-
Notifications
You must be signed in to change notification settings - Fork 1
/
gen_real_error.py
433 lines (355 loc) · 19.2 KB
/
gen_real_error.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
from postgres import *
from psql_explain_decoder import *
from sklearn.neighbors import KernelDensity
import random
import pandas as pd
from prep_error_list import plot_error, cal_rel_error
from querylets import *
import copy
file_name_to_save_real_error = ''
kk = '1=1'
cache_right = {}
db = 'imdb'
def gen_real_error():
global file_name_to_save_real_error
if db == 'imdb':
local_selections = pd.read_csv('lsc/job/LSC-JOB.csv')
# ['Table', 'Condition', 'Frequency']
condition_dict = {'k': [], 't': [], 'cn': [], 'n': [], 'mc': [], 'mi': [], 'it_pi': [], 'it_mi': [], 'it_miidx': [], 'an': [], 'lt': [], 'pi': [], 'ci':[], 'mi_idx':[], 'kt':[], 'ct':[], 'rt':[], 'cct':[], 'chn':[]}
if db == 'dsb':
local_selections = pd.read_csv('lsc/dsb/DSB-072.csv')
table_names_from_csv = local_selections['Table'].unique()
condition_dict = {key: [] for key in table_names_from_csv}
# condition_dict = {
# "call_center": [], "catalog_returns": [],
# "catalog_sales": [], "customer": [],
# "customer_address": [], "customer_demographics": [],
# "date_dim": [],
# "household_demographics": [],
# "income_band": [], "item": [],
# "ship_mode": [], "store": [], "store_sales": [],
# "warehouse": [], "web_sales": [],
# }
if db == 'stats':
local_selections = pd.read_csv('lsc/stats/LSC-Stats.csv')
condition_dict = {'b': [], 'c': [], 'u': [], 'ph': [], 'p': [], 'pl': [], 'v': []}
local_selections_grouped = local_selections.groupby('Table')
frequency_dict = copy.deepcopy(condition_dict)
frequency_dict['x'] = [2]
for table in condition_dict.keys():
for _, row in local_selections_grouped.get_group(table).iterrows():
if row['Condition'] == '1=1':
continue
condition_dict[table].append(row['Condition'])
frequency_dict[table].append(int(row['Frequency']))
condition_dict['x'] = ['1=1']
if db == 'imdb':
frequency_dict['mk'] =[1]
condition_dict['mk'] = ['1=1']
frequency_dict['akat'] =[1]
condition_dict['akat'] = ['1=1']
frequency_dict['cc'] =[1]
condition_dict['cc'] = ['1=1']
data_list = []
left = 'x'
left_single = 'ci_full'
right = "x"
right_single = 'mk_k_r'
template_name = f'template_mk_ci__k'
for id_1, left_condition in enumerate(condition_dict[left]):
for id_2, right_condition in enumerate(condition_dict[right]):
right_condition = "k.keyword ='character-name-in-title'"
template = querylet(db, left_condition, right_condition, template_name)
left_template = querylet(db, right_condition, left_condition, 'template_'+left_single)
right_template = querylet(db, left_condition, right_condition, 'template_'+right_single)
# template_full = querylet(db, left_condition, right_condition, 'template_'+template_name + '_full')
print(template)
file_name_to_save_real_error = template_name.split('template_')[1]
# data = cal_local_selectivity(template, template_full)
# input() # WARNING
data = cal_join_selectivity(template, left_template, right_template, id_2)
if data:
print(data, cal_rel_error(data[0], data[1]), math.log(data[1] / data[0]))
data_list.extend([data]*frequency_dict[right][id_2]*frequency_dict[left][id_1])
output = [str(data[0]) +" "+ str(data[1]) for data in data_list]
# print(output)
input()
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'w') as fp:
fp.write('\n'.join(output))
plot_pdf()
exit()
for right in condition_dict.keys():
for left in condition_dict.keys():
for l_r_b in ['l', 'r', 'both']:
data_list = []
right = right
template_name = f'{left}_{right}_{l_r_b}'
# fix 'it'
# if left.split('_')[0] == 'it' and left.split('_')[1] == right:
# template_name = f'{left}_{l_r_b}'
# elif right.split('_')[0] == 'it' and right.split('_')[1] == left:
# template_name = f'{right.split('_')[1]}_{right.split('_')[0]}_{l_r_b}'
# else:
# continue
# if template_name.split('_')[0] == 'it' or template_name.split('_')[1] == 'it':
# print(template_name)
# input()
# else:
# continue
# don't care those redundent template; only care simplest ones now
if not check_tempalte(db, 'template_'+template_name):
continue
if os.path.exists(f'./data/abs-error-imdb/{template_name}.txt'):
print("yes")
continue
else:
print("No", template_name)
input()
# continue
for id_1, left_condition in enumerate(condition_dict[left]):
for id_2, right_condition in enumerate(condition_dict[right]):
if l_r_b == 'l':
template = querylet(db, right_condition, left_condition, 'template_'+template_name)
else:
template = querylet(db, left_condition, right_condition, 'template_'+template_name)
if l_r_b == 'l':
right_single = right+'_full'
freq_right = 'x'
else:
right_single = right
freq_right = right
if l_r_b == 'r':
left_single = left + '_full'
