line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
#!/usr/bin/perl |
2
|
|
|
|
|
|
|
|
3
|
|
|
|
|
|
|
# |
4
|
|
|
|
|
|
|
# dbcolscorrelate.pm |
5
|
|
|
|
|
|
|
# Copyright (C) 1998-2015 by John Heidemann |
6
|
|
|
|
|
|
|
# $Id: bdcf1da03251b46bded7f59984e64e7f5060ae46 $ |
7
|
|
|
|
|
|
|
# |
8
|
|
|
|
|
|
|
# This program is distributed under terms of the GNU general |
9
|
|
|
|
|
|
|
# public license, version 2. See the file COPYING |
10
|
|
|
|
|
|
|
# in $dblibdir for details. |
11
|
|
|
|
|
|
|
# |
12
|
|
|
|
|
|
|
|
13
|
|
|
|
|
|
|
package Fsdb::Filter::dbcolscorrelate; |
14
|
|
|
|
|
|
|
|
15
|
|
|
|
|
|
|
=head1 NAME |
16
|
|
|
|
|
|
|
|
17
|
|
|
|
|
|
|
dbcolscorrelate - find the coefficient of correlation over columns |
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
=head1 SYNOPSIS |
20
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
dbcolscorrelate column1 column2 [column3...] |
22
|
|
|
|
|
|
|
|
23
|
|
|
|
|
|
|
=head1 DESCRIPTION |
24
|
|
|
|
|
|
|
|
25
|
|
|
|
|
|
|
Compute the coefficient of correlation over two (or more) columns. |
26
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
The output is one line of correlations. |
28
|
|
|
|
|
|
|
|
29
|
|
|
|
|
|
|
With exactly two columns, a new column I is created. |
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
With more than two columns, correlations are computed for each |
32
|
|
|
|
|
|
|
pairwise combination of rows, and each output column |
33
|
|
|
|
|
|
|
is given a name which is the concatenation of the two source rows, |
34
|
|
|
|
|
|
|
joined with an underscore. |
35
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
By default, we compute the I |
37
|
|
|
|
|
|
|
(usually designed rho, E<0x03c1>) |
38
|
|
|
|
|
|
|
and assume we see all members of the population. |
39
|
|
|
|
|
|
|
With the B<--sample> option we instead compute the |
40
|
|
|
|
|
|
|
I, usually designated I. |
41
|
|
|
|
|
|
|
(Be careful in that the default here to full-population |
42
|
|
|
|
|
|
|
is the I of the default in L.) |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
This program requires a complete copy of the input data on disk. |
45
|
|
|
|
|
|
|
|
46
|
|
|
|
|
|
|
=head1 OPTIONS |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
=over 4 |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
=item B<--sample> |
51
|
|
|
|
|
|
|
|
52
|
|
|
|
|
|
|
Select a the Pearson product-moment correlation coefficient |
53
|
|
|
|
|
|
|
(the "sample correlation coefficient", usually designated I). |
54
|
|
|
|
|
|
|
|
55
|
|
|
|
|
|
|
=item B<--nosample> |
56
|
|
|
|
|
|
|
|
57
|
|
|
|
|
|
|
Select a the Pearson product-moment correlation coefficient |
58
|
|
|
|
|
|
|
(the "sample correlation coefficient", usually designated I). |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
|
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
=item B<-f FORMAT> or B<--format FORMAT> |
63
|
|
|
|
|
|
|
|
64
|
|
|
|
|
|
|
Specify a L-style format for output statistics. |
65
|
|
|
|
|
|
|
Defaults to C<%.5g>. |
66
|
|
|
|
|
|
|
|
67
|
|
|
|
|
|
|
=item B<-T TmpDir> |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
where to put tmp files. |
