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
|
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Tests for algorithms related to association rules.
"""
import pytest
from efficient_apriori.apriori import apriori
from efficient_apriori.rules import Rule
from efficient_apriori.itemsets import ItemsetCount
def test_api():
transactions = [
("a", "c", "e"),
("a", "c", "e"),
("a", "d", "e"),
("b", "d", "e"),
("b", "d", "f"),
("b", "c", "f"),
("b", "c", "f"),
]
itemsets, rules = apriori(transactions, 0.2, 0.2)
assert itemsets[1] == {("a",): 3, ("c",): 4, ("e",): 4, ("d",): 3, ("b",): 4, ("f",): 3}
assert all(isinstance(rule, Rule) for rule in rules)
for count, itemsets_dict in itemsets.items():
assert isinstance(itemsets_dict, dict)
for itemset, count in itemsets_dict.items():
actual_count = sum(1 if set(itemset).issubset(set(trans)) else 0 for trans in transactions)
assert count == actual_count
itemsets, rules = apriori(transactions, 0.2, 0.2, output_transaction_ids=True)
for count, itemsets_dict in itemsets.items():
assert isinstance(itemsets_dict, dict)
for itemset, counter in itemsets_dict.items():
assert isinstance(counter, ItemsetCount)
actual_count = sum(1 if set(itemset).issubset(set(trans)) else 0 for trans in transactions)
assert counter.itemset_count == actual_count
def test_against_R_implementation_1():
"""
The following R-code was used:
> install.packages("arules")
> col1 = c("a", "a", "a", "b", "b", "b", "b")
> col2 = c("c", "c", "d", "d", "d", "c", "c")
> col3 = c("e", "e", "e", "e", "f", "f", "f")
> df = data.frame(col1, col2, col3)
> df <- data.frame(sapply(df, as.factor))
> rules <- apriori(df, parameter = list(supp = 0.2, conf = 0.2))
> inspect(head(rules, by = "confidence"))
"""
transactions = [
("a", "c", "e"),
("a", "c", "e"),
("a", "d", "e"),
("b", "d", "e"),
("b", "d", "f"),
("b", "c", "f"),
("b", "c", "f"),
]
itemsets, rules = apriori(transactions, 0.2, 0.2)
assert Rule(("a",), ("e",)) in rules
for rule in rules:
if rule == Rule(("a",), ("e",)):
assert abs(rule.support - 0.4285714) < 10e-7
assert rule.confidence == 1
if rule == Rule(("c", "e"), ("a",)):
assert abs(rule.support - 0.2857143) < 10e-7
assert rule.confidence == 1
if rule == Rule(("e",), ("a",)):
assert abs(rule.support - 0.4285714) < 10e-7
assert rule.confidence == 3 / 4
def test_against_R_implementation_2():
"""
The following R-code was used:
> install.packages("arules")
> col1 = c("b", "b", "c", "b", "a", "a", "b", "c", "b", "b", "a", "b", "a",
"a", "a", "c", "b", "a", "b", "b", "b", "c", "a", "c", "a", "a", "c", "a",
"b", "b", "a", "c")
> col2 = c("e", "f", "e", "e", "f", "e", "d", "f", "e", "e", "e", "d", "e",
"e", "f", "d", "d", "d", "e", "f", "f", "d", "d", "f", "e", "e", "f", "f",
"f", "d", "e", "e")
> col3 = c("g", "i", "j", "i", "i", "j", "i", "h", "g", "j", "g", "h", "i",
"h", "g", "h", "g", "j", "h", "i", "g", "g", "i", "h", "h", "h", "h", "g",
"j", "i", "g", "g")
> df = data.frame(col1, col2, col3)
> df <- data.frame(sapply(df, as.factor))
> rules <- apriori(df, parameter = list(supp = 0.2, conf = 0.2))
> inspect(head(rules, by = "confidence"))
"""
transactions = [
("b", "e", "g"),
("b", "f", "i"),
("c", "e", "j"),
("b", "e", "i"),
("a", "f", "i"),
("a", "e", "j"),
("b", "d", "i"),
("c", "f", "h"),
("b", "e", "g"),
("b", "e", "j"),
("a", "e", "g"),
("b", "d", "h"),
("a", "e", "i"),
("a", "e", "h"),
("a", "f", "g"),
("c", "d", "h"),
("b", "d", "g"),
("a", "d", "j"),
("b", "e", "h"),
("b", "f", "i"),
("b", "f", "g"),
("c", "d", "g"),
("a", "d", "i"),
("c", "f", "h"),
("a", "e", "h"),
("a", "e", "h"),
("c", "f", "h"),
("a", "f", "g"),
("b", "f", "j"),
("b", "d", "i"),
("a", "e", "g"),
("c", "e", "g"),
]
