2017-sci-hub-downloads_per_month.tsv

Generated by:

from collections import defaultdict
dl_per_month = defaultdict(int)
days_per_month = defaultdict(dict)
for line in open('2017.statistics.tab','r'):
    la = line.strip().split("\t")
    date = la[0].split(' ')[0].split('-')
    dl_per_month[date[1]] += 1
    if date[2] not in days_per_month[date[1]].keys():
        days_per_month[date[1]][date[2]] = True
print("month\tdownloads\tdays_with_data")
for key in dl_per_month.keys():
    downloads = dl_per_month[key]
    days_in_month = len(days_per_month[key])
    print("{}\t{}\t{}".format(key,downloads,days_in_month))

@gedankenstuecke nice. It would also be good to have visitors in addition to downloads. I suspect visitors (unique IP addresses) will be less affected by automated downloads. As per our conversation here regarding the lack of column labels, how confident are you that the third column is an IP address code? If we're not confident, we should avoid reporting visitors and stick with downloads.

Thanks @gedankenstuecke for putting this together! It helped inspire this notebook where I calculate these numbers for the old and new logs.