Continuous wavelet transform for time series with cross wavelet transform, coherence, phase, and plotting.

[srbt...@gmail.com]

For a geophysical project I wanted to use wavelet transforms of time series and to do cross wavelet transforms to compare to signals. I had used the package pycwt before, but I found the options unnecessarily complicated for my application and the inner workings difficult to understand and hard to adapt to new problems. Therefore, I wrote some of my own using the pywavelets cwt functions.

For those interested, I am sharing the code I created to compute spectrograms, cross wavelet spectrograms, cross wavelet coherence, and cross-wavelet phase for time series.

You can find the jupyter notebook here containing it all, along with examples:

https://gist.github.com/aprovecharLab

https://gist.github.com/aprovecharLab/09ded80ae59ea510334e00926a15920f

Please ... If you find this useful, let me know and if you find errors or problems, please help me understand them and correct them.

Thanks,

Stephen Brown

[Gregory Lee]

Thanks Stephen!  We have had queries about doing wavelet coherence before (see https://github.com/PyWavelets/pywt/issues/370), but don't currently have a developer working on extending CWT functionality. If you are would be interested in potentially contributing this sort of thing to PyWavelets itself, please feel free to open an issue or pull request on GitHub.