freq_left = 'x'
else:
left_single = left
freq_left = left
left_template = querylet(db, right_condition, left_condition, 'template_'+left_single)
right_template = querylet(db, left_condition, right_condition, 'template_'+right_single)
# template_full = querylet(db, left_condition, right_condition, 'template_'+template_name + '_full')
print(template)
# input()
print(left_template)
print(right_template)
file_name_to_save_real_error = template_name
# data = cal_local_selectivity(template, template_full)
# input() # WARNING
data = cal_join_selectivity(template, left_template, right_template, id_2)
if data:
print(data, cal_rel_error(data[0], data[1]), math.log(data[1] / data[0]))
data_list.extend([data]*frequency_dict[freq_right][id_2]*frequency_dict[freq_left][id_1])
if l_r_b == 'l':
break # since we don't need to go through right conditions
if l_r_b == 'r':
break # since we don't need to go through left conditions
output = [str(data[0]) +" "+ str(data[1]) for data in data_list]
# print(output)
input()
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'w') as fp:
fp.write('\n'.join(output))
plot_pdf()
if db == 'stats':
data_list = []
left = 'x'
left_single = 'b_full'
right = 'c'
right_single = 'c_ph_l'
template_name = f'template_ph_b__c'
for id_1, left_condition in enumerate(condition_dict[left]):
for id_2, right_condition in enumerate(condition_dict[right]):
template = stats_complex_querylet(cc=right_condition, template=template_name)
left_template = stats_single_querylet(left_condition, left_single)
right_template = stats_join_querylet(left_alias='c', right_alias='ph', l_r_b='l',
cc=right_condition, kk=left_condition)
print(template)
file_name_to_save_real_error = 'ph_b__c'
# data = cal_local_selectivity(template, template_full)
# input() # WARNING
data = cal_join_selectivity(template, left_template, right_template, id_2)
if data:
print(data, cal_rel_error(data[0], data[1]), math.log(data[1] / data[0]))
data_list.extend([data]*frequency_dict[right][id_2]*frequency_dict[left][id_1])
output = [str(data[0]) +" "+ str(data[1]) for data in data_list]
# print(output)
input()
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'w') as fp:
fp.write('\n'.join(output))
plot_pdf()
exit()
# for l_id, left in enumerate(['p', 'c', 'ph', 'pl', 'v']):
# for r_id, right in enumerate(['p', 'c', 'ph', 'pl', 'v']):
# for l_id, left in enumerate(['u', 'c', 'b', 'v']):
# for r_id, right in enumerate(['u', 'c', 'b', 'v']):
for l_id, left in enumerate(['ph']):
for r_id, right in enumerate(['b']):
# if r_id <= l_id: continue
for l_r_b in ['both']:
data_list = []
# right = 'income_band'
template_name = f'{left}_{right}_{l_r_b}'
template_id = '_2'
for id_1, left_condition in enumerate(random.sample(condition_dict[left], 10)):
for id_2, right_condition in enumerate(random.sample(condition_dict[right], 10)):
file_name_to_save_real_error = template_name + template_id
template = stats_join_querylet(left, right, l_r_b, left_condition, right_condition)
if l_r_b == 'l':
right_single = right+'_full'
freq_right = 'x'
else:
right_single = right
freq_right = right
if l_r_b == 'r':
left_single = left + '_full'
freq_left = 'x'
else:
left_single = left
freq_left = left
left_template = stats_single_querylet(left_condition, left_single)
right_template = stats_single_querylet(right_condition, right_single)
print(template)
print(left_template)
print(right_template)
# data = cal_local_selectivity(template, template_full)
# input() # WARNING
data = cal_join_selectivity(template, left_template, right_template, id_2)
if data:
print(data, cal_rel_error(data[0], data[1]), math.log(data[1] / data[0]))
data_list.extend([data]*frequency_dict[freq_right][id_2]*frequency_dict[freq_left][id_1])
# input()
if l_r_b == 'l':
break # since we don't need to go through right conditions
if l_r_b == 'r':
break # since we don't need to go through left conditions
output = [str(data[0]) +" "+ str(data[1]) for data in data_list]
# print(output)
# input()
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'w') as fp:
fp.write('\n'.join(output))
plot_pdf()
if db == 'dsb':
data_list = []
left = 'x'
left_single = 'warehouse_full'
right = 'catalog_sales'
right_single = 'inventory_catalog_sales_r'
template_name = f'inventory_warehouse__catalog_sales'
query_let_type_is_join = True
file_name_to_save_real_error = 'inventory_warehouse__catalog_sales_072'
for id_1, left_condition in enumerate(condition_dict[left]):
for id_2, right_condition in enumerate(condition_dict[right]):
# left_condition = left_condition.replace("s2.", "")
right_condition = right_condition.replace("d1.", "")
if query_let_type_is_join:
template = querylet(db, left_condition, right_condition, 'template_'+ template_name)
