# Converting code from old CWT to new CWT

5 views (last 30 days)
Michael on 20 Mar 2020
Commented: Michael on 1 May 2020
I am trying to understand and re-implement an algorithm using someone else's code, which uses the old cwt function. Here is the portion of code in question:
fb = 13.5;
fc = 0.5;
scale = 3.7;
output = cwt(signal,scale,['cmor' num2str(fb) '-' num2str(fc)]);
The variable signal is time series data at 100Hz. The goal of this is to return the scale parameters of a bandpassed version of the signal at approximately 10-16Hz.
Here are my questions:
1. Is this actually getting the proper bandpass range? Presumably this is using scale parameter. However, Fb and Fc seem mislabeled in this code and seem redudant to the scale parameter.
2. How can I properly implement this using the new CWT functions? That is, given a fixed center frequency, bandwidth, and sampling frequency what is the best way to implement the appropriate single-wavelet transformation using a complex morlet wavelet using the new framework?
Thanks!

Prabhan Purwar on 28 Mar 2020
Edited: Prabhan Purwar on 28 Mar 2020
Hi,
Variables fb and fc represents properties of the Morse Wavelet used to determine the Continuous Wavelet Transform (CWT) of the signal.
Following link illustrates the use of filterbank in cwt
To verify your results take help of Magnitude Scalogram (as illustrated in the above example)
For further insights refer to the following links:

#### 1 Comment

Michael on 1 May 2020
Thank you.
If I just wanted to have the wavelet coefficient values for a single wavelet with a center frequency of Fc and a bandwidth of Fb at sampling frequency Fs, what would be the most direct way to go about it using the updated functions?
I cannot see how to directly do this with these functions, which seem more designed ot give you the scaleogram or filtered estimate from multiple wavelets, rather than the coefficients for a single wavelet of specific properties.

Translated by