SAS Programming

SAS stands for statistical analysis system

SAS programming enables you to write Sas programs to access, explore, prepare and analyse data. It can
read the data from common spreadsheet and databases and output the results of statistical analyses in
tables, graphs and RTF, HTML and PDF documents. Sas is used for statistical analysis and enables users to
perform tasks such as importing data from multiple sources as well as analysing and reporting.
SAS Programming Workflow:
1.Access Data
2.Explore Data
3.prepare Data
4.Anlyse and Report the Data
5.Export result
SAS Program Structure:

DATA STEP:
This step involves loading the required data set into SAS memory and identifying the variables (also called
columns) of the data set. It also captures the records (also called observations or subjects). the string
variables have a $ at the end and numeric values are without it.

Syntax:
Data data_set_name; #name the data set
Inputvar1,var2,var3; # define the variables in this data set
New var; # create new variables
Datalines; # enter the data
Run;
Example:
Data myfuel;
Set sashelp.fuel;
Input CPRatio EQRatio Fuel $ NOx;
Dataliner;
13 5.90 indian 7.6
12 3.54 hp 0.34
12 4.77 hp 5.12
14 2.54 indian 4.32
9 3.14 hp 4.81
12 6.54 hp 4.81
Run;
PROC STEP:
This step involves invoking a SAS built-in procedure to analyse the data.
Syntax:
Proc procedure_name options; #the name of the proc
Run;
Example:
Proc means;
Run;
THE OUTPUT STEP:
The data from the data sets can be displayed with conditional output statements.

Syntax:
Proc print data=data_set;
Options;
Run;
Example:
Proc print data=myfuel;
Where CPRatio>=10;
Run;
ACCESSING THE DATA:
PROC CONTENTS:
The proc contents procedure provides a summary of a dataset's contents, including details such as
variable names, types, and attributes (such as formats, informats, and labels). It also tells you the
number of observations and variables present in the dataset, as well as the creation date of the
dataset.
Syntax:
Proc contents data=dataset_name;
Run;
Example:
proc contents data=sashelp.cars;
run;
Accessing the data through library:
A libref is the name of the library that can be used in a SAS program to read data files.
The engine provides instructions for reading SAS files and other types of files.
The path provides the directory where the collection of tables is located.
The libref remains active until you clear it, delete it, or shut down SAS.

Syntax:

Libname libref engine “path”;

Libname libref clear; #delete the library reference.

Example :
Libname mylib base “s:/workshop/data”;
Libname statement creates a libref or library name ‘mylib’ that uses the base engine to read sas tables
located in s:/workshop/data, ‘base’ is a default engine.
Using a Library to Read Excel Files:

Libname libref xlsx “pathfile.name.xlsx”;

Example:
libname xlclass xlsx “s:/workshop/data/class.xlsx”;
IMPORTING UNSTRUCTURED DATA:
IMPORTING THE CSV FILE:
Syntax:
Proc import datafile=”pathfilename” dbms=filetype out=output_table
<replace>;
Run;
Example:
Proc import datafile=”s:/workshop/storm_damage.csv” dbms=”csv”
out=storm_damage_import replace;
Run;

IMPORTING THE XLSX FILE:

Example:
Proc import datafile=”s:/workshop/storm_damage.xlsx” dbms=”xlsx”
out=storm_damage_import replace;
Run;

Exploring Data:

PROC PRINT:
proc print lists all columns and rows in the input table by default
Obs=number of rows, Var statement=order and limits the columns listed.
Syntax:
Proc print data=input_table(obs=n);
Var col_name(s);
Run;
Example:

Proc print data=cars(obs=20);

Run;

PROC MEANS:

Proc means generates simple summary statistics for each numberic column in the input data by default.

Syntax:

Proc mean data=input_table;

Var colname;
Run;

Example:

Proc means data=pg1.storm_summary;

Var MaxWindMPH Minpressure;

Run;

PROC UNVARIATE:

Proc univariate also generates summary statistical for each numeric column in the data by default ,but
include more ddetailed statistics related to distribution and extreme values.