left_template = querylet(db, right_condition, left_condition.replace('d1.', ''), 'template_'+left_single)
right_template = querylet(db, left_condition, right_condition.replace('d2.', ''), 'template_'+right_single)
print(template)
data = cal_join_selectivity(template, left_template, right_template, id_2)
else:
template = querylet(db, left_condition, right_condition, 'template_'+ template_name)
template_full = querylet(db, left_condition, right_condition, 'template_'+ template_name + '_full')
print(template_full)
print(template)
data = cal_local_selectivity(template, template_full)
if data:
print(data, cal_rel_error(data[0], data[1]), math.log(data[1] / data[0]))
data_list.extend([data]*frequency_dict[right][id_2]*frequency_dict[left][id_1])
output = [str(data[0]) +" "+ str(data[1]) for data in data_list]
# print(output)
input()
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'w') as fp:
fp.write('\n'.join(output))
plot_pdf()
exit()
def cal_join_selectivity(join_template, left_template, right_template, id):
global cache_right
est_join_count, act_join_count = get_est_act_count(join_template)
est_left_count, act_left_count = get_est_act_count(left_template)
if id not in cache_right.keys():
est_right_count, act_right_count = get_est_act_count(right_template)
cache_right[id] = [est_right_count, act_right_count]
else:
est_right_count, act_right_count = cache_right[id]
print("=== Use cached")
print(f"join rows est: {est_join_count}, act: {act_join_count}")
print(f"left rows est: {est_left_count}, act {act_left_count}")
print(f"right rows est: {est_right_count}, act: {act_right_count}")
if est_left_count == 0 or act_left_count == 0 or est_right_count == 0 or act_right_count == 0 or act_join_count == 0:
return False
est_sel_join = max(1, est_join_count) / (est_left_count * est_right_count)
act_sel_join = max(1, act_join_count) / (act_left_count * act_right_count)
return [act_sel_join, est_sel_join]
def cal_local_selectivity(local_template, full_table_template):
est_count, act_count = get_est_act_count(local_template)
est_count_full, act_count_full = get_est_act_count(full_table_template)
if act_count_full == 0 or est_count_full == 0:
return False
else:
return [max(1, act_count)/act_count_full, max(1, est_count)/act_count_full]
def get_est_act_count(template):
if db=='imdb':
join_plans = get_real_latency('imdbloadbase', template, times=1, return_json=True, limit_time=False, limit_worker=True, drop_buffer=False)
else:
join_plans = get_real_latency(db, template, times=1, return_json=True, limit_time=False, limit_worker=True, drop_buffer=False)
join_plans = join_plans[0][0][0]['Plan']
node_type = join_plans['Node Type']
while True:
if node_type in ['Aggregate', 'Gather', 'Sort', 'Materialize', 'Sort', 'Hash', 'Gather Merge']:
join_plans = join_plans["Plans"][0]
node_type = join_plans['Node Type']
else:
break
print(join_plans)
est_join_count = join_plans['Plan Rows']
act_join_count = join_plans['Actual Rows']
return est_join_count, act_join_count
def plot_pdf():
print(file_name_to_save_real_error)
with open('./data/abs-error-'+db+'/' + file_name_to_save_real_error + '.txt', 'r') as fp:
lines = fp.readlines()
data = [x.strip().split() for x in lines]
abs_error_list = []
cleaned_data = []
for x in data:
# if float(x[0]) > 0.001:
# continue
# if float(x[0]) > 1 or float(x[1]) > 1 or float(x[0]) < 0 or float(x[1]) < 0:
# continue
# else:
cleaned_data.append([float(x[0])/134170, float(x[1])/134170])
# Err = true - est
abs_err = [float(x[0]) - float(x[1]) for x in cleaned_data]
abs_err = sorted(abs_err)
abs_err = np.array(abs_err).reshape(-1, 1)
print(max(abs_err), "max abs (true-est) error")
print(min(abs_err), "min abs (true-est) error")
count_1 = 0
count_2 = 0
for i in abs_err:
if i > 0:
count_1 += 1
else:
count_2 += 1
print(count_1/(count_1 + count_2), ": true>est ", count_2/(count_1 + count_2), ": true<est")
kde = KernelDensity(kernel="gaussian", bandwidth=0.3).fit(abs_err)
# plot_error(abs_err, kde, name="data/abs-error-dsb/cn-mc_abs")
relative_error_list = []
for x in data:
if float(x[0]) == 0 or 0 == float(x[1]):
continue
relative_error_list.append(-cal_rel_error(float(x[0]), float(x[1])))
# print(relative_error_list)
relative_error_list = np.array(relative_error_list).reshape(-1, 1)
print(max(relative_error_list), "max rel error")
print(min(relative_error_list), "min rel error")
kde = KernelDensity(kernel="gaussian", bandwidth=1).fit(relative_error_list)
plot_error(relative_error_list, kde, rel_error=True, name='data/abs-error-'+db+'/'+file_name_to_save_real_error)
def check_tempalte(db, template_name):
if not querylet(db, '', '', template_name):
return False
else:
return True
gen_real_error()
# db = 'imdb'
# file_name_to_save_real_error = "mk_ci__q17"
# plot_pdf()