tsfresh cwt analysis

[mbjp09]

hi, community member

I'm new with time series analysis, tsfresh, etc., and I have hit a block with the cwt analysis in tsfresh. I have a time series of data with about 300,000 rows (i.e., time points). I can run pywt cwt on my own and compare with the output from tsfresh cwt, but the output coefficients are not the same. My own run results in also 300,000 data points. When I set the parameters for tsfresh I only used coeffs 0-14. I thought that those 0-14 coeffs from tsfresh should correspond to the first 14 coeffs from my pywt cwt run, but this isn't the case. (granted, not the same magnitude since tsfresh is using a different package for cwt. but I would have thought they'd be close)

Anyone able to provide some clarity?

I am using the same width, and I cannot provide data since it's work related.