| line | stmt | bran | cond | sub | pod | time | code | 
| 1 |  |  |  |  |  |  | package AI::NeuralNet::Simple; | 
| 2 |  |  |  |  |  |  |  | 
| 3 | 3 |  |  | 3 |  | 109722 | use Log::Agent; | 
|  | 3 |  |  |  |  | 25169 |  | 
|  | 3 |  |  |  |  | 308 |  | 
| 4 |  |  |  |  |  |  |  | 
| 5 | 3 |  |  | 3 |  | 20 | use strict; | 
|  | 3 |  |  |  |  | 6 |  | 
|  | 3 |  |  |  |  | 86 |  | 
| 6 |  |  |  |  |  |  |  | 
| 7 | 3 |  |  | 3 |  | 13 | use vars qw( $REVISION $VERSION @ISA ); | 
|  | 3 |  |  |  |  | 9 |  | 
|  | 3 |  |  |  |  | 2883 |  | 
| 8 |  |  |  |  |  |  |  | 
| 9 |  |  |  |  |  |  | $REVISION = '$Id: Simple.pm,v 1.3 2004/01/31 20:34:11 ovid Exp $'; | 
| 10 |  |  |  |  |  |  | $VERSION  = '0.11'; | 
| 11 |  |  |  |  |  |  |  | 
| 12 |  |  |  |  |  |  | if ( $] >= 5.006 ) { | 
| 13 |  |  |  |  |  |  | require XSLoader; | 
| 14 |  |  |  |  |  |  | XSLoader::load( 'AI::NeuralNet::Simple', $VERSION ); | 
| 15 |  |  |  |  |  |  | } | 
| 16 |  |  |  |  |  |  | else { | 
| 17 |  |  |  |  |  |  | require DynaLoader; | 
| 18 |  |  |  |  |  |  | push @ISA, 'DynaLoader'; | 
| 19 |  |  |  |  |  |  | AI::NeuralNet::Simple->bootstrap($VERSION); | 
| 20 |  |  |  |  |  |  | } | 
| 21 |  |  |  |  |  |  |  | 
| 22 | 80042 |  |  | 80042 | 0 | 385810 | sub handle { $_[0]->{handle} } | 
| 23 |  |  |  |  |  |  |  | 
| 24 |  |  |  |  |  |  | sub new { | 
| 25 | 7 |  |  | 7 | 1 | 9927 | my ( $class, @args ) = @_; | 
| 26 | 7 | 100 |  |  |  | 39 | logdie "you must supply three positive integers to new()" | 
| 27 |  |  |  |  |  |  | unless 3 == @args; | 
| 28 | 6 |  |  |  |  | 14 | foreach (@args) { | 
| 29 | 16 | 100 | 66 |  |  | 165 | logdie "arguments to new() must be positive integers" | 
| 30 |  |  |  |  |  |  | unless defined $_ && /^\d+$/; | 
| 31 |  |  |  |  |  |  | } | 
| 32 | 5 |  |  |  |  | 162 | my $seed = rand(1);    # Perl invokes srand() on first call to rand() | 
| 33 | 5 |  |  |  |  | 89 | my $handle = c_new_network(@args); | 
| 34 | 5 | 50 |  |  |  | 20 | logdie "could not create new network" unless $handle >= 0; | 
| 35 | 5 |  |  |  |  | 39 | my $self = bless { | 
| 36 |  |  |  |  |  |  | input  => $args[0], | 
| 37 |  |  |  |  |  |  | hidden => $args[1], | 
| 38 |  |  |  |  |  |  | output => $args[2], | 
| 39 |  |  |  |  |  |  | handle => $handle, | 
| 40 |  |  |  |  |  |  | }, $class; | 
| 41 | 5 |  |  |  |  | 22 | $self->iterations(10000);    # set a reasonable default | 
| 42 |  |  |  |  |  |  | } | 
| 43 |  |  |  |  |  |  |  | 
| 44 |  |  |  |  |  |  | sub train { | 
| 45 | 80001 |  |  | 80001 | 1 | 253700 | my ( $self, $inputref, $outputref ) = @_; | 
