I have a dataset that looks like this

This is considered a “bad” dataset because of the periodic behavior happening in the trough. a “good” data set would be flat or have clear changes without periodic behavior. How would I go about detecting that wavelike/periodic behavior, and quantifying it?

I’ve looked into fourier transforms but that appears to be more focused on digital signal processing (which this is not). I’ve found the scipy module,but haven’t found what I’m looking for quite yet.
Any recommendations?