Syntax:

Proc univariate data=input_table;

Var col_name;

Run;

Example:

Proc univariate data pg1.storm_summary;

Var MaxWindMPH Minpressure;

Run;

PROC FREQ:

PROC FREQ creates a frequency table for each variable in the input table by default. You can limit the
variables analyzed by using the TABLES statement

Syntax:

Proc freq data =input_table;

Tables col_name(s) <options>

Run;

Example:

Proc freq data=pg1.storm_summary;

Tables MaxWindMPH Minpressure;

Run;

FILTERING ROWS:

WHERE STATEMENT:

Where statement is used for filter rows. if expression is true, rows are read. If expression is false, then rows
are not read.

Syntax:

Proc procedure_name:

Where expression;

Run;

Example:

Proc print data=sashelp.cars;

Var Make Model Type MSRP MPG_City MPG_Highway;

Where Type=”SUV” and MSRP <= 30000;

Run;

FILTERING ROWS WITH BASIC OPERATORS:

Syntax:

Where expression;

Basic Operators;
 = (or) EQ

^= (or) ~= (or) NE

> (or) GT

< (or) LT

>= (or) GE

<= (or) LE

SAS Date Constant:

When an expression includes a fixed date value, use the SAS date constant syntax: “ddmmmyyyy”d, where
dd represents a 1- or 2-digit day, mmm represents a 3-letter month in any case, and yyyy represents a 2- or
4-digit year.

“ddmmmyyyy”d ------------ (“01JAN2015”d)

Example:

Proc print data=pg1.storm_summary;

Where StartDate >= “01jan2010”d;

Run;

Combine Expression using “ AND “:

Example:

Proc print data=sashelp.cars;

Var Make Model Type MSRP MPG_City;

Where Type=”SUV” and MSRP <=30000;

Run;
Combine Expression using “ OR ”:

Example:

Proc print data=sashelp.cars;

Var Make Model Type MSRP MPG_City;

Where Type=”SUV” and MSRP <=30000;

Run;

IN Operator:

WHERE col-name IN(value-1<...,value-n>);

WHERE col-name NOT IN (value-1<…,value-n>);

Special WHERE Operators

WHERE col-name IS MISSING;

WHERE col-name IS NOT MISSING;

WHERE col-name IS NULL;

WHERE col-name BETWEEN value-1 AND value-2;

WHERE col-name LIKE "value";

WHERE col-name =* "value";

Filtering Rows with Macro Variables

%LET macro-variable=value;

All statement in the macro variables are start with “%”

Macro-variable – name of the macro variable

%LET statement ---create a macro variable

Macro variables can be referenced in a program by preceding the macro variable name with an &.
If a macro variable reference is used inside quotation marks, double quotation marks must be used.
Eample:

%let cartype=Wagon;

Proc print data=sashelp.cars;

Where Type=”&cartype”;

Var Type Make Model MSRP;

Run;

Proc means data=sashelp.cars;

Where Type=”&carType”;

Var MSRP MPG_Highway;

Run;

Proc freq data=sashelp.cars;

Where Type=”&carType”;

Tables original Make;

Run;

FORMATTING COLUMNS:

Formats are used to change the way values are displayed in data and reports.

Syntax:Proc print data=input_table;

Format col_name(s) format;

Run;
<$>format-name<w>.<d>

COMMON FORMAT FOR NUMERIC VALUES:

FORMAT NAME EXAMPLE VALUE FORMAT APPLIED FORMATTED VALUE

w.d 12345.67 5. 12346

w.d 12345.67 8.1 12345.7

COMMAw.d 12345.67 COMMA8.1 12,345.7

DOLLARw.d 12345.67 DOLLAR10.2 $12,345.67

DOLLARw.d 12345.67 DOLLAR10. $12,346

YENw.d 12345.67 YEN7. Y12,346

COMMON FORMAT FOR DATE VALUES:

VALUE FORMAT APPLIED FORMATTED VALUE

21199 DATE7. 15JAN18

21199 DATE9. 15JAN2018

21199 MMDDYY10. 01/15/2018

21199 DDMMYY8. 15/01/18

21199 MONYY7. JAN2018

21199 MONNAME. MARCH

21199 WEEKDATE. MONDAY,MARCH04,2023

Example:

Proc print data=pg1.class_birthdate;

Format Height2. Weight3. Birthdate date9.;

Run;

SORTING DATA:

Proc sorts the rows in a table on one or more character or numeric columns.

