-
Notifications
You must be signed in to change notification settings - Fork 0
/
mining.py
750 lines (635 loc) · 21.2 KB
/
mining.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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
import cPickle as pickle
from operator import attrgetter
import time
import datetime
import functools
import copy
import sqlite3
from dateutil.relativedelta import relativedelta
import scb.names
import settings
class memoized(object):
"""Decorator that caches a function's return value each time it is called.
If called later with the same arguments, the cached value is returned, and
not re-evaluated.
"""
def __init__(self, func):
self.func = func
self.cache = {}
def __call__(self, *args):
try:
return self.cache[args]
except KeyError:
value = self.func(*args)
self.cache[args] = value
return value
except TypeError:
# uncachable -- for instance, passing a list as an argument.
# Better to not cache than to blow up entirely.
return self.func(*args)
def __repr__(self):
"""Return the function's docstring."""
return self.func.__doc__
def __get__(self, obj, objtype):
"""Support instance methods."""
return functools.partial(self.__call__, obj)
def db_type(k):
return {
int: 'INTEGER',
float: 'FLOAT',
}.get(k)
WORKER = 'worker'
ONCALL = 'oncall'
SIGNUP = 'signup'
BUYER = 'buyer'
CUSTOMER = 'customer'
ALL = 'all'
MALE = 'male'
FEMALE = 'female'
UNKNOWN = 'unknown'
GENDERS = [MALE, FEMALE, UNKNOWN]
CLASSES = [WORKER, ONCALL, SIGNUP, BUYER, CUSTOMER] + GENDERS
CUP_LITERS = 0.237
@memoized
def adapt_datetype(ts):
return int(time.mktime(ts.timetuple()))
@memoized
def convert_datetype(klass, s):
return klass.fromtimestamp(float(s))
for klass, prefix in [
(datetime.datetime, 'datetime'),
(datetime.date, 'date')]:
convert = functools.partial(convert_datetype, klass)
sqlite3.register_adapter(klass, adapt_datetype)
sqlite3.register_converter('py%s' % prefix, convert)
conn = sqlite3.connect('mining.sqlite', detect_types=sqlite3.PARSE_DECLTYPES)
conn.row_factory = sqlite3.Row
def get_cursor():
return conn.cursor()
c = get_cursor()
import_schema = [
"DROP TABLE IF EXISTS sections;",
"""CREATE TABLE sections (
name CHARACTER VARYING(20) PRIMARY KEY
)""",
"DROP TABLE IF EXISTS classes;",
"""CREATE TABLE classes (
id CHARACTER VARYING(20) PRIMARY KEY,
name CHARACTER VARYING(20)
)""",
"DROP TABLE IF EXISTS users;",
"""CREATE TABLE users (
id INTEGER PRIMARY KEY,
name CHARACTER VARYING(50),
gender CHARACTER VARYING(20) DEFAULT "u",
section CHARACTER VARYING(20) DEFAULT ""
)""",
"DROP TABLE IF EXISTS buys;",
"""CREATE TABLE buys (
id INTEGER PRIMARY KEY,
put_at pydatetime,
user INTEGER REFERENCES users
)""",
"DROP TABLE IF EXISTS semesters;",
"""CREATE TABLE semesters (
id INTEGER PRIMARY KEY,
name CHARACTER(6),
start pydate,
end pydate
)""",
"DROP TABLE IF EXISTS shifts;",
"""CREATE TABLE shifts (
id INTEGER PRIMARY KEY,
date pydate,
span TYNYINT,
semester INTEGER REFERENCES semesters
)""",
"DROP TABLE IF EXISTS signups;",
"""CREATE TABLE signups (
id INTEGER PRIMARY KEY,
shift INTEGER REFERENCES shifts,
user INTEGER REFERENCES user,
klass CHARACTER VARYING(20) REFERENCES classes
)""",
"""CREATE INDEX buy_user_idx ON buys(user)""",
"""CREATE INDEX shift_semester_idx ON shifts(semester)""",
"""CREATE INDEX signup_shift_idx ON signups(shift)""",
"""CREATE INDEX signup_user_idx ON signups(user)""",
"""CREATE INDEX signup_klass_idx ON signups(klass)""",
]
for g in CLASSES:
import_schema.append(
