fix REGEX default value change in pd.str.replace

python/palin/utils/utils.py

@@ -60,8 +60,8 @@ def compute_prob_agreement(data_df, session_identifiers=['experimentor', 'type',
     trial_responses_df = trial_responses_df.join(trial_responses_df[response_identifier].str.split(expand=True).rename(columns={0: '%s1' % response_identifier, 1: '%s2' % response_identifier}))
 
     #clean the response data by removing any non-numeric characters
-    trial_responses_df['%s1' % response_identifier] = trial_responses_df['%s1' % response_identifier].str.replace(r'\D', '')
-    trial_responses_df['%s2' % response_identifier] = trial_responses_df['%s2' % response_identifier].str.replace(r'\D', '')
+    trial_responses_df['%s1' % response_identifier] = trial_responses_df['%s1' % response_identifier].str.replace(r'\D', '', regex=True)
+    trial_responses_df['%s2' % response_identifier] = trial_responses_df['%s2' % response_identifier].str.replace(r'\D', '', regex=True)
 
     # drop the original response column
     trial_responses_df = trial_responses_df.drop(columns=[response_identifier])
@@ -133,4 +133,4 @@ def simulate_observer(internal_noise_sigma,criteria, n_trials, n_blocks=1):
 	#probability of agreement between pass
 	prob_agreement = np.mean(all_response_pass1==all_response_pass2) 
 
-	return prob_agreement,prob_interval1
+	return prob_agreement,prob_interval1