line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package Algorithm::SetSimilarity::Jaccard; |
2
|
|
|
|
|
|
|
|
3
|
1
|
|
|
1
|
|
43858
|
use 5.008005; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
45
|
|
4
|
1
|
|
|
1
|
|
6
|
use strict; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
37
|
|
5
|
1
|
|
|
1
|
|
5
|
use warnings; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
59
|
|
6
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
our $VERSION = "0.0.0_03"; |
8
|
|
|
|
|
|
|
|
9
|
1
|
|
|
1
|
|
7
|
use Scalar::Util; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
1435
|
|
10
|
|
|
|
|
|
|
|
11
|
|
|
|
|
|
|
sub new { |
12
|
9
|
|
|
9
|
0
|
14350
|
my ($class, $param) = @_; |
13
|
9
|
|
|
|
|
18
|
my %hash = (); |
14
|
9
|
50
|
|
|
|
30
|
$hash{"is_sorted"} = $param->{is_sorted} if (exists $param->{is_sorted}); |
15
|
9
|
100
|
|
|
|
25
|
$hash{"data_type"} = $param->{data_type} if (exists $param->{data_type}); |
16
|
9
|
|
|
|
|
38
|
bless \%hash, $class; |
17
|
|
|
|
|
|
|
} |
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
sub estimate_data_type { |
20
|
6
|
|
|
6
|
0
|
1401
|
my ($self, $set1, $set2) = @_; |
21
|
6
|
|
|
|
|
8
|
my $is_estimate = 0; |
22
|
6
|
|
|
|
|
11
|
my ($type1, $type2) = ("string","string"); |
23
|
|
|
|
|
|
|
|
24
|
6
|
100
|
|
|
|
27
|
if (Scalar::Util::looks_like_number($set1->[0])) { |
25
|
2
|
|
|
|
|
3
|
$type1 = "number"; |
26
|
|
|
|
|
|
|
} |
27
|
6
|
100
|
|
|
|
20
|
if (Scalar::Util::looks_like_number($set2->[0])) { |
28
|
2
|
|
|
|
|
2
|
$type2 = "number"; |
29
|
|
|
|
|
|
|
} |
30
|
|
|
|
|
|
|
|
31
|
6
|
100
|
|
|
|
16
|
if ($type1 eq $type2) { |
32
|
4
|
|
|
|
|
15
|
$self->{data_type} = $type1; |
33
|
4
|
|
|
|
|
6
|
$is_estimate = 1; |
34
|
|
|
|
|
|
|
} |
35
|
6
|
|
|
|
|
14
|
return $is_estimate; |
36
|
|
|
|
|
|
|
} |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
sub get_similarity { |
39
|
21
|
|
|
21
|
0
|
8230
|
my ($self, $set1, $set2, $threshold) = @_; |
40
|
21
|
|
|
|
|
26
|
my $score = -1.0; |
41
|
21
|
100
|
33
|
|
|
209
|
if ((ref $set1 eq "ARRAY") && (ref $set2 eq "ARRAY") && (@$set1) && (@$set2)) { |
|
|
|
66
|
|
|
|
|
|
|
|
100
|
|
|
|
|
42
|
18
|
100
|
66
|
|
|
65
|
if ((defined $threshold) && ($threshold > 0.0)) { |
43
|
10
|
|
|
|
|
24
|
$score = $self->filt_by_threshold($set1, $set2, $threshold); |
44
|
|
|
|
|
|
|
} else{ |
45
|
8
|
|
|
|
|
22
|
($set1, $set2) = $self->_swap_set_descending_order($set1, $set2); |
46
|
8
|
|
