathering 54
Append Records 244
ASCII Export 285
ASCII file 13
-<- ASCII Import 278
Manually Enter 283
<n>th Select 264
Audio 324
Split File 264
Auto-Update 314

-A- -B-
Accurate Near 30, 295
Background 312
Add Column 193
Main Window 312
Analyze 193
Batch Processing 353
Browse 270
Blank Field Matching 36
Add New Output Field 74
Both Blank Fields 30, 36
Merge 74
Break Grouping 23
Add Output Field 94
Browse 238
Merge/Purge 94
Add Column 270
Add Values 54
Change Column 271
Gathering 54
Browse Font 272
Additional Output Fields CASS 106
Browser options 272
Address Field Mapping 40
Advanced Combination Settings 299
Advanced File Type 124
Purge 124
-C-
Advanced Rank 88 Canadian Postal Code Component 26
Merge/Purge 88 Canadian Users 59
Purge 125 CASS 319
Advanced Tab 114 Field Naming 319
Merge/Purge 114 Merge/Purge 104
Purge 132 Output Fields 320
Merge/Purge 86 Setup 152
Purge 124 CASS Input Fields 104, 152
Alphas Only 30 CASS Options 154
Analyze 189 CASS Output 110, 159
Print 189 CASS Processing 177
Vertical Display 191 CASS Reports 230
Analyzer fonts 332 CASS Subscription 12
CASS Updates 12
CASSmate on CD 10
Change Column 195 -D-
Analyze 195
Database vs. Table 13
Browse 271
DB2 21
Change Structure 275 File Requirements 21
Changing Mappings 40
dBASE file 13
Check Mapping 343
DBASE Syntax 358
Check Truncations 344
Deduping 23
City Component 26
Default Output Table 317
Clustering 23 User Settings 317
Combinations 25, 33
Delete Records 250
Advanced 299
Department/Title Component 26
Matchcode 298
Double Take Errors 355
Command Line 350
Dupe Group 57
Company Acronym Component 26
Dupe Matrix 206
Company Component 26 Reports 206
Component Mapping 40 Source Code Reports 225
Component Properties 30
Component Sequence 44
Optimization 44 -E-
Components 26
Matchcode Editor 290 Earliest Date 54
Concatenate Fields 246 Gathering 54
Concepts 23 Edit Results 182
Concurrent Matching 25 E-Mail Component 26
Consonants Only 30 Enter Key 310
Containment 30 Enter Key behavior 272
Containment (matching) 295 Errors 355
Contents of Field 268 EULA 364
Split File 268 Exit 67
Copy Records 248 Export ASCII 285
Copyright 364 Expression Builder 340
Count Records 248
Counting Method 326
Country Component 26
-F-
Credit Card Component 26 Fast Near 30, 295
Features 6
new 6
Field Info 282 First Component 33
Interactive Import 282 First Component Requirements 33
Field List 338 First Data 54
Field Mapping 40 Gathering 54
Merge/Purge 90 First Name Component 26
User Settings 316 First/Nickname Component 26
File Update 143 Fonts 332
Purge 128 Analyzer 332
Field Matching 70 Reporting 334
Merge 70 Foreign Characters 30, 325
Field Naming 315 Foreign Charactes 30
User Settings 315 FoxPro 13
Field Naming CASS 319 Frequency 30
Fields 36 Frequency (Matching) 295
Both Blank 36 Full Address Component 26
File Info 281 Fuzzy 30
Interactive Import 281 Optimization 44
File location default 313 Fuzzy Matching 30
File Requirements 13 Accurate Near 295
DB2 21 Containment (matching) 295
ODBC 15 Fast Near 295
Oracle 20 Frequency (Matching) 295
SQL Server 19 Phonetex 294
File Summary 203 Soundex 294
Reports 203
Source Code Reports 222
File Update 143 -G-
Field Mapping for Matchcode 143