The out= option specifies an output table. Without this option, PROC SORT changes the order of rows in
the input table.

The BY statement specifies one or more columns in the input table whose values are used to sort the rows.
By default, SAS sorts in ascending order.

Syntax:

Proc sort data=input_table <out=output_table>

By <descending> col_name(s);

Run;

Example:

Proc sort data=pg1.class_test out=test_sort;

By descending name;

Run;

REMOVING THE DUPLICATE DATA FROM PARTICULAR COLUMN USING SORT STATEMENT:

SYNTAX:

PROC SORT DATA=INPUT_TABLE <OUT=OUTPUT_TABLE>

NODUPKEY <DUPOUT=OUTPUT_TABLE>;
BY COL-NAME(S);

RUN;

EXAMPLE:

Proc sort data=pg1.class_test out=test_clean

Nodupkey dupout=test_dups;

By name;

Run;

REMOVING THE DUPLICATE DATA FROM ALL COLUMNS USING SORT STATEMENT:

SYNTAX:

PROC SORT DATA=INPUT_TABLE <OUT=OUTPUT_TABLE>

NODUPKEY <DUPOUT=OUTPUT_TABLE>;

BY_ ALL_;

RUN;

EXAMPLE:

Proc sort data=pg1.class_test out=test_clean

Nodupkey dupout=test_dups;

By _all_;

Run;

FILTERING AND SUBSETTING DATA:

Syntax:

Data output_table;
Set input_table;

Where expression ;

Format col_name format;

Run;

Example:

Data myclass;

Set sashelp.class;

Where age>=15;

Keep Name Age Weight Height; # keep the column in the new table

Format Height2. Weight3.

Drop weight; #remove the column

Run;

CREATING THE NEW COLUMN:

Syntax:

Data out-table;

Set input_table;

New_column=expression;

Run;

Example:

Data cars_new;

Set sashelp.cars;
Where origin ne “USA”;

Profit=MSRP-Invoice;

Source=”Non-US cars”;

Format profit dollar10.;

Keep Make Model MSRP Invoice profit source;

Run;

FUNCTION TO CREATE COLUMN:

Syntax:

Data output_table;

Set input_table;

New_column=function(arguments);

Run;

FUNCTION TO CALCULATE THE SUMMARY SATISTICS:

Common Numeric Functions What it Does

SUM (num1, num2, ...) Returns the sum of nonmissing arguments.
MEAN (num1, num2, ...) Returns the arithmetic mean (average) of nonmissing arguments.
MEDIAN (num1, num2, ...) Returns the median value of nonmissing arguments.
RANGE (num1, num2, ...) Returns the range of the nonmissing values.
MIN (num1, num2, ...) Returns the smallest nonmissing value from a list of arguments.
MAX (num1, num2, ...) Returns the largest value from a list of arguments.
N (num1, num2, ...) Returns the number of nonmissing numeric values.
NMISS (num1, num2, ...) Returns the number of null and SAS missing numeric values.

CHARACTER FUNCTION:

Character Function What it Does

UPCASE (char)
Changes letters in a character string to uppercase or lowercase
LOWCASE(char)
PROPCASE (char,<delimiters>) Changes the first letter of each word to uppercase and other letters to lowercase
Concatenates character strings and removes leading and trailing blanks from each
CATS (char1, char2, ...)
argument
Character Function What it Does
SUBSTR (char, position,
Returns a substring from a character string
<length>)
Eample:

Data storm_new;

Set pg1.storm_summary;

Drop Type Hem_EW Hem_NS Min pressure Lat Lon;

Basin=upcase(basin);

Name=propcase(name);

Hemisphere=cat(Hem_NS,Hem_EW);

