endolith

That's removing the DC component https://github.com/endolith/waveform_analysis/blob/master/waveform_analysis/thd.py#L61

@spidermomo Is there anything else you need from me or can I close this?

because I'm lazy busy, and my computer has plenty of memory to fit any files I want to measure? of course it would be more efficient to read the file...

Somewhat working! Now: - [x] Get rid of test warnings - [x] Integrate with Codecov - [ ] Achieve 100% coverage

Not really, no. If the chirp is slow enough, you could run it through the THD function in small chunks, though.

@abrahamcn Is there anything you need from me on this or can I close it? I think you would need a different algorithm to process actual sweep signals. Stepped sweep...

I'm not sure if the frequency estimation function should handle this or not. If it should, it should probably return `None`, but maybe it should be the calling function's responsibility...

> basinh is dominated by all other solvers. Makes me wonder whether we want to deprecate that or declare it legacy. GitHub search says it's used in a lot of...

@tirthasheshpatel I don't know about the nan policy so you can do it in another PR if you want

Copying the ambiguous part from the other issue: > Hmm... Going back to [2004](https://github.com/matplotlib/matplotlib/blob/e34a333d00814124d3e19d462b9d78ac35e7a49a/lib/matplotlib/mlab.py#L1225) the code always had `abs()` in it, which I assume is for complex numbers. > >...