munger.py 文件源码

def get_dataframe_list(args, data_fields=('gene', 'raw_counts')):

    # get a list of dataframes
    dfs, files = [], args['files'] or []
    # create an index using the filenames
    # this will prevent having an overlong command line for 100's or 1000's of files
    if args['file_index']:
        with open(args['file_index']) as fp:
            files.extend(fp.readlines())
    files = sorted(filter(None, set([f.strip() for f in files])))
    # now iterate over the files and get the looooong list of dataframes
    for f in files:
        # Get only specific columns with usecols
        df = pd.read_table(f, usecols=data_fields)
        dfs.append(df)
    return dfs, files # a list of dataframes and the files index