Events files are read in with onset as the index, causing downstream issues

[tsalo]

In a few places within the package, events files are read in using pandas.read_csv, as below:

IDConn/idconn/connectivity/build_networks.py

Line 86 in 5e9bc32

timing = pd.read_csv(event_files[0], header=0, index_col=0, sep='\t')

Setting index_col=0 means that the first column will be read in as the index, and in BIDS datasets that first column would generally be the "onset" column. Later, however, the "onset" column is directly referenced, but it won't be available since it'll be the index (see below).

IDConn/idconn/connectivity/build_networks.py

Line 132 in 5e9bc32

blocks.append((cond_timing.loc[i]['onset'] / tr, ((cond_timing.loc[i]['onset'] + cond_timing.loc[i]['duration']) / tr) + 1))

I think the solution is to drop index_col=0 and header=0 from all read_csv calls targeting events files. Also, switching to read_table will let you drop sep="\t", so it would just be df = pd.read_table(events_file).

[62442katieb]

Good find! I'll go ahead and edit it. read_table will be much more efficient. Thanks!