Gathering 53
General Tab 137
Gathering Methods 54
Information About 142
Gathering settings 327
Input Field 146
Gender Component 26
Input Sources 141
General Component 26
Input Tables 141
General Tab 137
Matchcode 138
File Update 137
Matchcode Component 145
Merge/Purge 79
Matchcode Mapping 143
Purge 118
Ranking 139
Global Scatter/Gather 346
Setup 136
Merge/Purge 346
File Update Processing 174
Global Update 150
Find Record 241
 Interface 310
User Settings 310
-H- International records 62
Introduction 5
Hide Rows 186
Analyze 186
How to Dedupe 23 -J-
-I- Join 54
Gathering 54
Gathering 54
Import ASCII 278
Index 243
Browse 243
Inferred Matching 25
-L-
Information About 142 Label 30
File Update 142 Last Name Component 26
Merge Setup 69 Latest Date 54
Merge/Purge 84 Gathering 54
Purge 122 License 364
Initial Only 30 Limited Warranty 367
Input Field 146, 148 Locate Again 185
File Update 146, 148 Locate Record 241
Merge 73 Locate Row 185
Merge/Purge 92, 97 Longest 54
Purge 131 Gathering 54
CASS 104, 152
Input Sources 141
File Update 141 -M-
Merge/Purge 83
Main Window 312
Purge 122
User Settings 312
Input Tables 141
Manually Enter 283
File Update 141
ASCII Import 283
Merge Setup 68
Mapping 90, 128
Merge/Purge 83
Merge/Purge 90, 128
Purge 121
Mapping Components 40
Installation 10
Mapping Fields 40
Installing CASSmate 10
Mapping issues 343
Interactive Import 279
Matchcode 138
Field Info 282
File Update 138
File Info 281
Matchcode 138 Merge Processing 165
Print 307 Merge Setup 68
Purge 119 Information about 69
Matchcode Combinations 298 Input Tables 68
Matchcode Component 291 Merge/Purge 83
File Update 145 Input Sources 83
Merge/Purge 91 Input Tables 83
Purge 130 Matchcode Mapping 90, 128
Matchcode Components 26, 30 Output Fields 93
Matchcode Editor 288 Setup 79
Matchcode Mapping 40 Merge/Purge CASS 104
File Update 143 Merge/Purge Processing 168
Merge/Purge 90, 128 Methods 54
Matchcode Matrix 216 Gathering 54
Reports 216 Middle Name Component 26
Matchcode Quality 219 Middle/Nickname Component 26
Reports 219 Minimum Value 54
Matchcode Rules 80 Gathering 54
Merge/Purge 80 Modify Structure 275
Matchcodes 25 Multi Threading 330
Matchcode Editor 289 Multi-Buyer Dupe Matrix 208
Matching 295 Reports 208
Accurate Near 295 Multi-Buyer Quality Matrix 213
Containment (matching) 295 Reports 213
Fast Near 295 MultiBuyer Source Code 102
Frequency (Matching) 295
Phonetex 294
Soundex 294 -N-
Matching Strategies 293
Neighborhood Sorting 23
Maximum Value 54
New Features 6
Gathering 54
New Setup 63
Maximum Words 30
Numerics Only 30
mceditor.exe 288
Merge 74
Add New Output Field 74
Input Field 73
-O-
Output Field 72 ODBC 15
Output Field Matching 70 File Requirements 15
Output Table 75 One Blank Field 30
Reports 229 Open Setup 64
Optimization 44 PO Box Component 26
Component Sequence 44 Prefix Component 26
Fuzzy Algorithms 44 Print Matchcode 307
Unnecessary Combinations 44 Print Records 251
Unnecessary Components 44 Print Rows 189
Optimize Matchcode 308 Analyze 189
Options 196 Processing 177
Analyze 196 CASS 177
CASS 154 File Update 174
Options CASS 321 Merge 165
OR Values 54 Merge/Purge 168
Gathering 54 Purge 171
Oracle 20 Processing Mode 273
