The Perl Toolchain Summit 2025 Needs You: You can help 🙏 Learn more

BINARY / Machine-Epsilon-1.0.2 / lib / Machine / Epsilon.pm 

            

              
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

              
package Machine::Epsilon;
use strict;
use warnings;
use Exporter;
our @ISA    = qw(Exporter);
our @EXPORT = qw(machine_epsilon);    ## no critic (ProhibitAutomaticExportation)
our $VERSION = '1.0.2';
=head1 NAME
Machine::Epsilon - The maximum relative error while rounding a floating point number
=head1 SYNOPSIS
    use Machine::Epsilon; # imports machine_epsilon() automatically
    my $epsilon = machine_epsilon();
=cut
=head1 FUNCTIONS
=head2 machine_epsilon
Returns the rounding error for the machine.
=cut
my $epsilon;
sub machine_epsilon {
    return $epsilon if $epsilon;
    # Machine accuracy for 32-bit floating point number
    my $ma_32bit_23mantissa = 1.0 / (2**23);
    # Machine accuracy for 64-bit floating point number
    my $ma_64bit_52mantissa = 1.0 / (2**52);
    # Machine accuracy for 128-bit floating point number (e.g. IBM AIX)
    my $ma_128bit_105mantissa = 1.0 / (2**112);
    # Always start with a power of 2 to avoid roundoff errors!!
    my $e = 1.0;
    while (1) {
        if (1.0 + $e / 2 == 1.0) { last; }
        $e = $e / 2.0;
        # Accuracy already better than a 128-bit machine!!
        if ($e < $ma_128bit_105mantissa) {
            warn "Machine accuracy seems too good to be true!! "
                . "Do we have such a powerful machine? Assuming that "
                . "something isn't right, and returning machine accuracy "
                . "for 64 bit double.";
            $e = $ma_64bit_52mantissa;
            last;
        }
    }
    # If accuracy is very bad, we return the minimum accuracy for a 32-bit double
    if ($e > $ma_32bit_23mantissa) {
        warn "Machine accuracy ($e greater than $ma_32bit_23mantissa) "
            . "seems worse than the primitive 32-bit double representation. "
            . "Setting to minimum accuracy of $ma_32bit_23mantissa.";
        return $ma_32bit_23mantissa;
    }
    $epsilon = $e;
    return $e;
}
=head1 AUTHOR
binary.com, C<< <perl at binary.com> >>
=head1 BUGS
Please report any bugs or feature requests to C<bug-machine-epsilon at rt.cpan.org>, or through
the web interface at L<http://rt.cpan.org/NoAuth/ReportBug.html?Queue=Machine-Epsilon>.  I will be notified, and then you'll
automatically be notified of progress on your bug as I make changes.
=head1 SUPPORT
You can find documentation for this module with the perldoc command.
    perldoc Machine::Epsilon
You can also look for information at:
=over 4
=item * RT: CPAN's request tracker (report bugs here)
L<http://rt.cpan.org/NoAuth/Bugs.html?Dist=Machine-Epsilon>
=item * AnnoCPAN: Annotated CPAN documentation
L<http://annocpan.org/dist/Machine-Epsilon>
=item * CPAN Ratings
L<http://cpanratings.perl.org/d/Machine-Epsilon>
=item * Search CPAN
L<http://search.cpan.org/dist/Machine-Epsilon/>
=back
=head1 REFERENCES
    http://en.wikipedia.org/wiki/Machine_epsilon
=head1 LICENSE AND COPYRIGHT
Copyright 2014 binary.com.
This program is free software; you can redistribute it and/or modify it
under the terms of either: the GNU General Public License as published
by the Free Software Foundation; or the Artistic License.
See L<http://dev.perl.org/licenses/> for more information.
=cut
1;    # End of Machine::Epsilon