## Introduction to Pareto Chart

Pareto charts highlight the most important among a (typically large) set of factors by combining both bars and line graph, with individual values represented by bars in descending order, and the cumulative total represented by the line. They are useful when analyzing data where many courses of action are competing for attention, for instance indicating the frequency or causes of problems, as well as their cumulative impact. Pareto charts help to find the problems to prioritize in order to observe the greatest overall improvement.

Learn more about visualizations and dashboards with Spotfire®

## Creating a Pareto Chart in Spotfire®

This configuration works with all versions of Spotfire. This tutorial uses the "Sales and Marketing" sample analysis available in the Spotfire library (Library > Samples > Sales and Marketing). Open the "Sales & Marketing" sample analysis and create a new blank page.

The Pareto chart is created with a Combination Chart. Open the visualization panel and add a Combination Chart to the canvas.

Select a category column in the X-axis and select 2 columns on the Y-axis. In this example we use the (State) column as X-axis and (Class Sales) as Y-axis twice.

Change the aggregation type for one Y-axis to "Cumulative Sum".

Sort X-axis by the calculation column to re-order all bars including the cumulative column. Open Combination Chart properties > Appearance and set Sort x-axis by: to "Sum(Class Sales)" in this example.

Make sure the X-axis only uses the currently filtered data instead of showing all data by default which may cause the line chart to display incorrectly when filtering data. Open Combination Chart properties > X-axis > Settings... and set Evaluate axis expression on: to "Current filtering only".

Now, in order to configure the Combination Chart as a Pareto Chart, the Y-axis cumulative calculation must be modified. Right-click on its Y-axis selector and select "Custom Expression..."

Modify the expression to use the "AllNext" function instead of the default "AllPrevious" function. In this example, the expression should be as follows: `Sum([Class Sales]), Sum([Class Sales]) THEN Sum([Value]) OVER (AllNext([Axis.X]))`

Clicking OK should result with the following visualization:

You will notice that the cumulative calculation still does not look as expected. This is because the computation is still based on the original order of the (State) column. To re-order the original sequence by including the height of the bars we need to modify the X-axis expression. Right-click on the X-axis selector and select "Custom Expression"...

Modify the expression to use the height of the bars. In this example, the expression should be as follows: `<Sum([Class Sales])over([state]) NEST [state]>`

Clicking OK should result with the following visualization:

Now we can change the cumulative bards to show as a line to get closer to a Pareto chart. From the Combination Chart legend, click the bars icon for the cumulative calculation Y-axis and set it to "Show as lines".

## Additional Configurations

### Allowing each Y-axis to have its own scale

Right-click the Y-axis scale and set it to "Multiple Scales"

### Turning cumulative calculation into percentage

Right-click on the Y-axis selector for the cumulative calculation and select "Custom Expression...". Modify the expression as follows:

`Sum([Class Sales]) as [sales], Sum([Class Sales]) THEN Sum([Value]) OVER (AllNext([Axis.X])) / (Sum([Value]) OVER (All([Axis.X]))) as [Cum%]`

Clicking OK should result with the following visualization:

Display the (Cum%) as percentage. Open Combination Chart properties > Formatting and set Y:Cum% to percentage.

Set the Y-axis for cumulative calculation to "Show as Iines" again.

### Changing labels orientation

Changing the X-axis labels orientation will make them readable. Right-click the X-axis scale and set it to "Vertical Labels".

The result should be a Pareto chart that looks like this:

## Recommended Comments

There are no comments to display.