Management Information Systems: Lab 5 - Tableau: Basic Operation
Management Information Systems: Lab 5 - Tableau: Basic Operation
Objective
• At the end of this lesson, you should able to
– Identify what data visualization is
– Perform basic operation in Tableau
Tableau
• Business Intelligence software
• Tableau excels at displaying data visually. Tableau’s
core purpose is to help you draw conclusions from
your data visually
• Support different databases and data sources,
including standard corporate databases, “big data”
data source etc.
• Analyzing data in Tableau is incredibly fast. What
used to take traditional Business Intelligence tools
hours to reveal can be discovered in Tableau in
seconds or minutes
Demonstration
• Coffee Chain
Demonstration
• From the excel file, can you easily tell what states
have products that are not making a profit?
• We can show it in Tableau in a few clicks
• The core approach to visual design in Tableau:
connect to a desired data source, and drag various
data fields to desired parts of the Tableau screen
• The result is a simple visualization that can then be
enhanced and modified by dragging additional
data fields to different destinations in the
workspace
Choose additional
matching fields
The interface
Data Pane
• After importing the data, you will see the column
names in the excel file appear on the left
• Dimensions - categorical fields, such as date,
customer and Category. Dimensions are often
discrete.
• Measures – the metrics, such as COGS, sales and
profit. They are the numbers we want to analyze.
Measures are often continuous.
Show Me
• Show Me displays a series of
thumbnail image representing the
different types of charts you can
create with just a few clicks
• If fields are selected, or you
already have a chart created that
you wish to change, Show Me will
only show chart types that are
appropriate for fields that are in
use with your current chart or that
have been selected in the Data
pane
Show Me
• Depending on the number and type of fields selected,
Show Me will highlight the recommended thumbnail with
a blue box (this is yet another example of Tableau’s
attempt to employ visual best practices for you)
• If you hover your mouse over a Show Me thumbnail, the
bottom of the dialog box will show the name of the
chart type you are hovering over, as well as the number
of dimensions and /or measures that are required for
that chart type
• If you want to use that type chart type, just Ctrl-Click the
desired dimensions /measures in the Data pane until the
desired thumbnail is enabled. Then, click the thumbnail
to create the new chart
Marks Card
• Tableau will make some default
assumptions about the type of chart
created or, more specifically, the mark
type that will be used
• For example, if you initially use a date
dimension for a chart, you will find a
line chart (and thus, a line mark type)
being chosen by default.
• If you choose no dimension but instead
choose more than one measure
(resulting in a scatter plot), you will find
the mark type defaults to an open
circle
Color Options
• If your visualization consists of only one
color, clicking Color on the Marks card will
display a color palette
• Coloring on a dimension will create distinct
color values for each dimension member
• Dragging a measure onto Color will create
a shaded-variance color result based on
the “continuous” behavior of numeric
measures
• For example, a profit measure that ranges
from negative to positive values will color
the mark in a shaded variation based on
whether profit is negative or positive
Edit Colors
• A color legend will appear
on the right
• Clicking the small arrow on
the legend and choosing
Edit Colors from the
resulting menu will present
options for customizing
color behavior
Filters
• Filtering is the process of narrowing
down your chart or graph to only the
data that is relevant to your current
need.
• No matter what visualization or
analytical tool you use, one of the first
things you will need to do is filter data
• For example, if you have an historical
data warehouse that contains 10
years of records, and almost certainly
you want to limit your view to a
specific year or a few years. Or you
only want to show certain regions,
sales that are at least $10,000
• General
– If you drag a string
dimension to the Filters
shelf, a filter dialog will
appear, showing all
members of the dimension
– To exclude certain values
in your worksheet, check
the values you wish to
exclude, as well as the
Exclude checkbox in the
lower right
– When you click OK, the
filter will be immediately
applied and the chart will
immediately update to
reflect the new filter
• Wildcard
– Allows freeform filters
using wildcard searches,
such as “Contains”,
“Starts with”, “Ends with”
and “Exactly matches”
• Top
– Filter on an aggregated
numeric value
– Limit the filter to the top
or bottom “N”
occurrences of the
dimension
• Condition
– Filters the chosen
dimension, but not on the
actual dimension
members themselves
– You may specify a
condition based on an
aggregated numeric
value
– For example, if you only
wish to include
categories that are
unprofitable, select Sum
of Profit, and specify a
less than operator and
zero comparison value
Filtering on Measure
• Filtering on a measure
presents a different type
of filter dialog that
provides numeric options
based on the range of
values within the measure
• First need to choose the
type of aggregation
calculation you wish to
use for the filter (sum,
count, and so forth)
• You can choose different
ways of bounding the
values
Aggregation function
• There are several
aggregation functions in
Tableau, such as sum,
average, max, min etc.
• To change the
aggregation function of
the Measure, you can
click the measure that is
in the shelf, then choose
Measure
Reference Line
• It may based on a variety of
calculations, using existing
measures in your chart
• You may also base the value
on a constant number that
your “hard code” into the
dialog box, or the value of a
parameter
• If you want to base a
reference line on a measure
other than the one actually
being displayed on the chart,
drag the desired measure to
Detail on the Marks first
Reference Band
• The beginning and ending
values may be based on a
variety of calculations,
using existing measures in
your chart
• You may also base the
beginning and ending
values on a constant
number that your “hard
code” into the dialog box,
or the value of a
parameter
Reference Distribution
• Rather than just specifying a
beginning and ending point for a
distribution, you may specify
several points, which are shaded
with various gradients
• Reference bands are particularly
helpful for certain chart types,
such as bullet charts and box plots
Trend Line
• A line that appears on a chart
indicating the general trend the
underlying marks on the chart are
exhibiting
• Trend lines require both the Column
and Row contain a numeric measure.
A trend line is also permitted if a
chart is based on a date or
date/time dimension
• To create a trend line, right click the
chart and choose “Trend Lines”,
“Show trend lines”
• You can choose different model
types in the “Edit Trend Lines”
Forecasting
• Tableau provides forecasting, the ability to examine
data and trends in existing date- or date/time-
based data and forecast what trend may occur in
the future
• The first requirement when forecasting is to create a
chart based on a date or date/time dimension. If
you choose to change the default discrete Year
date level of the dimension, you must choose a
continuous date level. Discrete date levels other
than Year won’t permit forecasting
Practice 1
• Based on the Coffee Chain dataset build a chart to
answer the following questions
• Q1 Show the sum of COGS of each product, which
product has the least COGS?
• Q2 Show the average of COGS in each Market,
which Market has the highest COGS?
• Q3 Show the sum of margin of each product, which
product has the highest Margin?
• Q4 Show the total expenses in each Market Size,
which market size has the highest expenses?
Practice 2
• Based on the Coffee chain dataset to answer the
following questions
• Q1 Find what product type is given the highest and
the lowest sales
• Q2 Which states is doing the best and the worst in
each product type in terms of sales
Practice 3
• Based on the Coffee Chain dataset to answer the
following questions
• Q1 Show the sales of the Market from 2010 to 2013
• Q2 Show the trend of sales of the Market from 2010
to 2013
• Q3 Show the forecasting of sales of the Market in
2014
• Q4 What are the predicted sales of the 4 Market in
2014?
Reference
• Tableau Learning
• Tableau: The official guide
Hands-on Submission
• Submit you work on Moodle by the end of next
Friday 17:00 HKT (Apr 10)
Due 17:00