Occean=substr(Basin,2,1);

Run;

DATE FUNCTION FOR CREATING COLUMNS:

Date Function What it Does

MONTH (SAS-date) Returns a number from 1 through 12 that represents the month
YEAR (SAS-date) Returns the four-digit year
DAY (SAS-date) Returns a number from 1 through 31 that represents the day of the month
WEEKDAY (SAS-date) Returns a number from 1 through 7 that represents the day of the week (Sunday=1)
QTR (SAS-date) Returns a number from 1 through 4 that represents the quarter

Date Function What it Does

Returns the current date as a numeric SAS date value (no argument is required
TODAY ()
because the function reads the system clock)
MDY (month, day, year) Returns a SAS date value from numeric month, day, and year values
YRDIF (startdate,
Calculates a precise age between two dates
enddate, 'AGE')
Example:

Data Storm_new;

Set pg1.storm_damage;

Drop summary;

Yearspassed=yrdif(Date,today(),”age”);
Anniversary=mdy(month(Date),day(Date),year(today()));

Format yearspassed 4.1 Date anniversary mmddyy10.;

Run;

CONDITIONAL STATEMENTS:

CONDITONAL PROCESSING WITH IF-THEN LOGIC:

Syntax:

IF expression_1 THEN statement_1;

IF expression_2 THEN statement_2;

Example:

Data cars;

Set sashelp.cars;

If MSRP<30000 then cost _Group=1;

If MSRP>=30000 then Cost_Group=2;

Keep Make Model Type MSRP Cost_Group;

Run;

CSYNTAXONDITIONAL PROCESSING WITH IF-THEN /ELSE:

Syntax:

IF expression THEN statement;

<ELSE IF expression THEN statement;>

<ELSE IF expression THEN statement;>

Example:

Data cars;

Set sashelp.cars;

If MSRP<20000 then Cost_Group=1;

Else if MSRP<40000 then Cost_Group=2;

Else if MSRP<60000 then Cost_Group=3;

Else Cost_Group=4;
Keep Make Model,Type, MSRP Cost_Group;

Run;

CREATING CHARACTER COLUMNS WITH THE LENGTH STATEMENT:

Length char_column $ length;

Example:

Length CarType $6;

PROCESSING MULTIPLE STATEMNTS WITH IF-THEN/DO:

Syntax:

IF expression THEN DO;

<executable statements>

END;

ELSE IF expression THEN DO;

<executable statements>

END;

ELSE DO:

<executable statements>

END;

Example:

Data cars2;

Set sashelp.cars;

Length Cost_Type $ 4;

If MSRP<20000 then Cost_Group =1 and Cost_Type=”Low”;

Else if MSRP<40000 then Cost_Group=2 and Cost_Type=”Mid”;

Else Cost_group=3 and Cost_Type=”High”;

Run;
ANALYZING AND REPORTING ON DATA

ENHANCING REPORTS WITH TITLES

TITLE is a global statement that establishes a permanent title for all reports created in your SAS session.

You can have up to 10 titles, in which case you would just use a number 1-10 after the keyword TITLE to
indicate the line number. TITLE and TITLE1 are equivalent.

Titles can be replaced with an additional TITLE statement with the same number. TITLE; clears all titles.

Syntax:

TITLE <n>”title_text”;

FOOTNOTE<n>”footnote-text”;

Title1 “class Report”;

Title2 “All Students”;

Footnote1 “Report Generated on 01sep2018”;

Proc print data=pg1.class_birthdate;

Run;

APPLYING TEMPORARY LABELS TO COLUMNS:

Syntax:

LABEL col_name=”label_text”;

Example:

Proc means data=sashelp.cars;

Where type=”sedan”;

Var MSRP,MPG_Highway;

Label MSRP=”Manufacturer suggested retails price”

MRG_Highway=”Highway Miles per Gallon”;

Run;

PROC FREQ data=sashelp.heart <PROC options>;

tables chol-status/ options

run;

CREATING FREQUENCY REPORTS AND GRAPHS:

Syntax:

Ods graphics on;