|
|
|
16
|
my $s1 = $#$set1 + 1; |
47
|
8
|
|
|
|
|
13
|
my $s2 = $#$set2 + 1; |
48
|
8
|
100
|
|
|
|
23
|
my $is_estimate = $self->estimate_data_type($set1, $set2) unless (exists $self->{data_type}); |
49
|
8
|
50
|
66
|
|
|
33
|
if (($is_estimate) || (exists $self->{data_type})) { |
50
|
8
|
50
|
|
|
|
18
|
unless ( $self->{is_sorted} ) { |
51
|
8
|
100
|
|
|
|
19
|
if ($self->{data_type} eq "number") { |
52
|
2
|
|
|
|
|
7
|
@{$set1} = sort {$a <=> $b} @$set1; |
|
2
|
|
|
|
|
6
|
|
|
38
|
|
|
|
|
33
|
|
53
|
2
|
|
|
|
|
4
|
@{$set2} = sort {$a <=> $b} @$set2; |
|
2
|
|
|
|
|
8
|
|
|
38
|
|
|
|
|
38
|
|
54
|
|
|
|
|
|
|
} else { |
55
|
6
|
|
|
|
|
17
|
@{$set1} = sort {$a cmp $b} @$set1; |
|
6
|
|
|
|
|
16
|
|
|
17
|
|
|
|
|
43
|
|
56
|
6
|
|
|
|
|
13
|
@{$set2} = sort {$a cmp $b} @$set2; |
|
6
|
|
|
|
|
14
|
|
|
17
|
|
|
|
|
21
|
|
57
|
|
|
|
|
|
|
} |
58
|
|
|
|
|
|
|
} |
59
|
8
|
|
|
|
|
15
|
my $match_num = 0; |
60
|
8
|
|
66
|
|
|
42
|
for (my ($att1, $att2) = (0, 0); ($att1 < $s2) && ($att2 < $s2);) { |
61
|
45
|
|
|
|
|
46
|
my $judge = -1; |
62
|
45
|
100
|
|
|
|
82
|
if ($self->{data_type} eq "number") { |
63
|
25
|
|
|
|
|
32
|
$judge = ($set1->[$att1] <=> $set2->[$att2]); |
64
|
|
|
|
|
|
|
} else { |
65
|
20
|
|
|
|
|
29
|
$judge = ($set1->[$att1] cmp $set2->[$att2]); |
66
|
|
|
|
|
|
|
} |
67
|
|
|
|
|
|
|
|
68
|
45
|
100
|
|
|
|
73
|
if ($judge == -1) { |
|
|
100
|
|
|
|
|
|
69
|
13
|
|
|
|
|
48
|
$att1++; |
70
|
|
|
|
|
|
|
} elsif ($judge == 1) { |
71
|
8
|
|
|
|
|
28
|
$att2++; |
72
|
|
|
|
|
|
|
} else { |
73
|
24
|
|
|
|
|
28
|
$match_num++; |
74
|
24
|
|
|
|
|
22
|
$att1++; |
75
|
24
|
|
|
|
|
75
|
$att2++; |
76
|
|
|
|
|
|
|
} |
77
|
|
|
|
|
|
|
} |
78
|
8
|
|
|
|
|
28
|
my $diff_num = ($s1 - $match_num) + ($s2 - $match_num); |
79
|
8
|
|
|
|
|
19
|
$score = $match_num / ($match_num + $diff_num); |
80
|
|
|
|
|
|
|
} |
81
|
|
|
|
|
|
|
} |
82
|
|
|
|
|
|
|
} |
83
|
21
|
|
|
|
|
49
|
return $score; |
84
|
|
|
|
|
|
|
} |
85
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
sub _swap_set_ascending_order { |
87
|
2
|
|
|
2
|
|
794
|
my ($self, $set1, $set2) = @_; |
88
|
2
|
100
|
|
|
|
9
|
if ($#$set1 > $#$set2) { |
89
|
1
|
|
|
|
|
2
|
my $tmp_ref = $set1; |
90
|
1
|
|
|
|
|
2
|
$set1 = $set2; |
91
|
1
|
|
|
|
|
2
|
$set2 = $tmp_ref; |
92
|
|
|
|
|
|
|
} |
93
|
2
|
50
|
|
|
|
10
|
return wantarray ? ($set1, $set2) : [$set1, $set2]; |
94
|
|
|
|
|
|
|
} |
95
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