"""INSERT INTO classes (id, name) VALUES ("%s", "%s");""" % (g, g))
def do_queries(queries):
for query in queries:
try:
c.execute(query)
except sqlite3.OperationalError:
print query
raise
conn.commit()
class HashError(Exception):
pass
class Hash(object):
def __init__(self, row):
if row is None:
raise HashError(row)
self.cols = []
self.values = []
for idx, col in enumerate(row.keys()):
self.cols.append(col)
self.values.append(row[idx])
setattr(self, col, row[idx])
self.items = zip(self.cols, self.values)
def __str__(self):
return str(self.__dict__)
def __repr__(self):
return repr(self.__dict__)
def __hash__(self):
return hash(self.id)
def __int__(self):
return self.id
def hashes(res):
return [Hash(r) for r in res]
@memoized
def get_interval(interval):
c.execute(
"SELECT * FROM intervals WHERE start = ? AND end = ? LIMIT 1",
(interval,))
return Hash(c.fetchone())
@memoized
def get_intervals():
c.execute("SELECT * FROM intervals ORDER BY start")
return hashes(c.fetchall())
@memoized
def buys_within(interval):
c.execute(
"SELECT * FROM interval_buys WHERE interval = ?;",
(interval.id,))
return hashes(c.fetchall())
@memoized
def user_buys_within(interval, user):
c.execute(
"""SELECT count(*) as count FROM interval_buys
WHERE interval = ? AND user = ?;""",
(interval.id, user.id))
return Hash(c.fetchone()).count
@memoized
def shifts_within(interval):
c.execute(
"SELECT * FROM interval_shifts WHERE interval = ?;""",
(interval.id,))
return hashes(c.fetchall())
@memoized
def signups_within(interval, g=None):
extra = ""
args = [interval.id,]
if g in [WORKER, ONCALL]:
extra = ' AND klass = ?',
args.append(g)
c.execute(
"""SELECT * FROM interval_signups WHERE interval = ?%s;""" % extra,
args)
return hashes(c.fetchall())
@memoized
def get_users(iterable, get_set=False):
user_ids = set()
for i in iterable:
if isinstance(i, int):
user_ids.add(str(i))
else:
user_ids.add(str(i.user))
c.execute(
"SELECT * FROM users WHERE id IN (%s);" % ", ".join(user_ids))
return hashes(c.fetchall())
@memoized
def with_gender(gender, users):
assert gender in GENDERS
return filter(lambda user: user.gender == gender, users)
@memoized
def class_within(interval, g):
if g in [WORKER, ONCALL]:
return get_users(signups_within(interval, g))
elif g == SIGNUP:
return get_users(signups_within(interval))
elif g == BUYER:
return get_users(buys_within(interval))
elif g == CUSTOMER:
getter = attrgetter('user')
def get_set(func):
return set(map(getter, func(interval)))
buyers = get_set(buys_within)
signups = get_set(signups_within)
customers = buyers.difference(signups)
return get_users(customers)
elif g == ALL:
iterable = []
iterable.extend(signups_within(interval))
iterable.extend(buys_within(interval))
return get_users(iterable)
elif g in GENDERS:
return with_gender(g, class_within(interval, ALL))
assert False, g
@memoized
def classes_within(interval):
classes = {}
for g in AGG_CLASSES:
classes[g] = class_within(interval, g)
return classes
@memoized
def traverse(obj, func, test=None, context=None):
if context is not None:
context = copy.copy(context)
if isinstance(obj, dict):
for key, val in obj.items():
if not callable(test) or test(key):
obj[key] = traverse(val, func, test, context)
return obj
return func(obj)
@memoized
def _gender_node(obj):
new = {}
for gender in GENDERS:
func = functools.partial(with_gender, gender)
new[gender] = func(obj)
new[ALL] = obj
return new
def _gender_test(key):
return key not in [ALL] + GENDERS