Proc freq data=input_table<proc -options>;

Tables col-name(s) /options;

Run;

Proc freq statements options:

Order=freq|formatted|data nlevels

Tables statemnts options:

Nocum

Nopercent

Plots=freqplot(must turn on ods graphics)

Out=output_table

Example:

Ods graphics on;

Ods nonproctitle;

Title “frequency report for basin and storm Month”;

Proc freq data=pg1.storm_final order freq nlevels;

Tables Basinname stateDate /nocum plots=freqplot(orient=horizontal scale =percent);

Format startDate monname.;

Label BasinNAme=”Basin” StartDate=”Storm Month”;

Run;

Title;

Ods proctitle;
CREATING TWO-WAY FREQUENCY REPORTS:

Syntax:

Proc freq data=input_table<options>;

Tables col-name*col-name </options>;

Run;

Proc freq statemnts options:

Noprint

Tables statements options:

Norow ,nocol,nopercent

Crosslist ,list

Out=output-table

Example:

Proc freq data=pg1.storm_final noprint;

Tables BP_status*chol_stauts /out=stormcounts;

Format StartDate=monname.;

Run;

CREATING A SUMMARY STATISTICAL REPORT:

proc mean is used for calculating basic summary statistics

var statement is used to specify the numeric column to analyze.

<stat-list> is used to specify the statistics. that we want to calculate and how they should be display.

Syntax:

Proc means data=input-table stat-list;

Var col-name(s);

Class col-name(s);

Ways n;

Run;
Example:

Proc means data=pg1.storm_final mean median min max maxdec=0;

Var MaxwindMPH;

Class BasinName StormType;

Ways 0,1;

Run;

CREATING AN OUTPUT SUMMARY TABLE:

Syntax:

Output out=output-table <statistic=col-name>

Example:

Proc means data=sashelp.heart noprint;

Var weight;

Class col_status;

Ways 1;

Output out=heart_stats mean=AvgWeight;

Run;

EXPORTING DATA

proc export can export a SAS table to a variety of file formats outside SAS.

Proc export data=input_table outfile=”output_file”

<dbms=identifier><replace>;

Run;

EXPORTING DATA TO AN EXCEL WORKBOOK:

To export data using libname engine.

Syntax:

Libname libref xlsx “path/file.xlsx”;

 <use libref for output table(s)>

Libname libref clear;

Example:

Libname myxl xlsx “&output/cars.xlsx”;

Libname myxl clear;

EXPORTING REPORTS:

The SAS Output Delivery System (ODS) is programmable and flexible, making it simple to automate the
entire process of exporting reports to other formats.

Ods<destination><destination-specifications>;

/*SAS code that produces output */

Ods<distination>close;

EXPORTING RESULTS TO excel:

The ods excel destination creates an .xlsx file.

By default, each procedure output is written to a separate worksheet with a default worksheet name. The
default style is also applied.

Use the style= option on the ods excel statement to apply a different style.

Use the options(sheet_name=’label’) on the ods excel statement to provide a custom label for each
worksheet.

Syntax:

Ods excel file=”filename.xlsx” style=style options(sheet_name=”label”);

/* SAS code that produces output*/

Ods excel options(sheet_name=’label’);

/* SAS code that produces output */

Ods excel close;

Example:

Ods excelfile=”&outpath/wind.xlsx” style= style=sasdocprinter

Options(sheet_name=”windstats’);

Title “wind statistics by Basin”;

Ods nonproctitle;

Proc means data=pg1.Strom_final min mean median max maxdec=0;

Class BasinName;

Var MaxWindMPH;

Run;

Ods excel options(sheet_name=’Wind Distribution”);

Title “distribution of Maxmum Wind”;

Proc sgplot data=pg1.strom_final;

Histogram MaxWindMPH;

Density MaxWindMPH;

Run;

Title;

Ods proctitle;

Ods excel close;

EXPORTING RESULTS TO CSV:

By using csvall destination, the ods statement are exporting the result of the procedure.so by using ods
csvall with proc print, you can specify the order and format of the columns in your output csv file.