sub _swap_set_descending_order { |
97
|
64
|
|
|
64
|
|
669
|
my ($self, $set1, $set2) = @_; |
98
|
64
|
100
|
|
|
|
152
|
if ($#$set1 < $#$set2) { |
99
|
1
|
|
|
|
|
3
|
my $tmp_ref = $set1; |
100
|
1
|
|
|
|
|
2
|
$set1 = $set2; |
101
|
1
|
|
|
|
|
3
|
$set2 = $tmp_ref; |
102
|
|
|
|
|
|
|
} |
103
|
64
|
50
|
|
|
|
206
|
return wantarray ? ($set1, $set2) : [$set1, $set2]; |
104
|
|
|
|
|
|
|
} |
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
sub filt_by_threshold { |
107
|
55
|
|
|
55
|
0
|
19268
|
my ($self, $set1, $set2, $threshold) = @_; |
108
|
55
|
|
|
|
|
66
|
my $score = -1.0; |
109
|
55
|
100
|
33
|
|
|
682
|
if ((ref $set1 eq "ARRAY") && (ref $set2 eq "ARRAY") && |
|
|
|
33
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
|
66
|
|
|
|
|
110
|
|
|
|
|
|
|
(@$set1) && (@$set2) && |
111
|
|
|
|
|
|
|
($threshold > 0.0) && ($threshold <= 1.0)) { |
112
|
|
|
|
|
|
|
|
113
|
54
|
|
|
|
|
119
|
for(my $r = 0; $r <= 0; $r++) { |
114
|
54
|
|
|
|
|
104
|
($set1, $set2) = $self->_swap_set_descending_order($set1, $set2); |
115
|
54
|
|
|
|
|
88
|
my $s1 = $#$set1 + 1; |
116
|
54
|
|
|
|
|
64
|
my $s2 = $#$set2 + 1; |
117
|
|
|
|
|
|
|
|
118
|
54
|
50
|
|
|
|
153
|
last unless (($s1 * $threshold) <= $s2); #size filtering |
119
|
|
|
|
|
|
|
|
120
|
54
|
100
|
|
|
|
113
|
my $is_estimate = $self->estimate_data_type($set1, $set2) unless (exists $self->{data_type}); |
121
|
54
|
50
|
66
|
|
|
200
|
if (($is_estimate) || (exists $self->{data_type})) { |
122
|
54
|
50
|
|
|
|
113
|
unless ( $self->{is_sorted} ) { |
123
|
54
|
100
|
|
|
|
102
|
if ($self->{data_type} eq "number") { |
124
|
4
|
|
|
|
|
15
|
@{$set1} = sort { $a <=> $b } @$set1; |
|
4
|
|
|
|
|
13
|
|
|
76
|
|
|
|
|
75
|
|
125
|
4
|
|
|
|
|
10
|
@{$set2} = sort { $a <=> $b } @$set2; |
|
4
|
|
|
|
|
11
|
|
|
76
|
|
|
|
|
70
|
|
126
|
|
|
|
|
|
|
} else { |
127
|
50
|
|
|
|
|
122
|
@{$set1} = sort { $a cmp $b} @$set1; |
|
50
|
|
|
|
|
130
|
|
|
102
|
|
|
|
|
141
|
|
128
|
50
|
|
|
|
|
91
|
@{$set2} = sort { $a cmp $b} @$set2; |
|
50
|
|
|
|
|
126
|
|
|
102
|
|
|
|
|
118
|
|
129
|
|
|
|
|
|
|
} |
130
|
|
|
|
|
|
|
} |
131
|
|
|
|
|
|
|
|
132
|
54
|
|
|
|
|
127
|
my $min_overlap = int(($threshold / (1 + $threshold)) * ($s1 + $s2)); |
133
|
54
|
|
|
|
|
68
|
my $alpha = $s2 - ($min_overlap + 1) + 1; |
134
|
54
|
|
|
|
|
60
|
my $match_num = 0; |
135
|
54
|
|
|
|
|
75
|
my ($att1, $att2) = (0, 0); |
136
|
|
|
|
|
|
|
|
137
|
54
|
|
100
|
|
|
203
|
