Tableau for Beginners (session 4)

Warning

This is currently a draft version of the materials for this training session, and so may contain minor errors and inaccuracies.

No feedback found for this session

Session outline

• In this session, we’re going to build out the GP practice size dashboard from last time
• we’re going to use some new features of Tableau to help us with that
  • Calculated Fields
  • Table calculations
  • Level of Detail (LOD) expressions

Session files

• four .csv files
  • 1 - practices.csv from PHS’s opendata portal
  • 2 - boards.csv rom PHS’s opendata portal
  • 3b - demographics_p.csv pivoted data from PHS’s opendata portal
  • 4 - postcodes.csv from the Scottish Postcode Directory - although note that the supplied file is a small subset of this larger directory

Getting started

Task

• find your dashboard from last time
  • or create a new dashboard based on the following four data sources for this week
    
• create a new, empty, worksheet

Calculated Fields

• Calculated Fields are the basic tool used to summarize data in Tableau
• let’s start with a simple example: how big is the largest GP practice?
• we’ll answer that by writing a simple formula: MAX([Practice List Size])
  • MAX is a function that finds the largest value in some data
  • [Practice List Size] refers to one of our columns of data

MAX()

Task

• right-click your Practice List Size field, and select Create > Calculated Field...
  
• name your Calculated Field largest_practice
• then use the following (case sensitive) formula to calculate the largest practice: MAX([Practice List Size])
  

Finding your calculated fields

• note that your new largest_practice field will show up at the very bottom of your sets of data:
  
• note too the equals sign in the icon, indicating a calculated field

MIN()

Task

• now repeat to find the smallest practice using MIN()
  

COUNT()

Task

• now repeat to find the number of practices using COUNT()
  

AVG()

Task

• finally, find the average practice size using AVG()
  

Adding calculated fields to worksheets

Task

• Drag any one of your new calculated fields to Text Marks, then add HB Name to the column shelf:
  

• then (in a slightly quirky way) you should be able to add each calculation by double-clicking the name of the calculated field:
  

Finding functions

• from the right-click menu, you can edit calculations
  
• note that the right-hand edge of the calculation interface contains a mini help section for the available functions
  

Combining functions

• we can use several functions in a formula to do interesting analytic work
• we’ll look at three helpers for combining functions:
  • brackets
  • +
  • STR()

Brackets

• you can combine several functions together with brackets
• the functions will run from the inside-out:
  • ROUND(AVG([dat])) will:
    • average [dat], then
    • round that result

Task

• try making a rounded average of practice list size to the nearest 100
• ROUND(AVG([Practice List Size]), -2)

+

• + joins text values together

Task

• create a new sheet with HB Name on the Rows shelf
• then create a new calculation with the formula "The first practice name is " + MIN([GP Practice Name])
• add that calcuation to Text Marks
• you may need to play with the Fit menu to see your results:
  

STR()

• if you try to use + to add text to numeric calculations, you’ll run into a problem
• STR() converts numbers to text

Task

• try "The smallest practice has " + STR(MIN([Practice List Size])) + " patients"

IF and friends

• IF something THEN action1 ELSE action2 END
• ELSEIF something else THEN action2
• IFF(something, action1, action2)

Table calculations

• Table calculations are calculations on visualizations, rather than the underlying data
• We’ll run through a quick table calculation now

Quick table calculations

Task

• build a new sheet with largest and smallest practices by board
  
• right-click your largest_practice calculation, and select Quick Table Calculation > Rank
  
• that converts largest_practice to display as a rank between 1 and 14

• clear the table calculations from the right-click menu:
  

Level of detail calculations

• LoD calculations allow you to control the level of detail that your calculations work at
• this is just a taster as this gets into advanced (and messy) territory pretty fast

Task

• create a table showing the average of each demographic group per NHS board:
  
• now add a LoD calculation that will show the average for each attribute bracket for Scotland overall using {FIXED [Attribute] : AVG([Value])}:
  

That gives us a slightly messy way of benchmarking our average for our bracket within a board to the bracket nationally:

We could also try the more ambitious:

AVG([Value]) / ATTR({FIXED [Attribute] : AVG([Value])})

(you need the ATTR function to trick Tableau into mixing and matching these different values)