def get_gendered(interval, obj):
return traverse(obj, _gender_node, _gender_test)
@memoized
def _interval_buy_node(interval, users):
return [user_buys_within(interval, user) for user in users]
def get_buys(interval, obj):
func = functools.partial(_interval_buy_node, interval)
return traverse(obj, func)
def get_counted(interval, obj):
return traverse(obj, len)
def get_summed(interval, obj):
return traverse(obj, sum)
def insert_interval(name, start, end):
c.execute(
"SELECT * FROM buys WHERE put_at >= ? AND put_at < ?",
(start, end + relativedelta(days=1)))
buys = hashes(c.fetchall())
c.execute(
"SELECT * FROM shifts WHERE date >= ? AND date <= ?",
(start, end))
shifts = hashes(c.fetchall())
c.execute("SELECT count(*) AS offset FROM interval_shifts;")
c.execute(
"SELECT * FROM signups WHERE shift in (%s);" % \
", ".join([str(shift.id) for shift in shifts]))
signups = hashes(c.fetchall())
c.execute(
"""INSERT INTO intervals
(name, start, end, buy_count)
VALUES (?, ?, ?, ?);""",
(name, start, end, len(buys)))
idx = c.lastrowid
c.executemany(
"INSERT INTO interval_buys (interval, user, put_at) VALUES (?, ?, ?);",
[(idx, buy.user, buy.put_at) for buy in buys])
c.executemany(
"""INSERT INTO interval_shifts
(id, interval, date, span, semester) VALUES (?, ?, ?, ?, ?);""",
[(shift.id, idx, shift.date, shift.span, shift.semester) for shift in shifts])
c.executemany(
"""INSERT INTO interval_signups
(interval, shift, user, klass) VALUES (?, ?, ?, ?);""",
[(idx, signup.shift, signup.user, signup.klass) for signup in signups])
extra_schema = [
"DROP TABLE IF EXISTS intervals;",
"""CREATE TABLE intervals (
id INTEGER PRIMARY KEY,
name CHARACTER VARYING(50) DEFAULT "",
start pydate,
end pydate,
buy_count INTEGER
)""",
"DROP TABLE IF EXISTS interval_shifts;",
"""CREATE TABLE interval_shifts (
id INTEGER PRIMARY KEY,
interval INTEGER REFERENCES intervals,
date pydate,
span TYNYINT,
semester INTEGE REFERENCES semesters
)""",
"DROP TABLE IF EXISTS interval_signups;",
"""CREATE TABLE interval_signups (
id INTEGER PRIMARY KEY,
interval INTEGER REFERENCES intervals,
shift INTEGER REFERENCES interval_shifts,
user INTEGER REFERENCES users,
klass CHARACTER VARYING(20) REFERENCES classes
)""",
"DROP TABLE IF EXISTS interval_buys;",
"""CREATE TABLE interval_buys (
id INTEGER PRIMARY KEY,
interval INTEGER REFERENCES intervals,
user INTEGER REFERENCES users,
put_at pydatetime
)""",
"DROP TABLE IF EXISTS interval_buy_counts;",
"""CREATE TABLE interval_buy_counts (
id INTEGER PRIMARY KEY,
interval INTEGER REFERENCES intervals,
user INTEGER REFERENCES users,
count INTEGER
)""",
"""CREATE INDEX i_shift_interval_idx ON interval_shifts(interval)""",
"""CREATE INDEX i_shift_semester_idx ON interval_shifts(semester)""",
"""CREATE INDEX i_signup_interval_idx ON interval_signups(interval)""",
"""CREATE INDEX i_signup_shift_idx ON interval_signups(shift)""",
"""CREATE INDEX i_signup_user_idx ON interval_signups(user)""",
"""CREATE INDEX i_signup_klass_idx ON interval_signups(klass)""",
"""CREATE INDEX i_buy_interval_idx ON interval_buys(interval)""",
"""CREATE INDEX i_buy_user_idx ON interval_buys(user)""",
"""CREATE INDEX i_buy_count_interval_idx ON interval_buy_counts(interval)""",
"""CREATE INDEX i_buy_count_user_idx ON interval_buy_counts(user)""",
]
def transform(interval, base, steps):
if steps:
return transform(interval, steps[0](interval, base), steps[1:])
return base
@memoized
def get_metric(interval, metric, klass, gender):
