#RR_sec=dc.detrend_data(dc.quotient_filt(dc.square_filt(RR_sec_unclean))) #RR_sec=dc.detrend_data((RR_sec_unclean) RR_sec = dc.quotient_filt(dc.square_filt(RR_sec_unclean)) #RR_sec=dc.square_filt(RR_sec_unclean) #RR_sec=RR_sec_unclean ##### Extract delta RR intervals here ####### delta_RR_sec = pr.get_delta_rr(RR_sec) ###### Calculating GOLBAL statistical features for RR_sec VALUES ################# ###calculating AVG/mean of RR intervals ### mean_global = np.mean(RR_sec) * 1000 mean_global_arr.append(mean_global) #do we need this? feature_rec.append(mean_global) global_vocab, index_of_features = cl.fill_global_vocab( "mean_global", index_of_features, global_vocab) # if "mean_global" not in global_vocab.keys(): # global_vocab["mean_global"]=index_of_features; # index_of_features=index_of_features+1; #print("mean_global is: " + str(mean_global)) #sdrr sdrr_raw = np.std(RR_sec) #print (" sdrr of RR_sec (raw) in sec is: " +str(sdrr_raw)); sdrr_raw_ms = sdrr_raw * 1000 std_dev_global = sdrr_raw_ms feature_rec.append(std_dev_global) global_vocab, index_of_features = cl.fill_global_vocab( "std_dev_global", index_of_features, global_vocab)
#RR_sec=dc.detrend_data(dc.quotient_filt(dc.square_filt(RR_sec_unclean))) #RR_sec=dc.detrend_data((RR_sec_unclean) #RR_sec=dc.quotient_filt(dc.square_filt(RR_sec_unclean)) RR_sec = dc.square_filt(RR_sec_unclean) #RR_sec=RR_sec_unclean ##### Extract delta RR intervals here ####### delta_RR_sec = pr.get_delta_rr(RR_sec) ###### Calculating GOLBAL statistical features for RR_sec VALUES ################# ###calculating AVG/mean of RR intervals ### mean_global = np.mean(RR_sec) * 1000 mean_global_arr.append(mean_global) #do we need this? feature_rec.append(mean_global) global_vocab, index_of_features = cl.fill_global_vocab( "mean_global", index_of_features, global_vocab) # if "mean_global" not in global_vocab.keys(): # global_vocab["mean_global"]=index_of_features; # index_of_features=index_of_features+1; #print("mean_global is: " + str(mean_global)) #sdrr sdrr_raw = np.std(RR_sec) #print (" sdrr of RR_sec (raw) in sec is: " +str(sdrr_raw)); sdrr_raw_ms = sdrr_raw * 1000 std_dev_global = sdrr_raw_ms feature_rec.append(std_dev_global) global_vocab, index_of_features = cl.fill_global_vocab( "std_dev_global", index_of_features, global_vocab)
# ###### Calculating GOLBAL statistical features for RR_sec VALUES ################# # if "n" in record: # ####plot sodp and save fig #### # fig_std,plot_std_p=graphs.plot_simple(rec_name,range(len(std_dev_window_arr)), std_dev_window_arr, "Window number", "Standard Deviation", "SDW Series ", 'b',xlim_lo=0, xlim_hi=1, ylim_lo=0, ylim_hi=0.2) # fig_std.savefig(output_folder+"std_dev_window_"+record+".pdf",format='pdf') # if "p" in record: # ####plot sodp and save fig #### # fig_std,plot_std_p=graphs.plot_simple(rec_name,range(len(std_dev_window_arr)), std_dev_window_arr, "Window number", "Standard Deviation", "SDW Series ", 'r',xlim_lo=0, xlim_hi=1, ylim_lo=0, ylim_hi=0.2) # fig_std.savefig(output_folder+"std_dev_window_"+record+".pdf",format='pdf') ###calculating AVG/mean of RR intervals ### mean_global=np.mean(RR_sec)*1000; mean_global_arr.append(mean_global) #do we need this? feature_rec.append(mean_global) global_vocab,index_of_features=cl.fill_global_vocab("mean_global", index_of_features, global_vocab) # if "mean_global" not in global_vocab.keys(): # global_vocab["mean_global"]=index_of_features; # index_of_features=index_of_features+1; #print("mean_global