File Requirements 20 Processing Speed 324
Other Software 361 Processing Summary 201
Other Uses for Swap Matching 50 Reports 201
Output 110, 159 Products 361
CASS 110, 159 Properties 36
Analyze 192 Both Blank Fields 36
Output Field 149 Component 30
File Update 149 Purge 124
Merge 72, 74 Advanced File Type 124
Merge/Purge 96 Advanced Rank 125
Merge/Purge 93 Advanced Tab 132
Output Field Matching 70 Advanced Table Type 124
Merge 70 Field Mapping for Matchcode 128
Output Fields 93 General Tab 118
Merge/Purge 93 Information About 122
Output Fields CASS 320 Input Field 131
Output Table 75 Input Sources 122
Merge 75 Input Tables 121
Merge/Purge 99 Matchcode 119
Matchcode Component 130
Ranking 120
-P- Setup 118
Purge Processing 171
Partitioning 23
Peoplesmith Software 361
Phone Component 26
Phonetex 30, 294
-Q- -S-
Quality Matrix 211 Save Setup 65
Reports 211 Save Setup As 66
Source Code Reports 227 Scatter/Gather 346
Merge/Purge 346

-R- Scattering 56
Screen Savers 310
Screen Updates 324
Random Select 266
Split File 266 Search & Replace 256
Set Filter 242
Ranking 139
File Update 139 Set Index 243
Merge/Purge 81 Settings 233
Report 233
Purge 120
Short/Empty 297
Recall Records 253
Setup 152
Replace Records 254
CASS 152
Report Settings 233
File Update 136
Reporting Fonts 334
Merge 68
Reports 230
Merge/Purge 79
CASS 230
Purge 118
Dupe Matrix 206
File Summary 203 Short/Empty Match 30
Matchcode Matrix 216 Short/Empty Settings 297
Matchcode Quality 219 Shortest 54
Gathering 54
Merge Matrix 229
Multi-Buyer Dupe Matrix 208 Show/Hide Rows 186
Analyze 186
Multi-Buyer Quality Matrix 213
Processing Summary 201 Size 30
Quality Matrix 211 Software License 364
Source Code 220 Sort Records 259
View 198 Sort Rows 188
Analyze 188
Reports: Templates 235
Soundex 30, 294
Requirements 13
File 13 Source Code Format 101
Merge/Purge 101
System 9
Source Code MultiBuyer 102
Source Code Reports 220
Dupe Matrix 225
Source Code Reports 220 Truncations issues 344
File Summary 222
Quality Matrix 227
Specifying Matchcode Combinations 33 -U-
Split File 261
Unnecessary Combinations 44
<n>th Select 264
Optimization 44
by Records/Bytes/Files 262
Unnecessary Components 44
Contents of Field 268
Optimization 44
Random Select 266
Update Methods 54
SQL Server 19
Update Tab 147
File Requirements 19
Updates 314
Stack Group 54 CASS 12
Stacking 54
Uppercasing 30
Stacking settings 328
User Settings 310
Start 30 Default Output Table 317
State/Province Component 26 Field Mapping 316
Status Codes 329 Field Naming 315
Street Name Component 26 File Location 313
Street Number Component 26 Interface 310
Street Post-Directional Component 26
Using CASS CD 10
Street Pre-Directional Component 26
Street Secondary Component 26
Street Suffix Component 26 -V-
Subscription 12
CASS 12 Vertical Display 191
Suffix Component 26 Analyzer 191
Swap Match 30, 306 View File 336
Swap Matching 50 View Reports 198
System Requirements 9 Visual FoxPro 13
Vowels Only 30

-T- -W-
Table vs. Database 13
Templates 235 Warnings 323
Tip of the Day 310 User Settings 323
Tool Tips 310 Warranty 367
Toolbar size 272, 312 What's New 6
Translation Table 325
Trim 30
-Z-
Zip4 Component 26
Zip5 Component 26
Zip9 Component 26