itemsets, rules = apriori(transactions, 0.2, 0.2)
for rule in rules:
if rule == Rule(("a",), ("e",)):
assert abs(rule.support - 0.21875) < 10e-7
assert abs(rule.confidence - 0.5833333) < 10e-7
if rule == Rule(("e",), ("a",)):
assert abs(rule.support - 0.21875) < 10e-7
assert abs(rule.confidence - 0.5000000) < 10e-7
def test_against_R_implementation_3():
"""
The following R-code was used:
> install.packages("arules")
> col1 = c("b", "b", "c", "a", "b", "b", "a", "a", "b", "b", "a", "a", "c",
"b", "a", "c")
> col2 = c("e", "d", "e", "e", "e", "e", "d", "e", "e", "e", "d", "e", "e",
"e", "d", "e")
> col3 = c("i", "g", "h", "j", "i", "g", "h", "j", "i", "g", "j", "i", "j",
"j", "i", "i")
> df = data.frame(col1, col2, col3)
> df <- data.frame(sapply(df, as.factor))
> rules <- apriori(df, parameter = list(supp = 0.2, conf = 0.2))
> inspect(head(rules, by = "confidence"))
"""
transactions = [
("b", "e", "i"),
("b", "d", "g"),
("c", "e", "h"),
("a", "e", "j"),
("b", "e", "i"),
("b", "e", "g"),
("a", "d", "h"),
("a", "e", "j"),
("b", "e", "i"),
("b", "e", "g"),
("a", "d", "j"),
("a", "e", "i"),
("c", "e", "j"),
("b", "e", "j"),
("a", "d", "i"),
("c", "e", "i"),
]
itemsets, rules = apriori(transactions, 0.2, 0.2)
for rule in rules:
if rule == Rule(("b",), ("e",)):
assert abs(rule.support - 0.3750) < 10e-7
assert abs(rule.confidence - 0.8571429) < 10e-7
if rule == Rule(("i",), ("e",)):
assert abs(rule.support - 0.3125) < 10e-7
assert abs(rule.confidence - 0.8333333) < 10e-7
if rule == Rule(("j",), ("e",)):
assert abs(rule.support - 0.2500) < 10e-7
assert abs(rule.confidence - 0.8000000) < 10e-7
if rule == Rule(("e",), ("b",)):
assert abs(rule.support - 0.3750) < 10e-7
assert abs(rule.confidence - 0.5000000) < 10e-7
def test_minimal_input():
"""
The with some minimal inputs, and make sure the correct errors are raised.
"""
transactions = []
itemsets, rules = apriori(transactions, 0.2, 0.2)
assert itemsets == {} and rules == []
with pytest.raises(ValueError):
itemsets, rules = apriori(transactions, -0.2, 0.2)
with pytest.raises(ValueError):
itemsets, rules = apriori(transactions, 0.2, -0.2)
with pytest.raises(ValueError):
itemsets, rules = apriori(transactions, "asdf", 1)
itemsets, rules = apriori([(1, 2), (1, 2), (1, 3)], 1, 1)
itemsets, rules = apriori([(1, 2), (1, 2), (1, 3)], 1.0, 1.0)
def test_iterator_input():
"""
Minimal test using transactions from iterators.
"""
empty_iterator = iter(())
transactions = empty_iterator
itemsets, rules = apriori(transactions, 0.2, 0.2)
assert itemsets == {} and rules == []
transactions = [(1, 2), (1, 2), (1, 3), (1, 4), (1, 3)]
transactions_iter = iter(transactions)
itemsets1, rules1 = apriori(transactions_iter, 0.2, 1)
itemsets2, rules2 = apriori(transactions, 0.2, 1)
assert len(rules1) == len(rules2)
for i in range(len(rules1)):
assert rules1[i] == rules2[i]
def test_empty_H_1():
"""
An example of the case where there are itemsets without any Rule with
single item in right hand side that satifies the required minimum confidence.
The issue is raised in #57.
"""
# The results are received from commit 01d174379c51758aa2f6d2926b473124928dc631
true_itemsets_raw = {1: {(1,): 4, (2,): 5, (3,): 4}, 2: {(1, 2): 4, (1, 3): 3, (2, 3): 4}, 3: {(1, 2, 3): 3}}
true_rules = [
Rule((2,), (1,), 4, 5, 4, 5),
Rule((1,), (2,), 4, 4, 5, 5),
Rule((3,), (2,), 4, 4, 5, 5),
Rule((2,), (3,), 4, 5, 4, 5),
Rule((1, 3), (2,), 3, 3, 5, 5),
]
transactions = [(1, 2, 3), (1, 2, 3), (1, 2, 3), (1, 2), (2, 3)]
itemsets_raw, rules = apriori(transactions, 0.4, 0.8)
assert itemsets_raw == true_itemsets_raw
assert all(rule == true_rule for rule, true_rule in zip(rules, true_rules))
if __name__ == "__main__":
pytest.main(args=[".", "--doctest-modules", "-v"])
|
