CODEWITHSUNDEEP
pythonpandasfilterscipysignal-processing
Mari
Mari

Reputation: 698

Remove unwanted portion from a signal in python

I have a signal which measures current and save it the form of 1D array. The signal follows a typical patters and I will be doing some research work on that pattern.

But before the pattern appears, there lies some noise or unwanted signal. I want to filter or neglect that unwanted portion from my signal.

Refer picture below for better understanding

Also given some sample data to play with. (2 columns included - index & current values )

I tried with many approaches like sliding_window/finding_peaks from scipy. but nothing turns out.

Thanks

Upvotes: 1

Views: 968

Answers (1)

Joe
Joe

Reputation: 7121

You need to identify what is unique in that part of the signal in all your samples.

Do you want to remove

  • 10 seconds
  • the second valley below a certain value
  • two little valleys below a certain threshold

Or if the part you want to remove looks exactly or is very simliar all the time you can use cross-correlation to find out where that part is. You can use the scipy.signal.correlate to calculate it.

Upvotes: 1

Related Questions