Sanitize NAN values from timeseries data

[aestoltm]

Description

Add check to sanitize job timeseries data of NAN values due to Symfony throwing double NAN error when json encoding. Added helper function which replaces any NAN values in the series data with null.

Motivation and Context

Charts with NAN data would return a null record.

Tests performed

Observed changes on xdmod-dev. Job that has a NaN value(s) can be seen here .

After checking mongodb-access, it was confirmed the value is NaN inside Mongodb for the job linked above. There was an error for the cpuuser value that was NaN stating that it couldn't collect enough data for the specified time period.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My code follows the code style of this project as found in the CONTRIBUTING document.
• I have added tests to cover my changes.
• All new and existing tests passed.

[aaronweeden]

Can you please explain an example of how to reproduce the problem in the "Motivation and Context" section and explain in more detail how you tested it on xdmod-dev in the "Tests performed" section?

Without being sure how to reproduce the error, I'm not sure yet if changing NaN to zero is the right solution.

[jpwhite4]

Having thought about this - I think that the NaN values should be removed from the data series rather than set to 0 - since 0 could be a valid value from a valid measurement and NaN typically indicates that there was not a valid measurement.

[aaronweeden]

Looks like the PR description should be updated to say it replaces them with null instead of 0.

[aestoltm]

Looks like the PR description should be updated to say it replaces them with null instead of 0.

Updated the description that the value is replaced with null