.gitmodules

@@ -17,3 +17,7 @@
 	path = src/speech
 	url = osf://f5j3e
 	datalad-id = cacc9298-faed-11e9-9b76-e86a64c80559
+[submodule "src/confounds"]
+	path = src/confounds
+	url = https://github.com/psychoinformatics-de/studyforrest-data-confoundsannotation
+	datalad-id = 5f4c2889-56f4-4ef2-81c7-ddf31c0059f7

code/importer/.confounds.py.swp

@@ -0,0 +1 @@
+../../.git/annex/objects/P4/67/MD5E-s16384--e6b8bf0b4df9d58655a8377fb4d7ef28.py.swp/MD5E-s16384--e6b8bf0b4df9d58655a8377fb4d7ef28.py.swp

code/importer/confounds.py

@@ -0,0 +1,128 @@
+#!/usr/bin/env python3
+'''merges and copies (already segmented) annotations of confounds
+'''
+
+
+from glob import glob
+import argparse
+import numpy as np
+import os
+import pandas as pd
+import re
+
+
+# constants
+# no. of segments
+SEGMENTS = range(0, 8)
+# currently relevant annotations
+VIS_CONFS = ['brmean', 'brlr', 'brud', 'phash', 'normdiff']
+AUD_CONFS = ['rms', 'lrdiff']
+
+def parse_arguments():
+    '''
+    '''
+    parser = argparse.ArgumentParser(
+        description='merges and copies annotations of confounds'
+    )
+
+    parser.add_argument('-i',
+                        help='input directory where files are located',
+                        default='src/confounds/annotation/audio/'
+                        )
+
+    parser.add_argument('-s',
+                        help='the annotated stimulus' +
+                        '(\'aomovie\' or \'avmovie\')',
+                        default='aomovie'
+                        )
+
+    parser.add_argument('-o',
+                        default='test',
+                        help='output directory')
+
+    args = parser.parse_args()
+
+    in_dir = args.i
+    stimulus = args.s
+    out_dir = args.o
+
+    return in_dir, stimulus, out_dir
+
+
+def find_files(pattern):
+    '''
+    '''
+    found_files = glob(pattern)
+
+    return sorted(found_files)
+
+def create_merged_datafr(inputs):
+    '''
+    '''
+    # initial dataframe providing 'duration', too
+    df = pd.read_csv(inputs[0],
+                     header=0,
+                     sep='\t',
+                     index_col=0,
+                     dtype='str'  # read all input as string
+                     )
+
+    # populate the dataframe with the data from remaining files
+    for input_f in inputs[1:]:
+        df_new = pd.read_csv(input_f,
+                             usecols=[0, 2],
+                             index_col=0, # well, thats hard coded
+                             header = 0,
+                             sep='\t',
+                             dtype='str'  # read all input as string
+                             )
+
+        df[df_new.columns[0]] = df_new
+
+    return df
+
+
+
+if __name__ == '__main__':
+    # get command line arguments
+    in_dir, stimulus, out = parse_arguments()
+
+    out_path = os.path.join(out, stimulus)
+    os.makedirs(out_path, exist_ok=True)
+
+    if stimulus == 'aomovie':
+        file_pattern = 'fg_ad_seg?_*.tsv'
+#        in_fpathes = find_files(os.path.join(in_dir, file_pattern))
+    elif stimulus == 'avmovie':
+        file_pattern = 'fg_av_ger_seg?_*.tsv'
+#        in_fpathes = find_files(os.path.join(in_dir, file_pattern))
+    else:
+        raise ValueError('stimulus must be \'aomovie\' or \'avmovie\'')
+
+    # segments = [re.search(r'seg\d{1}', in_fpath) for in_fpath in in_fpathes]
+    # segments = sorted(list(set([segment.group() for segment in segments])))
+
+    for segment in SEGMENTS:
+        run = f'run-{segment + 1}'
+        # subtitute for current segment
+        inputs = os.path.join(in_dir, file_pattern)
+        inputs = inputs.replace('seg?', f'seg{str(segment)}')
+
+        if 'visual' in in_dir:
+            inputs = [inputs.replace('_*.tsv', f'_{conf}.tsv') for conf in VIS_CONFS]
+            out_file = 'conf_visual_' + run + '_events.tsv'
+
+        elif 'audio' in in_dir:
+            inputs = [inputs.replace('_*.tsv', f'_{conf}.tsv') for conf in AUD_CONFS]
+            out_file = 'conf_audio_' + run + '_events.tsv'
+
+        # read and merge the inputs
+        merged_df = create_merged_datafr(inputs)
+
+        # prepare saving
+        out_fpath = os.path.join(out_path, out_file)
+        # format the index in such a way
+        # that is written to file nicely formatted
+        merged_df.reset_index(level=0, inplace=True)
+        merged_df['onset'] = merged_df['onset'].map('{:.2f}'.format)
+        merged_df.to_csv(out_fpath, sep='\t', index=False)

src/confounds

@@ -0,0 +1 @@
+Subproject commit f7d47364b54bcbfff7927f0b6aa0bab4d5b11020