70
|
|
|
|
|
|
|
Also uses environment variable TMPDIR, if -T is |
71
|
|
|
|
|
|
|
not specified. |
72
|
|
|
|
|
|
|
Default is /tmp. |
73
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
=back |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
=for comment |
77
|
|
|
|
|
|
|
begin_standard_fsdb_options |
78
|
|
|
|
|
|
|
|
79
|
|
|
|
|
|
|
This module also supports the standard fsdb options: |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
=over 4 |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
=item B<-d> |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
Enable debugging output. |
86
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
=item B<-i> or B<--input> InputSource |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
Read from InputSource, typically a file name, or C<-> for standard input, |
90
|
|
|
|
|
|
|
or (if in Perl) a IO::Handle, Fsdb::IO or Fsdb::BoundedQueue objects. |
91
|
|
|
|
|
|
|
|
92
|
|
|
|
|
|
|
=item B<-o> or B<--output> OutputDestination |
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
Write to OutputDestination, typically a file name, or C<-> for standard output, |
95
|
|
|
|
|
|
|
or (if in Perl) a IO::Handle, Fsdb::IO or Fsdb::BoundedQueue objects. |
96
|
|
|
|
|
|
|
|
97
|
|
|
|
|
|
|
=item B<--autorun> or B<--noautorun> |
98
|
|
|
|
|
|
|
|
99
|
|
|
|
|
|
|
By default, programs process automatically, |
100
|
|
|
|
|
|
|
but Fsdb::Filter objects in Perl do not run until you invoke |
101
|
|
|
|
|
|
|
the run() method. |
102
|
|
|
|
|
|
|
The C<--(no)autorun> option controls that behavior within Perl. |
103
|
|
|
|
|
|
|
|
104
|
|
|
|
|
|
|
=item B<--help> |
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
Show help. |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
=item B<--man> |
109
|
|
|
|
|
|
|
|
110
|
|
|
|
|
|
|
Show full manual. |
111
|
|
|
|
|
|
|
|
112
|
|
|
|
|
|
|
=back |
113
|
|
|
|
|
|
|
|
114
|
|
|
|
|
|
|
=for comment |
115
|
|
|
|
|
|
|
end_standard_fsdb_options |
116
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
|
118
|
|
|
|
|
|
|
=head1 SAMPLE USAGE |
119
|
|
|
|
|
|
|
|
120
|
|
|
|
|
|
|
=head2 Input: |
121
|
|
|
|
|
|
|
|
122
|
|
|
|
|
|
|
#fsdb name id test1 test2 |
123
|
|
|
|
|
|
|
a 1 80 81 |
124
|
|
|
|
|
|
|
b 2 70 71 |
125
|
|
|
|
|
|
|
c 3 65 66 |
126
|
|
|
|
|
|
|
d 4 90 91 |
127
|
|
|
|
|
|
|
e 5 70 71 |
128
|
|
|
|
|
|
|
f 6 90 91 |
129
|
|
|
|
|
|
|
|
130
|
|
|
|
|
|
|
=head2 Command: |
131
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
cat DATA/more_grades.fsdb | dbcolscorrelate test1 test2 |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
=head2 Output: |
135
|
|
|
|
|
|
|
|
136
|
|
|
|
|
|
|
#fsdb correlation |
137
|
|
|
|
|
|
|
0.83329 |
138
|
|
|
|
|
|
|
# | dbcolscorrelate test1 test2 |
139
|
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
=head1 SEE ALSO |
142
|
|
|
|
|
|
|
|
143
|
|
|
|
|
|
|
L, |
144
|
|
|
|
|
|
|
L, |
145
|
|
|
|
|
|
|
L. |
146
|
|
|
|
|
|
|
|
147
|
|
|
|
|
|
|
|
148
|
|
|
|
|
|
|
=head1 CLASS FUNCTIONS |
149
|
|
|
|
|
|
|
|
150
|
|
|
|
|
|
|
=cut |
151
|
|
|
|
|
|
|
|
152
|
|
|
|
|
|
|
@ISA = qw(Fsdb::Filter); |
153
|
|
|
|
|
|
|
$VERSION = 2.0; |
154
