Data Visualization
• First step in data analysis is understanding what the data mean.
• Measurements levels are classifications of variable.
• Important to understand these as they determine the arithmetic and
statistical operations that can be carried out on the data.
• Four categories of measurement levels:
Nominal
Ordinal
Interval
Ratio
• A single variable can be measured at different measurement levels.
3
[Link]
4
Weakest
Level Description Example
Nominal Distinct categories, no ranking is implied; nothing can be implied gender,
about the magnitude or quantitative difference between the country
values (if numbers used)
Ordinal Ranking is implied; nothing can be inferred about the differences Placing in
between the categories. Ordering only implies which category is race,
greater/preferred – not by how much
Interval Ordered, the difference between measurements is a meaningful Degrees
quantity, no true zero point, every additional unit of measurement Celsius,
represents the same amount (interval). Shoe Size
Ratio Measurements can have a value of zero which represents the height, age,
absence of the quantity being measured length
5Strongest
6
• Categorical Variables:
Variables on which calculations are not meaningful
e.g. Nominal and ordinal
• Metric Variables:
Calculations are meaningful on these
E.g. Interval and ratio
• Discrete and continuous is another classification
• Some attributes can be classified in different ways
e.g. Age, Date
• How an attribute is defined will determine the techniques that can be
applied
• Tableau uses dimension and measure 7
Connecting to Data
Foundations for building visualizations
Visualizing the data
Bar Graphs
Line Graphs
Geographic Visualizations
Creating Variables
Using Parameters
• Connect -> To a File
-> Text file
• Select file to open
• You can preview the
file
• Select Sheet1 on
bottom left to start
visualizing
1. Menu
2. Toolbar
3. Sidebar contains tabs
Data tab – Data pane
Dimensions
Measures
1 Analytics tab – Analytics
pane
2 4. Shelves and Cards
4 Shelves
3 Columns
5 6 Rows
Pages
Filter
Marks Card –
contains additional
shelves, Colour Size,
Text, Detail and
Tooltip
5. The Canvas
6. Show me
7. Edit Data Source or
7 navigate between sheets
9
8 and dashboards
8. Status bar
9. Sheet sorter
Data in the data pane can be dragged onto canvas or the various shelves and cards
– Rows, Columns, Colour or Size.
Measures
Measure are continuous;.
Values that are aggregated i.e summed, averaged, counted or have a min or max
Colour – Green; Colouring – gradient (flows from first to last)
Dimensions
Dimension are categorical/discrete;
Values that determine level of detail at which the measures are aggregated. Slicing the
measures or creating groups into which measures fit
Colour – Blue; Colouring – distinct colour for each category
Open Tableau
Connect to Text file Superstore
Create a Variable called Region – create a Calculated field
Q - Sales of each department in the
Parishes
Variables - Sales, Parish, Departments
Move departments to Colour Marks
Move parish to Colours Marks
Experiment with swap and sort icons
on the Toolbar
Level of detail – Based on the
dimensions
Put Parish as filter
Experiment with Marks – Colour (Put
department).
Q - Compare sales in the different
parishes by county
Experiment with Marks Cards
Colour
Size
Label
Formatting
Title
Caption
X- Axis
Y-Axis
Put Mark type as circle or use Show me
Put variables as shown in figures
Sum(Sales)
Department
Region
To put two variables in a mark press shift and place it there.
Tableau creates two variables measure names and measure
values to build certain views that involve multiple
measures.
Measure name - contains the names of all measures in your data,
collected into a single field with discrete values
Measure values - contains all the measures in your data, collected
into a single field with continuous values.
Drag Department in Columns
Drag Region in Rows
We want to display both profit and sales figures
Drag measure values in Text – select only
Profit and Sales
Drag measure names in Rows
Put Dept and Customer segment in rows and Sales in columns
Drag profit on x-axis [drag until the mouse changes to a double ruler]
Note both Sales and Profit are on same axis
Show Sales and Profit in different colours - drag measure names in marks-color
Rename the “value” axis dollars – right click on the x-axis region
Unselect “Show Header” when the mouse is over profit/sales on y-axis on the
graph
Put Dept and Customer segment on
• Open the second file Profit
target
• From Sheet1 Profit Target
• Drag Measure values
on Column
• Drag Measure Names
to Colour Marks
• Put Measure Names as
Size in Marks.
• Select Sheet2 Superstore
• Put Profit Ja as Measure
Values (below marks
card). You should have
only two variables.
• In Row put parish
• Edit colour to orange
and grey
• Fix Size
• Analysis-> Stacks
marks -> off
When to use
Movement of values over time
How two continuous variables
are related
Sales over time
Sales and order date
Change year to Quarter (2nd
option)
Experiment with Colour Marks
Put region in colour
Put category in colour
[If too many values then line graph
is confusing]
Dates as Dimensions and measures
Hierarchy in dates
Year- Quarter – Month – Week –Day
If time is included then Hours-Min
Profit margin = sum(Profit Ja)/sum(Sales Ja)
To display as % => Right click variable (Select Default Properties-> Number
format)
Select Category as row
Select Sales as column
Drag Profit on x-axis, drag Profit margin on x-axis put as different graphs.
Put Measure names in colour
Put labels on the graph
Do you like this Graph??
• Put Category in
rows
• Put Sales (Ja),
Profit (Ja), Profit
Margin in Column
• Put Measure
Names for Colour
To get labels on
Sales, Profit and
profit margin
Left click on Label
in marks card and
select Show marks
labels.
Put date field as a
dimension in pages
Parameters are useful when you want to add interactivity and flexibility to a report, or to
experiment with what-if scenarios.
Use parameters,
• If you are unsure which fields to include in your view
• Which layout would work best for your viewers.
• To let viewers choose how they want to look at the data.
For parameters consider the following things
1. Create a parameter.
2. Show parameter - The parameter control needs to be displayed so that viewers
can interact with it.
3. Use parameters in calculations or filters.
• Put Customer
Name in Filter
and you can set
how many top
customers are to
be displayed
• If you want it to
be a variable,
then create a
parameter
Click on right arrow in dimensions to create a Parameter – Top X Customers
• Right click on the
parameter and select
“Show Parameter
Control”
• It is displayed on the
right but notice it does
nothing
Customer Name in Filter
Edit Filter and Select Top
and the parameter
Now the parameter control
is working
You can click on the control
and changed slider to type
in
Create Parameter
– “Sort By” as List
Select Data type
as string
Allowable values
as List
In Value column
type Sales (JA),
Profit (JA), Profit
Margins
Show the
Parameter Control
by right clicking
the Parameter
Change values in
the control, it does
nothing as yet
Create a Calc field –
“Sort Order” using case
statements
CASE[Sort by]
WHEN "Sales" THEN
sum([Sales (JA)])
WHEN "Profit" THEN
sum([Profit (JA)])
ELSE [Profit Margin]
END
Right click “Category”
in Row shelf.
Sort the “category” in
descending order using
the calculated field “Sort
Order”
Now see how
Parameters change