1

Using the wmtsa R package, one can perform a continuous wavelet transform (CWT) on a vector of data. Here's an example:

library(wmtsa)
x=rnorm(8)
cwt=wavCWT(x=x,wavelet="gaussian2")

Where the gaussian2 wavelet refers to the mexican hat wavelet. Through the inverse CWT one should be able to retrieve the exact original data object.

However, I cannot find a function that's able to do so.

Does anyone have any suggestion for a function or how to implement the inverse CWT?

KoenVdB
  • 293
  • 2
  • 12
  • Does it have to be a CWT or could a DWT be suitable, too? – RHertel Aug 13 '15 at 08:55
  • @RHertel It has to be a CWT. The inverse DWT is available in the `wavethresh` package, amongst others. – KoenVdB Aug 13 '15 at 09:08
  • The `reconstruct()` function of the `wmtsa` package doesn't seem to work for CWT. I looked also at the `Rwave` package, but couldn't find anything that seemed to answer your question. Maybe [this answer](http://stackoverflow.com/a/22131410/4770166) helps. – RHertel Aug 13 '15 at 09:14

0 Answers0