c.execute(
"""SELECT * FROM %(db_plural)s WHERE
interval = ? AND
klass = ? AND
gender = ? LIMIT 1""" % get_metric_context(metric),
(interval.id, klass, gender))
return Hash(c.fetchone())
def dictzip(*dicts):
result = dict()
dict_keys = []
non_dict_keys = []
for k, v in dicts[0].items():
if isinstance(v, dict):
dict_keys.append(k)
else:
non_dict_keys.append(k)
result[k] = []
for k in dict_keys:
result[k] = dictzip(*[d[k] for d in dicts])
for d in dicts:
for k in non_dict_keys:
result[k].append(d[k])
return result
AGG_CLASSES = [CUSTOMER, WORKER]
METRICS = {
'buys': {
'py_type': int,
'basic': True,
'transforms': (get_buys, get_summed),
},
'people': {
'py_type': int,
'basic': True,
'transforms': (get_counted,),
},
'buys_per_person': {
'py_type': float,
'basic': False,
},
'liters': {
'py_type': float,
'basic': False,
},
'liters_per_person': {
'py_type': float,
'basic': False,
},
}
def get_metrics(metric_type=False):
metric_hashes = []
for metric, meta in METRICS.items():
if metric_type and metric != metric_type:
continue
context = get_metric_context(metric)
c.execute("SELECT * FROM %(db_plural)s ORDER BY id" % context)
metric_hashes.extend(hashes(c.fetchall()))
return metric_hashes
def clear_metrics(metric_type=False):
for metric, meta in METRICS.items():
if metric_type and metric != metric_type:
continue
context = get_metric_context(metric)
c.execute("DELETE FROM %(db_plural)s" % context)
def get_metric_context(metric):
meta = METRICS[metric]
return dict(
db_singular='metric_%s' % metric,
db_plural='metric_%ss' % metric,
py_type=meta['py_type'],
db_type=db_type(meta['py_type']),
)
def get_metric_schema(metric):
templates = [
"DROP TABLE IF EXISTS %(db_plural)s;",
"""CREATE TABLE %(db_plural)s (
id INTEGER PRIMARY KEY,
interval INTEGER REFERENCES intervals,
klass CHARACTER VARYING(20) REFERENCES classes,
gender CHARACTER VARYING(20) DEFAULT "u",
value %(db_type)s
)""",
"""CREATE INDEX %(db_plural)s_interval_idx ON %(db_plural)s(interval)""",
"""CREATE INDEX %(db_plural)s_klass_idx ON %(db_plural)s(klass)""",
]
return [template % get_metric_context(metric) for template in templates]
def insert_metric(interval, metric, klass, gender, value, commit=True):
sql = """INSERT INTO %(db_plural)s
(interval, klass, gender, value)
VALUES (?, ?, ?, ?);""" % get_metric_context(metric)
c.execute(sql, (int(interval), klass, gender, value))
if commit:
conn.commit()
def do_basic_metrics():
basic_metrics = [metric for metric, meta in METRICS.items() if meta['basic']]
for metric in basic_metrics:
intervals = get_intervals()
meta = METRICS[metric]
for interval in intervals:
gendered = get_gendered(interval, classes_within(interval))
data = transform(interval, copy.deepcopy(gendered), meta['transforms'])
for klass in AGG_CLASSES:
for gender in [ALL] + GENDERS:
args = (
interval,
metric,
klass,
gender,
data[klass][gender],
)
insert_metric(*args, commit=False)
conn.commit()
def do_extra_metrics():
buys = get_metrics('buys')
people = get_metrics('people')
clear_metrics('buys_per_person')
for b, p in zip(buys, people):
mean = float(b.value) / max(1.0, p.value)
insert_metric(
b.interval,
'buys_per_person',
b.klass,
b.gender,
mean, commit=False,
)
buys_per_person = get_metrics('buys_per_person')
clear_metrics('liters')
for b in buys:
insert_metric(
b.interval,
'liters',
b.klass,
b.gender,
b.value * CUP_LITERS, commit=False,
)
clear_metrics('liters_per_person')
for bpp in buys_per_person:
insert_metric(
bpp.interval,