| 46 | 80001 |  |  |  |  | 125972 | return c_train( $self->handle, $inputref, $outputref ); | 
| 47 |  |  |  |  |  |  | } | 
| 48 |  |  |  |  |  |  |  | 
| 49 |  |  |  |  |  |  | sub train_set { | 
| 50 | 1 |  |  | 1 | 1 | 23 | my ( $self, $set, $iterations, $mse ) = @_; | 
| 51 | 1 |  | 33 |  |  | 4 | $iterations ||= $self->iterations; | 
| 52 | 1 | 50 |  |  |  | 2 | $mse = -1.0 unless defined $mse; | 
| 53 | 1 |  |  |  |  | 3 | return c_train_set( $self->handle, $set, $iterations, $mse ); | 
| 54 |  |  |  |  |  |  | } | 
| 55 |  |  |  |  |  |  |  | 
| 56 |  |  |  |  |  |  | sub iterations { | 
| 57 | 5 |  |  | 5 | 1 | 10 | my ( $self, $iterations ) = @_; | 
| 58 | 5 | 50 |  |  |  | 24 | if ( defined $iterations ) { | 
| 59 | 5 | 50 | 33 |  |  | 48 | logdie "iterations() value must be a positive integer." | 
| 60 |  |  |  |  |  |  | unless $iterations | 
| 61 |  |  |  |  |  |  | and $iterations =~ /^\d+$/; | 
| 62 | 5 |  |  |  |  | 27 | $self->{iterations} = $iterations; | 
| 63 | 5 |  |  |  |  | 26 | return $self; | 
| 64 |  |  |  |  |  |  | } | 
| 65 | 0 |  |  |  |  | 0 | $self->{iterations}; | 
| 66 |  |  |  |  |  |  | } | 
| 67 |  |  |  |  |  |  |  | 
| 68 |  |  |  |  |  |  | sub delta { | 
| 69 | 3 |  |  | 3 | 1 | 1076 | my ( $self, $delta ) = @_; | 
| 70 | 3 | 100 |  |  |  | 12 | return c_get_delta( $self->handle )              unless defined $delta; | 
| 71 | 2 | 50 |  |  |  | 6 | logdie "delta() value must be a positive number" unless $delta > 0.0; | 
| 72 | 2 |  |  |  |  | 8 | c_set_delta( $self->handle, $delta ); | 
| 73 | 2 |  |  |  |  | 4 | return $self; | 
| 74 |  |  |  |  |  |  | } | 
| 75 |  |  |  |  |  |  |  | 
| 76 |  |  |  |  |  |  | sub use_bipolar { | 
| 77 | 3 |  |  | 3 | 1 | 10 | my ( $self, $bipolar ) = @_; | 
| 78 | 3 | 100 |  |  |  | 12 | return c_get_use_bipolar( $self->handle ) unless defined $bipolar; | 
| 79 | 2 |  |  |  |  | 6 | c_set_use_bipolar( $self->handle, $bipolar ); | 
| 80 | 2 |  |  |  |  | 17 | return $self; | 
| 81 |  |  |  |  |  |  | } | 
| 82 |  |  |  |  |  |  |  | 
| 83 |  |  |  |  |  |  | sub infer { | 
| 84 | 0 |  |  | 0 | 1 | 0 | my ( $self, $data ) = @_; | 
| 85 | 0 |  |  |  |  | 0 | c_infer( $self->handle, $data ); | 
| 86 |  |  |  |  |  |  | } | 
| 87 |  |  |  |  |  |  |  | 
| 88 |  |  |  |  |  |  | sub winner { | 
| 89 |  |  |  |  |  |  |  | 
| 90 |  |  |  |  |  |  | # returns index of largest value in inferred answer | 
| 91 | 16 |  |  | 16 | 1 | 951 | my ( $self, $data ) = @_; | 
| 92 | 16 |  |  |  |  | 49 | my $arrayref = c_infer( $self->handle, $data ); | 
| 93 |  |  |  |  |  |  |  | 
