plot.py

"""
zcat -f access* | grep userfiles | grep GET | awk -F' ' '{if ($6 == "\"GET" && $9 == "200") print $10;}'  | sort | uniq -c | sort -rn > histo_dump
"""

import matplotlib.pyplot as pl
import numpy as np
import csv

data = []

with open('histo_dump', newline='') as csvfile:
	reader = csv.reader(csvfile, delimiter=' ', skipinitialspace=True)
	for row in reader:
		mb = float(row[1]) / 1048576.0 / 1024.0
		total = float(row[0]) * mb
		data.append((total, mb))


data.sort(key=lambda tup: tup[0], reverse=True)

totalstor = sum(b for a, b in data)

step = 0.01
acc = 0.0
next = step
cbin = 0.0
bins = []
labels = []
for entry in data:
	acc += entry[1]
	cbin += entry[0]
	while acc / totalstor > next:
		bins.append(cbin)
		labels.append(next * totalstor)
		print("after accounting for " + str(round(next * totalstor, 1)) + " GiB of stored data, " + str(round(sum(bins))) + " GiB of bandwidth was used")
		next += step
		cbin = 0.0

if cbin != 0:
	bins.append(cbin)
	labels.append(next * totalstor)

acc = np.add.accumulate(bins)

pl.bar(labels, acc, totalstor * step * 0.8)
pl.xlabel('Stored files sorted by bandwidth usage [GiB]')
pl.ylabel('Cumulative yearly bandwidth used [GiB]')
pl.show()