Introduction to Tableau

Tableau Tutorial: Do This on Your Own

Getting_started_transcript

Video on getting started

Dataset for the Tutorial

Build Your First Tableau Chart - Deaths by Counties

Build a basic Covid chart

1. Use the master file

2. Create a folder on your computer for the semester: Data_Spring21. Put all of your class assignments and data in that folder. Put master_file.csv in that folder

3. Tableau: Connect | To A File | Text file | master_file.csv

–this loads master_file.csv into Tableau

4. Data Interpreter (top left column, under “Files”)

5. Check the file formats.
   –Are Words as ABC?
   –Are the Numbers as #?

6. New Worksheet. Click Sheet 1 (bottom left)

7. Drag county_nam to rows.

8. Drag deaths to columns. Check that measure is SUM

9. deaths to colors.
   –Click colors. Green-Red diverging. Reverse colors

10. Drag total deaths to labels

11. Filter for a date in time
    –Date to Filter. Individual date. Pick yesterday.

12. Sort Descending. See toolbar, sorting icon, down arrow

13. Formatting.
    –Headline: Click “Sheet 1.” Write a headline Your end product should look somewhat like the chart below:

14. Export a jpg version of your chart
    –Worksheet | Export | Image

15. Engage in excessive self-congratulatory behavior.

Formatting a Graphic

Here are some common formatting errors on graphics:

--Round labels to whole numbers
--Format the Y axis in percent    
--Caption not needed. Delete   
--Delete legend.   

Click on image to expand

Filters and Calculations

Using Filters:
1) Drag County Nam to Filters, click OK
2) On the County Filter, click down arrow, show filter. It will appear on the right column as a list of county names
3) On the right column, down arrow, select Multiple Dropdown
4) Edit Title and Type some instructions to your readers:
“Select Your County
A) Select All to clear
B) Select your county”

Add a calculation rate: Deaths Per 10,000

1. Duplicate Sheet
   –Left Click on the tab, duplicate
   –This copies your whole work to a new sheet

2. Calculate a Deaths Per 10,000
   –Click on Deaths - down arrow | Create | Calculated Field
   –Name the Calculation “Deaths Per 10,000” 
   –Divide Total Deaths into Population: 
   –Add a Division Symbol after ([Deaths]/[Pop 5Yr Est 2018])*10000

3. Drag Deaths Per 10,000 to Columns –Remove Deaths
   –Sort

4. Check your colors and labels.
   –Drag Deaths Per 10,000 to Colors
   –Drag Deaths Per 10,000 to labels

5. Headline, Format Legends, Export jpg

6. Your graphic should look something like the image below.

Your Turn

Create calculated fields and make the following charts:

--Tested/population, percentage for the most recent five days.  
--Active cases per 10,000 residents
--A stacked bar chart with Ever on a Vent and Ever Hospitalized

Post on a Google Doc, include link in your weekly memo.

More on Tableau, Basic Excel Skills: FBI Data

More Tableau and Excel skills training: FBI Data

Tableau Calculations

Calculations sample workbook

Introduction to calculations video

Intro to calculations, video, link 2

Tableau for Data Journalism Lecture Week 3: Data joining, blends and table calculations video

Simple calculations video

Introduction to Calculations Transcript

Showing Context – Total In Tableau

Format Numbers in Tableau

Create a Table Calculation:

This exercise uses 2018 student debt data for Arkansas

Import the data into a new Tableau workbook

Add a new column that divides graduate debt into enrollment

--In Measures Data Tab, Left Click on GRAD_DEBT_MDN, Create Calculated Field
--Put in this formula: sum[GRAD_DEBT_MDN]/[UGDS]
-–Rename This as: Grad Debt Per Capita Enrollment
--Drag INSTNM to Rows
--Grad Debt Per Capita Enrollment to Columns
--Format with a Title, add appropriate annotations.
--Color the graphic, include labels.

Build a Histogram

This is a detailed walk-through of how to add a cumulative distribution to a histogram in Tableau.

This process involves finding a median value, creating bins to reflect that value, creating a histogram.

1. Use the master file

2. We’ll use population. Determine the median value.

   –Drag Pop 5Yr Est 2018 to Rows. Down arrow pill, switch from sum to median.
   –Filter out state total. Drag County Nam to Filters, select all except Arkansas_all_counties.
   –Filter for a single date. Mydate to Filter, select Oct 11 –Write down that value: 18,313
   –We have 75 counties in Arkansas. That means half are below the population of 18,313.
   –Knowing this, we want to have some detail on our graph that’s useful. We’ll select 5,000 as the size of the bins. It’s arbitrary but this helps you determine visualization.