| 94 | 16 |  |  |  |  | 22 | my $largest = 0; | 
| 95 | 16 |  |  |  |  | 56 | for ( 0 .. $#$arrayref ) { | 
| 96 | 32 | 100 |  |  |  | 108 | $largest = $_ if $arrayref->[$_] > $arrayref->[$largest]; | 
| 97 |  |  |  |  |  |  | } | 
| 98 | 16 |  |  |  |  | 85 | return $largest; | 
| 99 |  |  |  |  |  |  | } | 
| 100 |  |  |  |  |  |  |  | 
| 101 |  |  |  |  |  |  | sub learn_rate { | 
| 102 | 12 |  |  | 12 | 1 | 4536 | my ( $self, $rate ) = @_; | 
| 103 | 12 | 100 |  |  |  | 46 | return c_get_learn_rate( $self->handle ) unless defined $rate; | 
| 104 | 6 | 100 | 66 |  |  | 73 | logdie "learn rate must be between 0 and 1, exclusive" | 
| 105 |  |  |  |  |  |  | unless $rate > 0 && $rate < 1; | 
| 106 | 5 |  |  |  |  | 13 | c_set_learn_rate( $self->handle, $rate ); | 
| 107 | 5 |  |  |  |  | 25 | return $self; | 
| 108 |  |  |  |  |  |  | } | 
| 109 |  |  |  |  |  |  |  | 
| 110 |  |  |  |  |  |  | sub DESTROY { | 
| 111 | 6 |  |  | 6 |  | 1633 | my $self = shift; | 
| 112 | 6 |  |  |  |  | 19 | c_destroy_network( $self->handle ); | 
| 113 |  |  |  |  |  |  | } | 
| 114 |  |  |  |  |  |  |  | 
| 115 |  |  |  |  |  |  | # | 
| 116 |  |  |  |  |  |  | # Serializing hook for Storable | 
| 117 |  |  |  |  |  |  | # | 
| 118 |  |  |  |  |  |  |  | 
| 119 |  |  |  |  |  |  | sub STORABLE_freeze { | 
| 120 | 1 |  |  | 1 | 0 | 203 | my ( $self, $cloning ) = @_; | 
| 121 | 1 |  |  |  |  | 3 | my $internal = c_export_network( $self->handle ); | 
| 122 |  |  |  |  |  |  |  | 
| 123 |  |  |  |  |  |  | # This is an excellent example where "we know better" than | 
| 124 |  |  |  |  |  |  | # the recommended way in Storable's man page... | 
| 125 |  |  |  |  |  |  | # Behaviour is the same whether we're cloning or not --RAM | 
| 126 |  |  |  |  |  |  |  | 
| 127 | 1 |  |  |  |  | 8 | my %copy = %$self; | 
| 128 | 1 |  |  |  |  | 3 | delete $copy{handle}; | 
| 129 |  |  |  |  |  |  |  | 
| 130 | 1 |  |  |  |  | 118 | return ( "", \%copy, $internal ); | 
| 131 |  |  |  |  |  |  | } | 
| 132 |  |  |  |  |  |  |  | 
| 133 |  |  |  |  |  |  | # | 
| 134 |  |  |  |  |  |  | # Deserializing hook for Storable | 
| 135 |  |  |  |  |  |  | # | 
| 136 |  |  |  |  |  |  | sub STORABLE_thaw { | 
| 137 | 1 |  |  | 1 | 0 | 116 | my ( $self, $cloning, $x, $copy, $internal ) = @_; | 
| 138 | 1 |  |  |  |  | 7 | %$self = %$copy; | 
| 139 | 1 |  |  |  |  | 20 | $self->{handle} = c_import_network($internal); | 
| 140 |  |  |  |  |  |  | } | 
| 141 |  |  |  |  |  |  |  | 
| 142 |  |  |  |  |  |  | 1; | 
| 143 |  |  |  |  |  |  |  | 
| 144 |  |  |  |  |  |  | __END__ |