'liters_per_person',
bpp.klass,
bpp.gender,
bpp.value * CUP_LITERS, commit=False,
)
conn.commit()
def num(obj):
try:
return len(obj)
except TypeError:
return int(obj)
def ratio(obj, pool):
return float(num(obj)) / max(1.0, num(pool))
def formatpc(obj, pool):
r = ratio(obj, pool)
pad = 7
if r > 1.0:
return "#" * pad
else:
s = "%.2f%%" % (100.0 * r)
return s.rjust(pad)
def get_semesters():
c.execute("SELECT * FROM semesters ORDER BY start;")
return hashes(c.fetchall())
def main():
if settings.DO_IMPORT:
do_queries(import_schema)
name_gender = {}
male_names = scb.names.male()
female_names = scb.names.female()
for names, gender in [(male_names, MALE), (female_names, FEMALE)]:
for name in names:
name_gender[name] = gender
with open('mining.pkl', 'rb') as f:
d = pickle.load(f)
del f
for sec in d['sections']:
c.execute("INSERT INTO sections (name) VALUES (?)", (sec,))
conn.commit()
for uid in d['users']:
name = d['user_name'][uid]
gender = name_gender.get(name, UNKNOWN)
section = d['user_section'][uid]
c.execute(
"""INSERT INTO users
(id, name, gender, section)
VALUES (?, ?, ?, ?)
""",
(uid, name, gender, section))
conn.commit()
buy_args = []
for user, buys in d['user_orders'].items():
buy_args.extend((put_at, user) for put_at in buys)
c.executemany(
"""INSERT INTO buys
(put_at, user)
VALUES (?, ?)
""",
buy_args)
conn.commit()
for sem in d['semesters']:
name = d['semester_name'][sem]
start, end = d['semester_startend'][sem]
c.execute(
"""INSERT INTO semesters
(id, name, start, end)
VALUES (?, ?, ?, ?)
""",
(sem, name, start, end))
conn.commit()
for shift in d['shifts']:
date = d['shift_date'][shift]
span = d['shift_span'][shift]
sem = d['shift_semester'][shift]
c.execute(
"""INSERT INTO shifts
(id, date, span, semester)
VALUES (?, ?, ?, ?)
""",
(shift, date, span, sem))
conn.commit()
signup_args = []
for signup_key, signup_class in [
('user_workshifts', WORKER),
('user_oncallshifts', ONCALL)]:
for user, shifts in d[signup_key].items():
signup_args.extend([(shift, user, signup_class) for shift in shifts])
c.executemany(
"""INSERT INTO signups
(shift, user, klass)
VALUES (?, ?, ?)
""",
signup_args)
conn.commit()
if settings.DO_EXTRAS:
do_queries(extra_schema)
for sem in get_semesters():
insert_interval(sem.name, sem.start, sem.end)
conn.commit()
basic_metrics = [metric for metric, meta in METRICS.items() if meta['basic']]
if settings.DO_BASIC_METRICS:
for metric in basic_metrics:
do_queries(get_metric_schema(metric))
do_basic_metrics()
extra_metrics = [metric for metric, meta in METRICS.items() if not meta['basic']]
if settings.DO_EXTRA_METRICS:
for metric in extra_metrics:
do_queries(get_metric_schema(metric))
do_extra_metrics()
if settings.DO_CSV:
for metric, meta in METRICS.items():
csv_file = "%s.csv" % metric
fields = ['interval']
lookups = []
for klass in AGG_CLASSES:
for gender in [ALL, MALE, FEMALE]:
lookups.append((klass, gender))
fields.append("%s_%s" % (klass, gender))
with open(csv_file, 'w') as csv:
line = ", ".join(map(str, fields))
csv.write(line + '\n')
for interval in get_intervals():
try:
values = []
for lookup in lookups:
values.append(get_metric(interval, metric, *lookup).value)
total = sum(values)
if total == 0:
continue
if metric == 'buys' and total < 500:
continue
if metric == 'people' and total < 250:
continue
line = ",".join(map(str, [interval.name] + values))
with open(csv_file, 'a+') as csv:
csv.write(line + '\n')
except HashError:
pass
if __name__ == '__main__':
main()