Repository navigation
Algorithm for inverse continuous wavelet transform #751
Description
Activity
Hi there, I am not sure if I can help directly but I found a few resources that might be of help:
StackExchange posts
- https://dsp.stackexchange.com/questions/87097/why-is-inverse-cwt-inexact-inaccurate
- https://stackoverflow.com/questions/21412465/inverse-wavelet-transform-xpost-signalprocessing
Other projects
- https://github.com/regeirk/pycwt (seems to be maintained, might not be fully correct without modifications, though: inverse cwt regeirk/pycwt#30
- https://github.com/duducosmos/kPyWavelet (old and unmaintained, likely broken because it uses
easy_install)
Reacted by clementegiselavThank you @agriyakhetarpal ! I was reading the documentation and contributions that you suggest, but they are the same ones that I had seen a while ago. I see that an algorithm for the icwt has not yet been implemented for wavelets other than Morlet, Paul, DOG and Mexican Hat, that is, those used by Torrence and Compo.
Reacted by Agriya KhetarpalReacted by GMDragon
I would like to take this opportunity to ask you if we have any inverse continuous wavelet transform algorithm that allows us to reconstruct the signal with wavelets, such as daubechies. I did a fairly exhaustive search and this leads me to the icwt algorithm based on the paper by Torrence and compo (1998), which is not implemented for the wavelets we are talking about. Thank you very much for your contributions!