/
revision_stats.py
166 lines (123 loc) · 4.36 KB
/
revision_stats.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
# -*- coding: utf-8 -*-
# <nbformat>3</nbformat>
# <codecell>
import requests, json
from pprint import pprint
article_title = 'Coffee'
rev_json = requests.get('http://ortelius.toolserver.org:8089/revisions/'+article_title)
revs = json.loads(rev_json.text)['result']
# <codecell>
print sorted(revs[0].keys())
# <codecell>
import numpy
import matplotlib.pyplot
from scipy import stats
import json
# <codecell>
sorted(revs, key=lambda x: int(x['rev_id']), reverse=True)[0]
DATE_FORMAT = '%Y%m%d%H%M%S'
from datetime import timedelta, datetime, date
from collections import defaultdict
def parse_date(d_or_dlist):
try:
return datetime.strptime(d_or_dlist, DATE_FORMAT)
except TypeError:
try:
return [datetime.strptime(d, DATE_FORMAT) for d in d_or_dlist]
except TypeError:
raise TypeError('parse_date() expects a date string or iterable of date strings')
def get_time_diffs(times):
ret = []
dtimes = []
tds = []
tds_seconds = []
for t in times:
dtimes.append(parse_date(t))
for x, y in zip(dtimes, dtimes[1:]):
tds.append(y - x)
tds_seconds.append(tds[-1].total_seconds())
return tds_seconds
DEFAULT_CUTOFF = 2010
def edits_by_day(edits, cutoff=DEFAULT_CUTOFF):
ed_dict = defaultdict(list)
for ed in edits:
dt = parse_date(ed['rev_timestamp'])
ed_dict[dt.utctimetuple()[:3]].append(ed)
return [(date(*dtup), len(eds)) for (dtup,eds) in ed_dict.iteritems()
if dtup[0] >= cutoff]
# <codecell>
def get_reverted_by_summary(revs):
reverted_revs = []
for revnum,rev in enumerate(revs):
if 'revert' in rev['rev_comment'].lower():
rev_range_end = revnum
for i in range(revnum - 1, revnum - 5, -1):
if rev['rev_sha1'] == revs[i]['rev_sha1']:
rev_range_start = i
reverted_revs.append((rev_range_start, rev_range_end, rev_range_end - rev_range_start))
break
return reverted_revs
rev_by_sum_range = get_reverted_by_summary(revs)
# <codecell>
#plot([x for x in range(0, len(rev_by_sum_range))], [y[2] for y in rev_by_sum_range])
#rev_by_sum_range[0]
max([y[2] for y in rev_by_sum_range])
# <codecell>
REVERT_LOOKAHEAD = 1
REVERT_THRESHOLD = 0
from __builtin__ import sum, any
def get_diff_sizes(revs, absolute=False):
if absolute:
return [(b['rev_user_text'], abs(b['rev_len']-a['rev_len']))
for a,b in
zip(revs, revs[1:])]
else:
return [(b['rev_user_text'], b['rev_len']-a['rev_len'])
for a,b in
zip(revs, revs[1:])]
def get_editor_bytes(revs):
editors = {}
reverted_revs = []
diff_sizes = get_diff_sizes(revs, absolute=True)
for editor, size in diff_sizes:
tot_size, count = editors.get(editor, (0,0))
editors[editor] = ((tot_size + size), (count + 1))
return editors
def get_reverted_revs(revs):
reverted_revs = []
diff_sizes = get_diff_sizes(revs)
for rev_num, (editor, size) in enumerate(diff_sizes):
rev_range = [x[1] for x in diff_sizes[rev_num:rev_num+REVERT_LOOKAHEAD+1]]
if any([ r != 0 for r in rev_range ]) and sum(rev_range) == REVERT_THRESHOLD:
reverted_revs.append(rev_num)
return reverted_revs
print 'User, (bytes, edits)'
sorted(get_editor_bytes(revs).items(), key=lambda x: x[1], reverse=True)
reverted_revs = get_reverted_revs(revs)
len(reverted_revs)
# <codecell>
revs[9:12]
# <codecell>
ed_data_unsorted = edits_by_day(revs, 2005)
ed_data = sorted(ed_data_unsorted, key=lambda x: x[1], reverse=True)
ed_counts = [e[1] for e in ed_data]
def print_stats(datums):
print 'Mean:', stats.tmean(datums)
print 'Median:', stats.cmedian(datums)
print 'Std Dev:', stats.tstd(datums)
print 'Variation:', stats.variation(datums)
print 'Kurtosis:', stats.kurtosis(datums, fisher=False)
print 'Skewness:', stats.skew(datums)
print_stats(ed_counts)
plot(ed_counts)
# <codecell>
#sorted_revs = sorted(revs, key=lambda x: x['rev_len'], reverse=True)
data_unsorted = get_time_diffs([x['rev_timestamp'] for x in revs])
data = sorted(data, reverse=True)
print stats.kurtosis(data, fisher=False)
print stats.skew(data)
print stats.describe(data)
plot(data_unsorted)
# <codecell>
plot(data)
# <codecell>