|
|
|
|
|
|
|
|
155
|
1
|
|
|
1
|
|
7358
|
use strict; |
|
1
|
|
|
|
|
4
|
|
|
1
|
|
|
|
|
40
|
|
156
|
1
|
|
|
1
|
|
7
|
use Pod::Usage; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
118
|
|
157
|
1
|
|
|
1
|
|
8
|
use Carp; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
75
|
|
158
|
|
|
|
|
|
|
|
159
|
1
|
|
|
1
|
|
10
|
use Fsdb::Filter; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
33
|
|
160
|
1
|
|
|
1
|
|
8
|
use Fsdb::IO::Reader; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
29
|
|
161
|
1
|
|
|
1
|
|
6
|
use Fsdb::IO::Writer; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
42
|
|
162
|
1
|
|
|
1
|
|
8
|
use Fsdb::Support qw($is_numeric_regexp); |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
98
|
|
163
|
1
|
|
|
1
|
|
8
|
use Fsdb::Support::NamedTmpfile; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
36
|
|
164
|
1
|
|
|
1
|
|
7
|
use Fsdb::Filter::dbpipeline qw(dbpipeline_open2 dbpipeline_close2_hash dbcolstats); |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
1693
|
|
165
|
|
|
|
|
|
|
|
166
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
=head2 new |
168
|
|
|
|
|
|
|
|
169
|
|
|
|
|
|
|
$filter = new Fsdb::Filter::dbcolscorrelate(@arguments); |
170
|
|
|
|
|
|
|
|
171
|
|
|
|
|
|
|
Create a new dbcolscorrelate object, taking command-line arguments. |
172
|
|
|
|
|
|
|
|
173
|
|
|
|
|
|
|
=cut |
174
|
|
|
|
|
|
|
|
175
|
|
|
|
|
|
|
sub new ($@) { |
176
|
0
|
|
|
0
|
1
|
|
my $class = shift @_; |
177
|
0
|
|
|
|
|
|
my $self = $class->SUPER::new(@_); |
178
|
0
|
|
|
|
|
|
bless $self, $class; |
179
|
0
|
|
|
|
|
|
$self->set_defaults; |
180
|
0
|
|
|
|
|
|
$self->parse_options(@_); |
181
|
0
|
|
|
|
|
|
$self->SUPER::post_new(); |
182
|
0
|
|
|
|
|
|
return $self; |
183
|
|
|
|
|
|
|
} |
184
|
|
|
|
|
|
|
|
185
|
|
|
|
|
|
|
|
186
|
|
|
|
|
|
|
=head2 set_defaults |
187
|
|
|
|
|
|
|
|
188
|
|
|
|
|
|
|
$filter->set_defaults(); |
189
|
|
|
|
|
|
|
|
190
|
|
|
|
|
|
|
Internal: set up defaults. |
191
|
|
|
|
|
|
|
|
192
|
|
|
|
|
|
|
=cut |
193
|
|
|
|
|
|
|
|
194
|
|
|
|
|
|
|
sub set_defaults ($) { |
195
|
0
|
|
|
0
|
1
|
|
my($self) = @_; |
196
|
0
|
|
|
|
|
|
$self->SUPER::set_defaults(); |
197
|
0
|
|
|
|
|
|
$self->{_format} = "%.5g"; |
198
|
0
|
|
|
|
|
|
$self->{_columns} = []; |
199
|
0
|
|
|
|
|
|
$self->{_include_non_numeric} = undef; |
200
|
0
|
|
|
|
|
|
$self->{_sample} = undef; |
201
|
0
|
|
|
|
|
|
$self->{_fscode} = undef; |
202
|
0
|
|
|
|
|
|
$self->set_default_tmpdir; |
203
|
|
|
|
|
|
|
} |
204
|
|
|
|
|
|
|
|
205
|
|
|
|
|
|
|
=head2 parse_options |
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
$filter->parse_options(@ARGV); |
208
|
|
|
|
|
|
|
|
209
|
|
|
|
|
|
|
Internal: parse command-line arguments. |
210
|
|
|
|
|
|
|
|
211
|
|
|
|
|
|
|
=cut |
212
|
|
|
|
|
|
|
|
213
|
|
|
|
|
|
|
sub parse_options ($@) { |
214
|
0
|
|
|
0
|
1
|
|
my $self = shift @_; |
215
|
|
|
|
|
|
|
|
216
|
0
|
|
|
|
|
|
my(@argv) = @_; |
217
|
|
|
|
|
|
|
$self->get_options( |
218
|
|
|
|
|
|
|
\@argv, |
219
|
0
|
|
|
0
|
|
|
'help|?' => sub { pod2usage(1); }, |
220
|
0
|
|
|
0
|
|
|
'man' => sub { pod2usage(-verbose => 2); }, |
221
|
|
|
|
|
|
|
