What is the best way to represent historical data?

Question

I have a project with multiple test cases. Each of these test cases contain data related to their failure and success for each run as well as the time taken for the API calls during each execution.

I want to represent this data for each test case in different graphs, but I'm not sure if that is the best solution. What is the best way to represent this data?

Answer 1

My take on this problem would be:

• Provide a list to select one/multiple test cases.
• Provide "data-info" filters: what do you want to see mainly in the graph (in order to avoid an unreadable graph with thousands of lines)
• Provide a date filter: you rarely want to views years and years of data

Then, display:

• A graph with the "data-info" for all the test cases: and also being able to see easily the values in the graph (like Google Charts API graphs)
• Maybe some short insight about the data: a short text to resume the data
• A sortable-table (click to the legend name to sort) to view all the data individually in order to be able to review everything one by one

So here is a quick mock-up of this:

Quick notes:

• Save the settings parameters so the user doesn't have to re-select everything
• Allow to sort (and maybe search) in the table for easy reviewing
• Maybe add an indicator next to a test case if it reaches critical errors (for example, a fail/success rate too bad) to quickly show that this test has some issues

In conclusion, this kind of UI allows

• Quick and easy review of data
• Quick and easy customization to review precise data cases
• Quick overview thanks to the graph
• Detailed overview thanks to the table

So that should provide a rather nice user experience, I hope. :)

You can also check Google Public Data Explorer that provides similar "Choose what data you want and review"-UI.