while (($att2 < $alpha) && ($att1 < $s1)) { |
138
|
127
|
|
|
|
|
636
|
my $judge = -1; |
139
|
127
|
100
|
|
|
|
228
|
if ($self->{data_type} eq "number") { |
140
|
31
|
|
|
|
|
38
|
$judge = ($set1->[$att1] <=> $set2->[$att2]); |
141
|
|
|
|
|
|
|
} else { |
142
|
96
|
|
|
|
|
162
|
$judge = ($set1->[$att1] cmp $set2->[$att2]); |
143
|
|
|
|
|
|
|
} |
144
|
|
|
|
|
|
|
|
145
|
127
|
100
|
|
|
|
229
|
if ($judge == -1) { |
|
|
100
|
|
|
|
|
|
146
|
30
|
|
|
|
|
106
|
$att1++; |
147
|
|
|
|
|
|
|
} elsif ($judge == 1) { |
148
|
22
|
|
|
|
|
69
|
$att2++; |
149
|
|
|
|
|
|
|
} else { |
150
|
75
|
|
|
|
|
68
|
$match_num++; |
151
|
75
|
|
|
|
|
61
|
$att1++; |
152
|
75
|
|
|
|
|
197
|
$att2++; |
153
|
|
|
|
|
|
|
} |
154
|
|
|
|
|
|
|
} |
155
|
54
|
|
|
|
|
59
|
my $min = ($s2 - $att2); |
156
|
54
|
100
|
|
|
|
103
|
$min = ($s1 - $att1) if ($min > ($s1 - $att1)); |
157
|
54
|
50
|
|
|
|
90
|
if (($match_num) + $min < $min_overlap) { |
158
|
0
|
|
|
|
|
0
|
$match_num = 0; |
159
|
0
|
|
|
|
|
0
|
last; |
160
|
|
|
|
|
|
|
} |
161
|
|
|
|
|
|
|
|
162
|
54
|
100
|
|
|
|
89
|
last unless ($match_num >= 1); |
163
|
53
|
|
66
|
|
|
172
|
while (($att1 < $s1) && ($att2 < $s2)) { |
164
|
63
|
|
|
|
|
106
|
my $judge = -1; |
165
|
63
|
100
|
|
|
|
112
|
if ($self->{data_type} eq "number") { |
166
|
9
|
|
|
|
|
12
|
$judge = ($set1->[$att1] <=> $set2->[$att2]); |
167
|
|
|
|
|
|
|
} else { |
168
|
54
|
|
|
|
|
79
|
$judge = ($set1->[$att1] cmp $set2->[$att2]); |
169
|
|
|
|
|
|
|
} |
170
|
|
|
|
|
|
|
|
171
|
63
|
100
|
|
|
|
159
|
if ($judge == -1) { |
|
|
100
|
|
|
|
|
|
172
|
5
|
50
|
|
|
|
10
|
last if ($match_num + ($s1 - $att1) < $min_overlap); |
173
|
5
|
|
|
|
|
16
|
$att1++; |
174
|
|
|
|
|
|
|
} elsif ($judge == 1) { |
175
|
8
|
50
|
|
|
|
15
|
last if ($match_num + ($s2 - $att2) < $min_overlap); |
176
|
8
|
|
|
|
|
29
|
$att2++; |
177
|
|
|
|
|
|
|
} else { |
178
|
50
|
|
|
|
|
47
|
$match_num++; |
179
|
50
|
|
|
|
|
43
|
$att1++; |
180
|
50
|
|
|
|
|
133
|
$att2++; |
181
|
|
|
|
|
|
|
} |
182
|
|
|
|
|
|
|
} |
183
|
53
|
100
|
|
|
|
97
|
last unless ($match_num >= $min_overlap + 1); |
184
|
44
|
|
|
|
|
58
|
my $diff_num = ($s1 - $match_num) + ($s2 - $match_num); |
185
|
44
|
|
|
|
|
114
|
$score = $match_num / ($match_num + $diff_num); |
186
|
|
|
|
|
|
|
} |
187
|
|
|
|
|
|
|
} |
188
|
|
|
|
|
|
|
} |
189
|
55
|
|
|
|
|
119
|
return $score; |
190
|
|
|
|
|
|
|
} |
191
|
|
|
|
|
|
|
|
192
|
|
|
|
|
|
|
1; |
193
|
|
|
|
|
|
|
|
194
|
|
|
|
|
|
|
__END__ |