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
|
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Tests for algorithms related to association rules.
"""
import pytest
import os
from efficient_apriori.apriori import apriori
from efficient_apriori.rules import Rule
def test_adult_dataset():
"""
Test on the Adult dataset, which may be found here:
https://archive.ics.uci.edu/ml/datasets/adult
Some numeric columns were removed. The age was discretized.
The purpose of this test is to assure that the algorithm can deal with a
small 2.2 MB (30k rows) data set reasonably efficiently.
Test against R, from the following code
> library(arules)
> df <- read.csv("adult_data_cleaned.txt", header = FALSE)
> df <- data.frame(sapply(df, as.factor))
> rules <- apriori(df, parameter = list(supp = 0.4, conf = 0.4))
> inspect(head(rules, by = "confidence"m = 10))
"""
def transactions_from_file(filename):
with open(filename) as file:
for line in file:
yield tuple(k.strip() for k in line.split(","))
try:
base, _ = os.path.split(__file__)
filename = os.path.join(base, "adult_data_cleaned.txt")
except NameError:
filename = "adult_data_cleaned.txt"
transactions = transactions_from_file(filename)
itemsets, rules = apriori(transactions, min_support=0.2, min_confidence=0.2)
# Test that the rules found in R were also found using this implementation
rules_set = set(rules)
assert Rule(("Married-civ-spouse", "Husband", "middle-aged"), ("Male",)) in rules_set
assert (
Rule(
("Married-civ-spouse", "White", "middle-aged", "Male"),
("Husband",),
)
in rules_set
)
assert Rule(("<=50K", "young"), ("Never-married",)) in rules_set
assert (
Rule(
("Husband", "White", "Male", "middle-aged"),
("Married-civ-spouse",),
)
in rules_set
)
assert Rule(("young",), ("Never-married",)) in rules_set
# Test results against R package arules
for rule in rules:
if rule == Rule(("Married-civ-spouse", "Husband", "middle-aged"), ("Male",)):
assert abs(rule.support - 0.2356193) < 10e-7
assert abs(rule.confidence - 0.9998697) < 10e-7
assert abs(rule.lift - 1.494115) < 10e-7
if rule == Rule(
("Married-civ-spouse", "White", "middle-aged", "Male"),
("Husband",),
):
assert abs(rule.support - 0.2123399) < 10e-7
assert abs(rule.confidence - 0.9938192) < 10e-7
assert abs(rule.lift - 2.452797) < 10e-7
if rule == Rule(("<=50K", "young"), ("Never-married",)):
assert abs(rule.support - 0.2170081) < 10e-7
assert abs(rule.confidence - 0.7680435) < 10e-7
assert abs(rule.lift - 2.340940) < 10e-7
if rule == Rule(
("Husband", "White", "Male", "middle-aged"),
("Married-civ-spouse",),
):
assert abs(rule.support - 0.2123399) < 10e-7
assert abs(rule.confidence - 0.9995663) < 10e-7
assert abs(rule.lift - 2.173269) < 10e-7
if rule == Rule(("young",), ("Never-married",)):
assert abs(rule.support - 0.2200792) < 10e-7
assert abs(rule.confidence - 0.7379261) < 10e-7
assert abs(rule.lift - 2.249144) < 10e-7
if __name__ == "__main__":
pytest.main(args=[".", "--doctest-modules", "-v"])
|