is: " + str(mean_global)) #sdrr sdrr_raw=np.std(RR_sec); #print (" sdrr of RR_sec (raw) in sec is: " +str(sdrr_raw)); sdrr_raw_ms=sdrr_raw*1000; std_dev_global=sdrr_raw_ms; feature_rec.append(std_dev_global)
#print arr #calc stats for 1 list and make 4 features #print ('index_num is: ' + str(index_num)) size_val,min_max,mean_val,var_val,skewness_val,kurtosis_val=describe(arr) min_val=min_max[0] max_val=min_max[1] feature_rec.append(size_val) feature_rec.append(min_val) feature_rec.append(max_val) feature_rec.append(mean_val) feature_rec.append(var_val) feature_rec.append(skewness_val) feature_rec.append(kurtosis_val) for f_name in stat_feature_names: global_vocab_ecg,index_of_features_ecg=cl.fill_global_vocab(f_name+'_'+all_feature_names[index_num], index_of_features_ecg, global_vocab_ecg) index_num=index_num+1; all_features.append(feature_rec) print(' dict is: ' + str((global_vocab_ecg))) print("all features is: ") print all_features print(len(all_features)) #write to all_features to file rw.write_value(all_features,output_folder,"all_features_readable.txt",'w') rw.write_features_to_file(all_features,output_folder,"all_features_pickle.txt") rw.write_features_to_file(global_vocab_ecg,output_folder,"global_vocab_pickle.txt") rw.write_features_to_file(rec_name_array, output_folder, "rec_name_array_pickle.txt") exit()
#print ('index_num is: ' + str(index_num)) size_val, min_max, mean_val, var_val, skewness_val, kurtosis_val = describe( arr) min_val = min_max[0] max_val = min_max[1] feature_rec.append(size_val) feature_rec.append(min_val) feature_rec.append(max_val) feature_rec.append(mean_val) feature_rec.append(var_val) feature_rec.append(skewness_val) feature_rec.append(kurtosis_val) for f_name in stat_feature_names: global_vocab_ecg, index_of_features_ecg = cl.fill_global_vocab( f_name + '_' + all_feature_names[index_num], index_of_features_ecg, global_vocab_ecg) index_num = index_num + 1 all_features.append(feature_rec) print(' dict is: ' + str((global_vocab_ecg))) print("all features is: ") print all_features print(len(all_features)) #write to all_features to file rw.write_value(all_features, output_folder, "all_features_readable.txt", 'w') rw.write_features_to_file(all_features, output_folder, "all_features_pickle.txt") rw.write_features_to_file(global_vocab_ecg, output_folder, "global_vocab_pickle.txt")
# # #### APPLY FILTERS TO CLEAN DATA ####### # # # # RR_sec=dc.detrend_data(dc.quotient_filt(dc.square_filt(RR_sec_unclean))) # # RR_sec=dc.quotient_filt(dc.square_filt(RR_sec_unclean)) # # RR_sec=RR_sec_unclean # # #### Extract delta RR intervals here ####### # # delta_RR_sec = pr.get_delta_rr(RR_sec); ###### Calculating GOLBAL statistical features for RR_sec VALUES ################# ###calculating AVG/mean of RR intervals ### mean_global=np.mean(RR_sec)*1000; mean_global_arr.append(mean_global) #do we need this? feature_rec.append(mean_global) global_vocab,index_of_features=cl.fill_global_vocab("mean_global", index_of_features, global_vocab) # if "mean_global" not in global_vocab.keys(): # global_vocab["mean_global"]=index_of_features; # index_of_features=index_of_features+1; #print("mean_global is: " + str(mean_global)) #sdrr sdrr_raw=np.std(RR_sec); #print (" sdrr of RR_sec (raw) in sec is: " +str(sdrr_raw)); sdrr_raw_ms=sdrr_raw*1000; std_dev_global=sdrr_raw_ms; feature_rec.append(std_dev_global)