| line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
|
1
|
|
|
|
|
|
|
package Algorithm::Classifier::NaiveBayes::App::Command::classify; |
|
2
|
|
|
|
|
|
|
|
|
3
|
1
|
|
|
1
|
|
149133
|
use 5.006; |
|
|
1
|
|
|
|
|
3
|
|
|
4
|
1
|
|
|
1
|
|
5
|
use strict; |
|
|
1
|
|
|
|
|
1
|
|
|
|
1
|
|
|
|
|
28
|
|
|
5
|
1
|
|
|
1
|
|
3
|
use warnings; |
|
|
1
|
|
|
|
|
2
|
|
|
|
1
|
|
|
|
|
78
|
|
|
6
|
1
|
|
|
1
|
|
588
|
use Algorithm::Classifier::NaiveBayes (); |
|
|
1
|
|
|
|
|
4
|
|
|
|
1
|
|
|
|
|
38
|
|
|
7
|
1
|
|
|
1
|
|
9
|
use Algorithm::Classifier::NaiveBayes::App -command; |
|
|
1
|
|
|
|
|
1
|
|
|
|
1
|
|
|
|
|
23
|
|
|
8
|
1
|
|
|
1
|
|
453
|
use JSON::PP (); |
|
|
1
|
|
|
|
|
2
|
|
|
|
1
|
|
|
|
|
399
|
|
|
9
|
|
|
|
|
|
|
|
|
10
|
|
|
|
|
|
|
sub options { |
|
11
|
|
|
|
|
|
|
return ( |
|
12
|
4
|
|
|
4
|
0
|
36
|
[ 'm=s', 'Model JSON file path/name.', { 'default' => 'nb_model.json', 'completion' => 'files' } ], |
|
13
|
|
|
|
|
|
|
[ 's', 'Also print the log score of every class.' ], |
|
14
|
|
|
|
|
|
|
[ 'p', 'Also print the probability of every class.' ], |
|
15
|
|
|
|
|
|
|
[ 'json', 'Print the class, scores, and probs as JSON instead.' ], |
|
16
|
|
|
|
|
|
|
); |
|
17
|
|
|
|
|
|
|
} |
|
18
|
|
|
|
|
|
|
|
|
19
|
0
|
|
|
0
|
1
|
0
|
sub abstract { 'Classify the specified text' } |
|
20
|
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
sub description { |
|
22
|
0
|
|
|
0
|
1
|
0
|
return 'Classify the specified text using a saved model. |
|
23
|
|
|
|
|
|
|
|
|
24
|
|
|
|
|
|
|
The text is taken from the remaining args joined by a space, or from |
|
25
|
|
|
|
|
|
|
stdin if no args are given. The best matching class is printed. |
|
26
|
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
nb_tool classify -m model.json cheap pills for sale |
|
28
|
|
|
|
|
|
|
cat some_message.txt | nb_tool classify -m model.json -p |
|
29
|
|
|
|
|
|
|
'; |
|
30
|
|
|
|
|
|
|
} ## end sub description |
|
31
|
|
|
|
|
|
|
|
|
32
|
|
|
|
|
|
|
sub validate { |
|
33
|
4
|
|
|
4
|
0
|
8
|
my ( $self, $opt, $args ) = @_; |
|
34
|
|
|
|
|
|
|
|
|
35
|
4
|
100
|
|
|
|
134
|
if ( !-f $opt->{'m'} ) { |
|
36
|
1
|
|
|
|
|
14
|
$self->usage_error( '-m, "' . $opt->{'m'} . '", is not a file or does not exist' ); |
|
37
|
|
|
|
|
|
|
} |
|
38
|
|
|
|
|
|
|
|
|
39
|
3
|
|
|
|
|
9
|
return 1; |
|
40
|
|
|
|
|
|
|
} |
|
41
|
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
sub execute { |
|
43
|
3
|
|
|
3
|
1
|
13
|
my ( $self, $opt, $args ) = @_; |
|
44
|
|
|
|
|
|
|
|
|
45
|
3
|
|
|
|
|
21
|
my $nb = Algorithm::Classifier::NaiveBayes->new; |
|
46
|
3
|
|
|
|
|
13
|
$nb->load( $opt->{'m'} ); |
|
47
|
|
|
|
|
|
|
|
|
48
|
3
|
|
|
|
|
16
|
my ( $class, $scores, $probs ) = $nb->classify( $self->text_from($args) ); |
|
49
|
3
|
50
|
|
|
|
8
|
if ( !defined($class) ) { |
|
50
|
0
|
|
|
|
|
0
|
die('The model has not been trained yet'); |
|
51
|
|
|
|
|
|
|
} |
|
52
|
|
|
|
|
|
|
|
|
53
|
3
|
100
|
|
|
|
6
|
if ( $opt->{'json'} ) { |
|
54
|
1
|
|
|
|
|
4
|
print JSON::PP->new->canonical->pretty->encode( { 'class' => $class, 'scores' => $scores, 'probs' => $probs } ); |
|
55
|
1
|
|
|
|
|
438
|
return; |
|
56
|
|
|
|
|
|
|
} |
|
57
|
|
|
|
|
|
|
|
|
58
|
2
|
|
|
|
|
13
|
print $class. "\n"; |
|
59
|
|
|
|
|
|
|
|
|
60
|
2
|
100
|
|
|
|
74
|
if ( $opt->{'s'} ) { |
|
61
|
1
|
|
|
|
|
5
|
print "scores:\n"; |
|
62
|
1
|
|
|
|
|
10
|
foreach my $possible ( sort { $scores->{$b} <=> $scores->{$a} } keys %{$scores} ) { |
|
|
1
|
|
|
|
|
5
|
|
|
|
1
|
|
|
|
|
4
|
|
|
63
|
2
|
|
|
|
|
17
|
print ' ' . $possible . ': ' . $scores->{$possible} . "\n"; |
|
64
|
|
|
|
|
|
|
} |
|
65
|
|
|
|
|
|
|
} |
|
66
|
2
|
100
|
|
|
|
38
|
if ( $opt->{'p'} ) { |
|
67
|
1
|
|
|
|
|
2
|
print "probs:\n"; |
|
68
|
1
|
|
|
|
|
8
|
foreach my $possible ( sort { $probs->{$b} <=> $probs->{$a} } keys %{$probs} ) { |
|
|
1
|
|
|
|
|
2
|
|
|
|
1
|
|
|
|
|
3
|
|
|
69
|
2
|
|
|
|
|
25
|
print ' ' . $possible . ': ' . $probs->{$possible} . "\n"; |
|
70
|
|
|
|
|
|
|
} |
|
71
|
|
|
|
|
|
|
} |
|
72
|
|
|
|
|
|
|
} ## end sub execute |
|
73
|
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
1; |