Peak detection fails for high DC levels

[josueportiz]

In fit_peaks(), the moving average values to test list (ma_perc_list) starts at 5 %. Signals with high DC values related to the AC component (e.g. some PPG signals) add a high value to test so that the starting 5% leads to moving average values higher than any of the peaks of the signal.

Please, consider add fine grained values to the ma_perc_list: 0, 1, 2, 3, 4 %. Otherwise it does not work.

[paulvangentcom]

Thanks for the suggestion, I will put it on the list for the next version. Currently picking up heartpy again.

[SahebehDadboud]

@paulvangentcom that's one of the reasons I can't use Hearpy package as a built-in module, so I had to clone the project. If you add ma_perc_list as a parameter to the function (which I did already), it would be much more beneficial. Currently I had to divide the original ma_perc_list by 10 to let rolling mean find the peaks in our device(s).