mondejar / ecg-classification

Code for training and test machine learning classifiers on MIT-BIH Arrhyhtmia database
GNU General Public License v3.0
528 stars 193 forks source link

Work on all features #10

Closed banivyom closed 5 years ago

banivyom commented 5 years ago

Actually, I want to work on all 180 features rather than extracting few of them(wvlt, HOS, etc) So is there a way to do that? Thanks.

mondejar commented 5 years ago

Check the file _run_trainSVM.py. In that file, you can set in the variable compute_morph = {''} the desired computed features. Additionally, in order to use the raw signal, just set compute_morph = {'raw'}. In the file _loadMITBIH.py is considered that option. You are also free to compute your own features

banivyom commented 5 years ago

use_RR = False norm_RR = False compute_morph = {'raw'} So here I should use use_RR = False and norm_RR = False right?

mondejar commented 5 years ago

Yes, in order to just employ the raw signal is correct as you said. You can also adjust the desired windows size with winL and winR when you call the main method.

banivyom commented 5 years ago

Thanks, changed the window size to 180 each few days ago :-D