Syntax:

Ods csvall file =”filename.csv”;

/*SAS code that produces output */

Ods csvall close;

Example:

Ods csvall file =”&output/cars.csv”;

Proc print data=sashelp.cars noobs;

Var Make Model Type MSRP MPG_City MPG_Highway;

Format MSRP dollar8.;

Run;

Ods csvall close;

EXPORTING RESULTS TO PDF:

The ods pdf destination creates a .pdf file.

The pdftoc=n option controls the level of the expansion of the table of contents in PDF documents.

The ods proclabel statement enables you to change a procedure label.

Syntax:

ods pdf file=”filename.pdf” style=style startpage=no pfttoc=1;

ods proclabel=”label”;

/* SAS code produces output */

Ods pdf close;

EXPOTING RESULT TO POWERPOINT AND MICROSOFT:

Ods powerpoint file=”filename.ppt”

Style=style;

/*SAS code that produces the output*/

Ods powerpoint close;

Ods rtf file=”filename.pptx” startpage=no;

/* SAS code produces the output *?

Ods rtf close;

USING STRUCTUREES QUERY LANGUAGE (SQL) IN SAS:

SQL used in prepare data and analyse and report on data in SAS.

Proc sql creates a report by default.

The SELCET statement describes the query. List columns to include in the results after SELECT, separated by
commas. The FROM clause lists the input table(s).

The ORDER BY clause arranges rows based on the columns listed. The default order is ascending. Use DESC
after a column name to reverse the sort sequence.

PROC SQL ends with a QUIT statement.

Syntax:

proc sql;

select col-name ,col-name

from input-table;

quit;

Example:

Proc sql;

Select Name Age Height*2.54 as HeightCM format=2.3 BirthDate format=date9.;

From pg1.class_birthdate;

Quit;

SUBSETTING DATA:

The condition works on where clause.

Where clause in SAS works in data step.

Where expression is not separate statements from SQL but it is clause add to select statement.

Syntax:

WHERE expression;

SORTING DATA

In SAS, we use proc sort to sort the data.

We use order by clause to sort the data in ascending and descending order.

Syntax:

ORDER BY col-name <DESC>

Example:

Proc sql;

Select Name ,Age, Height,Birthdate format=date8.;

From pg1.class_birthdate;

Where age>14;

Order by Height desc;

Quit;

CREATING A TABLE IN QUERY:

We use CREATE TABLE is used to create table

Syntax:

CREATE TABLE table-name as

Example:

Proc sql;

Create table work.myclass as

Select Name,Age,Height

From pg1.class_birthdate;

Where Age >14;

Order by Height desc ;

Quit;

DELETEING THE DATA:

We use DROP TABLE to delete the table .

Syntax:

DROP TABLE table-name;

Example:

Proc sql;

Drop table work.myclass;

Quit;

JOINING TABLES USING SQL IN SAS:

An SQL inner join combines matching rows between two tables.

The two tables to be joined are listed in the FROM clause.

Syntax:

FROM from table1 INNER JOIN table2

ON table1.column = table2.column

Example:

INNER JOIN:

Proc sql;

Select Grade,Age ,Teacher

From pg1.class-update inner join pg1.class-teacher

On class-update=class.-teacher;

Quit;

LEFT JOIN:

Proc sql;

Select Grade,Age ,Teacher

From pg1.class-update left join pg1.class-teacher

On class-update=class.-teacher;

Quit;

OUTER JOIN:

Proc sql;

Select Grade,Age ,Teacher

From pg1.class-update outer join pg1.class-teacher

On class-update=class.-teacher;

Quit;

USING TABLE ALIASES:

Assign an alias (or nickname) to a table in the FROM clause by adding the keyword AS and the alias of your
choice. Then you can use the alias in place of the full table name to qualify columns in the other clauses of
a query.

Syntax:

FROM table1 AS alias1 , table2 AS alias2

Example:

Proc sql;

Select u.Name,Age,Grade,Teacher

From pg1.class-update as u inner join pg1.class-teacher as t;

On u.name = t.name;

Quit;