-
Notifications
You must be signed in to change notification settings - Fork 2
/
evaluate.py
209 lines (201 loc) · 10.9 KB
/
evaluate.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
"""
Small demo how to inspect results of the Python Package DBMS Benchmarker.
Mode resultsets prints the first difference in result sets for each query in the latest benchmark in the current folder.
Copyright (C) 2021 Patrick Erdelt
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as
published by the Free Software Foundation, either version 3 of the
License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
"""
from dbmsbenchmarker import *
import pandas as pd
from operator import itemgetter
import argparse
import ast
def convertToInt(var):
"""
Converts variable to float.
:param var: Some variable
:return: returns float converted variable
"""
#print(var)
#print(type(var))
try:
return int(var)
except Exception as e:
#print(str(e))
#print("Not convertible")
#print(var)
return var
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='A debug tool for DBMSBenchmarker. It helps to analyze a result folder. It depends on the evaluation cube, so that cube must have been created before.')
parser.add_argument('-r', '--result-folder', help='folder for storing benchmark result files, default is given by timestamp', default="./")
parser.add_argument('-e', '--experiment', help='code of experiment', default="")
parser.add_argument('-q', '--query', help='number of query to inspect', default=None)
parser.add_argument('-c', '--connection', help='name of DBMS to inspect', default=None)
parser.add_argument('-n', '--num-run', help='number of run to inspect', default=None)
parser.add_argument('-d', '--diff', help='show differences in result sets', action='store_true', default=False)
parser.add_argument('-rt', '--remove-titles', help='remove titles when comparing result sets', action='store_true', default=False)
parser.add_argument('mode', help='show debug infos about which part of the outcome', choices=['resultsets', 'errors', 'warnings', 'query'])
args = parser.parse_args()
# path of folder containing experiment results
resultfolder = args.result_folder#"./"
# create evaluation object for result folder
evaluate = inspector.inspector(resultfolder)
# list of all experiments in folder
evaluate.list_experiments
# dataframe of experiments
evaluate.get_experiments_preview()
if not len(args.experiment) > 0:
# pick last experiment
code = evaluate.list_experiments[len(evaluate.list_experiments)-1]
else:
code = args.experiment
# load it
evaluate.load_experiment(code)
list_connections = evaluate.get_experiment_list_connections()
list_queries = evaluate.get_experiment_list_queries()
if args.mode == 'resultsets':
for numQuery in list_queries:
if args.query is not None and int(args.query) != numQuery:
continue
query = tools.query(evaluate.benchmarks.queries[numQuery-1])
#print("PRECISION:")
#print(query.restrict_precision)
list_warnings = evaluate.get_warning(numQuery)
#print("Q"+str(numQuery))
#print(list_warnings)
numRun=0
df = evaluate.get_datastorage_df(numQuery, numRun)
data = df.values
if len(data) > 0:
# there are warnings: and sum([len(v) for k,v in list_warnings.items()]) > 0:
print("\n===Q{}: {}===".format(numQuery, query.title))
print(list_warnings)
pd.set_option('display.max_colwidth', None)
pd.set_option('display.max_rows', None)
if not args.diff:
print("Storage:", df)
df = pd.DataFrame(sorted(data, key=itemgetter(*list(range(0,len(data[0]))))), columns=df.columns)
#print(df)
for c in list_connections:
if c in list_warnings and len(list_warnings[c])>0:
s = evaluate.benchmarks.protocol['query'][str(numQuery)]['dataStorage']
r = evaluate.benchmarks.protocol['query'][str(numQuery)]['resultSets']
#print(r[c])
#print(r[c][numRun][0])
#print(s[numRun][0])
