Tableau Quick Reference

General Notes

Dimensions -> Categorical variables
Measures -> Numerical variables

When you place variables in columns or rows, the first variable acts as a category for the others. So you want to go (Country)(State)(City) so that it creates good categorization levels.

To create a calculated field right click in the Dimensions/Measures area and select “Create Calculated Field”. You can then use the arrow on the right to select specific calculation types as well as drag-and-drop items into the calculation sections.


Right click and format on the chart/graph is the best path, lots of options for style here.


Dragging a variable onto the color button will give each unique example of that variable a distinct colour.

When dragging a variable you can hold control to make a copy of it, thus not removing it from columns/row.



Use text file to connect to CSV file.

To create a data extract: right click on the data silo when looking at the sheet, then click extract data. The idea is to create an extract the data. The extra can be used in lieu of the actual data by clicking the silo, which should have a changed icon.

Saving workbooks prevents lost of progress, but is not the same as saving data. Just click file -> save as.

Double click on pills to get a break down of the calculation that makes them, the drop down can be used to look for different automatic calculations.

Color section controls markers on the graph line as well.

Go to analysis -> aggregation to remove auto aggregation, which will mean the graph just plots every individual data point.

By dragging a variable to colour you can further split the chart/graph/table into coloured variants of that variable.

There is a button for highlighting that exists on the right hand side of the screen, yada, yada.

The show me button at the top right controls what types of displays are used for charts/graphs.


Exclude to filter what you dont want to see


When Adjusting Colour ranges open the ‘advanced’ subsection to change the colour ranges.


Dashboard button is little window button next to sheet. The main function of dashboards is to create a one stop area to amalgamate that which has already been worked on.

Dashboards – Actions:

Right click menu at top of the individual pieces to add ‘use as filter’, or go to dashboard button at top of screen and select actions.

The sort of thing you can do is make the second worksheet on a page change when you hover over pieces of the map.

Highlighting – I dont understand highligthing after the video and need to review.

*** The diference between Action: filter and Action: highlight is that filter removes the underyling data whereas highlight functions on the complete data set.



Scenario: Table A and Table B

Inner Join: Only matching rows in both tables, the rest removed.

Left Join: All rows from Table A but only matching rows from B, null values where matching values do not exist in B. Those with no matching key excluded.

Right Join: All rows from Table B but only matching rows from A, null values where matching values do not exist in A. Those with no matching key excluded.

Outer Join: All rows, if they join they are merged else not but rows still kept.

When joining on multiple fields it may make sense to use two keys to avoid duplication.



To create a blend you need to add two data sources, but it can be two versions of the same data source.

??? Learn more here, why didnt it work for me?


Multi-axis charts:



Right click on a pill and select create, then bins to set up customized bins.



Right click in dimensions/measures section and click create parameter. After creating this you need to link it to something to make it usable…



Data Cleaning:

Sometimes the data interpreter will offer the option to activate when a new data source is added.

Review results for data interpreter
Red cells = header
Green cells = data

You can use drop down -> split to split data on a givern character or character set.



Storylines are like a slide show of different sheets and dashboards. Its almost like powerpoint, but with interactable pieces. What you are doing is telling a story with the different pieces of data. The idea is to add a description of what each ‘slide’ on the storyboard is trying to tell whoever is reading it.

Click button across the bottom ‘new story’.


sometimnes with geographic data some of the data will not correspond to a known map location and this will appear on the bottom right of the map as a map error called ‘unknown’.


Build your own custom Territory:

Approach 1 –

Use groups – Hold ctrl and select all of the states, then right click and in the menu choose group (generally all dimensions).

Then, remove the state/country/etc. The initial dimension.


Approach 2 –

Select the dimension pill desired and click the drop down, then geographic role -> create from -> variable.



Add a parameter/variable to detail and then in the pill drop down select ‘show highlighter’



For finding data that fits a group pattern, for example marketting data where the limit is all roughly the same on ROI.

Tableu uses k-means clustering by default.

Click analytics tab on left -> Cluster located under model -> Drag Cluster onto sheet.

Section 8, 61 error with data and population blah blah

Combining clustering with trend lines can be very effective to see what is and is not profitable.

For instance, if the linear investigation shows that Revenue = 0.98 * marketting_spend + 7, it would suggest that for every dollar spent on marketing you get 98 cents as a return.

Holt ctrl and drag clusters into dimensions to save them.


Cross Database Joins from Different Sources:

Tableu can join from different file types, not all but most.

Rows are a good indicator for duplication.