Join multiple pandas dataframes

join.py

import glob
import sys
import pandas


def read_filenames(names):
    for arg in names:
        if "*" in arg:
            for file in glob.glob(arg):
                yield file
        else:
            yield arg


data = []

for sample in read_filenames(sys.argv[1:]):
    frame = pandas.read_csv(
        sample,
        sep="\t",
        header=None,
        names=["chr", "strand", "coord", sample],
        index_col=[0, 1, 2]
    )
    data.append(frame)

data_joined = data[0].join(data[1:], how='outer')

sum = data_joined[data_joined.columns].sum(axis=1)
data_joined.insert(0, "scoreSum", sum)

print(data_joined.to_csv(
    sep="\t",
    na_rep=0,
    float_format='%.0f'
    ))

Answer on biostars question: https://www.biostars.org/p/317165