#print(s)
#exit()
for numRun, result in enumerate(s):
if args.num_run is not None and int(args.num_run) != numRun:
continue
if not args.diff:
print(c, evaluate.get_resultset_df(numQuery, c, numRun))
continue
df2={}
data_stored = s[numRun]
if len(data_stored) == 0:
continue
if len(r[c]) == 0:
continue
print("numRun: "+str(numRun+1))
#print(data)
#print(data_stored, r[c][numRun])
s2 = data_stored
r2 = r[c][numRun]
#print(s2)
#print(r2)
#print(s[numRun], s[numRun][0])
titles_storage = s[numRun][0]
titles_resultset = r[c][numRun][0]
# remove titles
if args.remove_titles:
#s2 = [l.pop(0) for l in s2]
titles_storage = list(range(len(s[numRun][0])))
s2.pop(0)
# remove titles
if args.remove_titles:
#r2 = [l.pop(0) for l in r2]
titles_resultset = list(range(len(r[c][numRun][0])))
r2.pop(0)
# convert datatypes
s2 = [[round(float(item), int(query.restrict_precision)) if tools.convertToFloat(item) == float else convertToInt(item) if convertToInt(item) == item else item for item in sublist] for sublist in s2]
r2 = [[round(float(item), int(query.restrict_precision)) if tools.convertToFloat(item) == float else convertToInt(item) if convertToInt(item) == item else item for item in sublist] for sublist in r2]
#print("storage", s2)
#print("result", r2)
#print(len(s2[0]), titles_storage)
if len(s2) > 0:
#print(s2)
#print(itemgetter(*list(range(0,len(s2[0])))))
#for x in s2:
# for y in x:
# print(y,type(y))
df = pd.DataFrame(sorted(s2, key=itemgetter(*list(range(0,len(s2[0]))))), columns=titles_storage)#df_tmp.columns)
#print(df)
if len(r2) > 0 and r2 != s2:
#df_tmp = evaluate.get_resultset_df(numQuery, c, numRun)
#print(df_tmp.columns)
#data = df_tmp.values
#r2 = [[round(float(item), int(query.restrict_precision)) if tools.convertToFloat(item) == float else convertToInt(item) if convertToInt(item) == item else item for item in sublist] for sublist in data]
data = r2
if len(data) > 0:
df2[c] = pd.DataFrame(sorted(data, key=itemgetter(*list(range(0,len(data[0]))))), columns=titles_resultset)#df_tmp.columns)
else:
df2[c] = pd.DataFrame()
diff = False
for c,d in df2.items():
if not inspector.getDifference12(df,d).empty:
print("Storage has more than {}:".format(c))
print(inspector.getDifference12(df,d))
diff = True
if not inspector.getDifference12(d,df).empty:
print("{} has more than storage".format(c))
print(inspector.getDifference12(d,df))
diff = True
#print("data", df,df2)
if not diff:
print("same")
if len(df2) > 0:
break
else:
print("no data")
elif args.mode == 'errors':
print(evaluate.get_total_errors())
for numQuery in list_queries:
if args.query is not None and int(args.query) != numQuery:
continue
query = tools.query(evaluate.benchmarks.queries[numQuery-1])
list_errors = evaluate.get_error(numQuery)
#print(list_errors)
list_errors = {k:v for k,v in list_errors.items() if len(v) > 0}
df2_errors = pd.DataFrame(list_errors, index=['ERROR'])
df2_errors = df2_errors.T.sort_index()
pd.set_option('display.max_colwidth', None)
if not df2_errors.empty:
print("===Q{}: {}===".format(numQuery, query.title))
print(df2_errors)
elif args.mode == 'warnings':
print(evaluate.get_total_errors())
for numQuery in list_queries:
if args.query is not None and int(args.query) != numQuery:
continue
query = tools.query(evaluate.benchmarks.queries[numQuery-1])
list_warnings = evaluate.get_warning(numQuery)
#print(list_warnings)
list_warnings = {k:v for k,v in list_warnings.items() if len(v) > 0}
df2_warnings = pd.DataFrame(list_warnings, index=['WARNINGS'])
df2_warnings = df2_warnings.T.sort_index()
pd.set_option('display.max_colwidth', None)
if not df2_warnings.empty:
print("===Q{}: {}===".format(numQuery, query.title))
print(df2_warnings)
elif args.mode == 'query':
query = evaluate.get_querystring(int(args.query), args.connection, int(args.num_run))
print(query)