Tulip allows you to analyze defect data in real time, and across any time period- daily, weekly, monthly and more.
But, if operators have been using Tulip apps for weeks or months, you will generate a large number of defect reports. You can use a "Pareto Chart" to quickly understand this data by instantly highlighting the most common defects.
This guide will show you how to create a Pareto chart within the Analytics Builder.
Before using this guide, you should understand the concept of "app completions", because that is how data will be generated from the app.
Structuring Your App
Before using the Analytics Builder, you will likely want to use a single variable to track all possible defects within the app.
This variable is commonly called "defect_name". You can use it to track the value of a dropdown in a Form Step, or use buttons to allow an operator to choose a defect from a list.
Here's an example of how to add a variable "defect_name" to a dropdown:
That is all you need within your app to create your Pareto chart.
Building the Analysis
In the app of your choice, click the "Analytics" tab on the App Summary View, then choose "New Analysis".
Then, choose the "One Operation" analysis for the list of templates.
Then, you will see 5 fields in the Context Pane:
For this analysis, we will just need to change the "X Axis" and "Filters" fields.
The "Y Axis" can stay as "Count of Completions" because we just want to find the count of each defect. Each defect can only be logged once per app completion in this model.
For the "X Axis" field, we need to choose the variable that we would like to separate out into bars for each individual value. In this case, that is "defect_name", because we want to learn more about the frequency of values.
Select "X Axis" then choose the word "change" next to the "X Axis" label.
Then, select "App variables" and choose the "defect_name" variable.
But here is the problem. A defect may only be logged 20% of the time that the app is used. In the other 80% of times that the app is used, there will not be any defects.
So, we need a way to filter out all the cases where no defect is logged. We can do this in the "Filters" field.
Select "Filters" and then choose "Add New Filter".
Then, choose the variable named "defect_name".
Finally, choose the orange "Equals" text and choose "Is Not Blank" under "Equalities". This will allow your analysis to only look at app completions where the "defect_name" variable had a value.