3. Create a binned calculation. In Data select Pop 5Yr Est 2018 | Create | Bins | In “Size of bins” write 5000

   –Drag that new Pop 5Yr Est 2018(bin) into Columns

A distribution chart appears

Make sure Rows has Count: “CNT(Pop 5 Yr2018)”

4. Drag county_nam to colors. Individual counties are colored –Drag Pop 5Yr Est 2018 to Tooltip. Population now is visible in Tooltip. –Edit tooltip, remove “Count of Pop 5Yr Est 2018:”

Your Turn

Build a histogram from CDC vaccination data

How to Tweak A Filter to Display Properly

Dual Axis Graphic

A version of this chart was published on the ArkansasCovid Twitter feed

Dual Axis Step-By-Step Tutorial

Use this data source

Import data to Tableau    
Drag "Sheet1" to data pane   
Create a visualization: New Worksheet   
Date to Columns   
Expand Date: Click + until you get to Day  
Delete Blue "Quarter" Pill  
Active Cases Total to Rows    
Confirmed Active Cases to Rows   
Select Confirmed Active Cases "Green Pill"  
    --down arrow: Dual Axis Chart   

Change to Active Cases to Lines  
Marks Card, select SUM(Active Cases) panel   
Change Automatic to Line (down arrow from Automatic)  

Change to Confirmed Cases to Circles  
Marks Card, select SUM(Confirmed Cases) panel   
Change Automatic to Circle (down arrow from Automatic)  

Colors  
Marks Card  
SUM(Confirmed Cases)  
Remove Active cases from color and drag Confirmed cases to color.   
Change color to purple  

Align the Y axes:  
(You have to do this in Tableau Desktop)  
Left click, synchronize axis  

Your Turn: Dual Axis, Deaths and New Cases

Use this data source You’ll use the same spreadsheet but a different tab on the sheet.

In Menu, Select Data | New Data Source |  active_cases_jan_27.xlsx   
Drag "countysnapshot" to data pane  
Build a dual axis chart with Total Deaths and New Cases Today 
Make the Deaths as Red Circles and the New Cases as Blue Columns  
Visualize by Months  

More information: See the Tableau Dual Axis chart tutorial

More documentation on dual axis

Your Turn:

Build a dual axis chart with the vaccine data: All_Doses_Received and All_Doses_Given

Chart Monthly Differences, COVID-19

1. Import master_file.csv

2. Sheet 1:
   –Mydate to Columns.
   –Active Cases, Deaths, Positive to Rows
   –Filter: County Nam to Filter: Select Arkansas_All_Counties
   –Mydate to Filter: Individually select the last day of each month. March 31 2020, April 30, 2020 etc.

–New: Or create a calculated field from Mydate. Call it End of Month Date.
Copy the following formula for the calculation: DATEADD(‘day,’-1,DATEADD(‘month,’1,DATETRUNC(‘month,’[Mydate])))
Drag End of Month Date to Columns. Remove the Mydate

–New: On Columns, End of Month Date, click down arrow, select Day option so the date is Month/Day/Year
You now have three charts, aggregated by last days of the month.

A quick video on the process:

Save and Headline Your Graphic.

Export Filtered Values from Tableau, Create New Calculations

1. Left click on the chart. Select view data. You’ll see a monthly summary of the data: Year, Month, Active Cases, Deaths, Positive. 19 rows

2. Export all. creates a new mini spreadsheet

3. Open spreadsheet in Excel
   copy underlying data into a new tab

Sort spreadsheet in descending order, Sept. 2021 as first value

6. Create three new columns: Active_Diff Deaths_Diff Positive_Diff

7. Active_Diff:
   -subtract September 2021 from August 2021
   –Check the math is correct
   –copy formula down to October 2020.
   –Copy formula over for Deaths_Diff, Positive_Diff

Save your work

Import new data back to Tableau

1. Select new worksheet
   
2. Menu Bar: Data | New Data Source
   
3. Import your new calculations
   
4. build a vialuzation
   –Dates in Column (Year and Month)
   –Positive Diff, Active Diff, Deaths_Diff in rows.
   –Marks Card, change to bar charts.
   You have three charts

Joins in Tableau

This exercise builds on the “Turning a PDF into a Spreadsheet” in the Excel chapter.

We are joining school data from the ADH Covid reports to the school population data and creating per-capita Covid calculations.

1. Download the school population data:
   –Examine the data. Which field has the names of the schools?

2. In Tableau, select a new sheet.

Add a new Data Source (Data | New Data Source)

Add the School_Districts_FY31_Districts_EnrollmentByRace.csv

3. Join with the ADH School data.

Look in Upper Right on Data Source pane for “Connections.” Select “Add”

–Add your ADH school data “School_Report_10162021.xlsx”

–In the main data pane window, Double Click on “School_Districts_FY31_EnrollmentByRace.csv”

–Drag the ADH data to the data pane and a Join window appears.

–Select “inner join” on the two fields that have the school names. Select the names from a drop-down dialog box.

4. Examine the results.

How could you maximize the number of matches? Look carefully at the columns you were trying to match.

–Describe your strategy to clean the data so the names will match.

More Tableau Materials on Next Page