'a|include-non-numeric!' => \$self->{_include_non_numeric}, |
222
|
|
|
|
|
|
|
'autorun!' => \$self->{_autorun}, |
223
|
|
|
|
|
|
|
'd|debug+' => \$self->{_debug}, |
224
|
|
|
|
|
|
|
'f|format=s' => \$self->{_format}, |
225
|
|
|
|
|
|
|
'F|fs|cs|fieldseparator|columnseparator=s' => \$self->{_fscode}, |
226
|
0
|
|
|
0
|
|
|
'i|input=s' => sub { $self->parse_io_option('input', @_); }, |
227
|
|
|
|
|
|
|
'log!' => \$self->{_logprog}, |
228
|
|
|
|
|
|
|
's|sample!' => \$self->{_sample}, |
229
|
|
|
|
|
|
|
'T|tmpdir|tempdir=s' => \$self->{_tmpdir}, |
230
|
0
|
|
|
0
|
|
|
'o|output=s' => sub { $self->parse_io_option('output', @_); }, |
231
|
0
|
0
|
|
|
|
|
) or pod2usage(2); |
232
|
0
|
|
|
|
|
|
push (@{$self->{_columns}}, @argv); |
|
0
|
|
|
|
|
|
|
233
|
|
|
|
|
|
|
} |
234
|
|
|
|
|
|
|
|
235
|
|
|
|
|
|
|
=head2 setup |
236
|
|
|
|
|
|
|
|
237
|
|
|
|
|
|
|
$filter->setup(); |
238
|
|
|
|
|
|
|
|
239
|
|
|
|
|
|
|
Internal: setup, parse headers. |
240
|
|
|
|
|
|
|
|
241
|
|
|
|
|
|
|
=cut |
242
|
|
|
|
|
|
|
|
243
|
|
|
|
|
|
|
sub setup ($) { |
244
|
0
|
|
|
0
|
1
|
|
my($self) = @_; |
245
|
|
|
|
|
|
|
|
246
|
0
|
|
|
|
|
|
$self->finish_io_option('input', -comment_handler => $self->create_pass_comments_sub); |
247
|
|
|
|
|
|
|
|
248
|
|
|
|
|
|
|
croak $self->{_prog} . ": at least two columns must be specified to compute a correlation.\n" |
249
|
0
|
0
|
|
|
|
|
if ($#{$self->{_columns}} < 1); |
|
0
|
|
|
|
|
|
|
250
|
0
|
|
|
|
|
|
my @output_columns; |
251
|
|
|
|
|
|
|
my %columns_processed; |
252
|
0
|
|
|
|
|
|
foreach my $i (0..$#{$self->{_columns}}) { |
|
0
|
|
|
|
|
|
|
253
|
0
|
|
|
|
|
|
my $column = $self->{_columns}[$i]; |
254
|
|
|
|
|
|
|
croak $self->{_prog} . ": column $column is double-listed as an input column (not allowed).\n" |
255
|
0
|
0
|
|
|
|
|
if (defined($columns_processed{$column})); |
256
|
0
|
|
|
|
|
|
$columns_processed{$column} = 1; |
257
|
0
|
|
|
|
|
|
$self->{_colis}[$i] = $self->{_in}->col_to_i($column); |
258
|
|
|
|
|
|
|
croak $self->{_prog} . ": column $column does not exist in the input stream.\n" |
259
|
0
|
0
|
|
|
|
|
if (!defined($self->{_colis}[$i])); |
260
|
0
|
|
|
|
|
|
foreach my $j (0..$#{$self->{_columns}}) { |
|
0
|
|
|
|
|
|
|
261
|
0
|
0
|
|
|
|
|
next if ($i >= $j); |
262
|
0
|
|
|
|
|
|
push(@output_columns, $self->{_columns}[$i] . "_" . $self->{_columns}[$j]); |
263
|
|
|
|
|
|
|
}; |
264
|
|
|
|
|
|
|
}; |
265
|
|
|
|
|
|
|
# if only one column, it has a special name |
266
|
0
|
0
|
|
|
|
|
$output_columns[0] = 'correlation' |
267
|
|
|
|
|
|
|
if ($#output_columns == 0); |
268
|
0
|
|
|
|
|
|
my @output_options = (-cols => \@output_columns); |
269
|
|
|
|
|
|
|
unshift (@output_options, -fscode => $self->{_fscode}) |
270
|
0
|
0
|
|
|
|
|
if (defined($self->{_fscode})); |
271
|
0
|
|
|
|
|
|
$self->finish_io_option('output', @output_options); |
272
|
|
|
|
|
|
|
}; |
273
|
|
|
|
|
|
|
|
274
|
|
|
|
|
|
|
|
275
|
|
|
|
|
|
|
=head2 run |
276
|
|
|
|
|
|
|
|
277
|
|
|
|
|
|
|
$filter->run(); |
278
|
|
|
|
|
|
|
|
279
|
|
|
|
|
|
|
Internal: run over each rows. |
280
|
|
|
|
|
|
|
|
281
|
|
|
|
|
|
|
=cut |
282
|
|
|
|
|
|
|
sub run ($) { |
283
|
0
|
|
|
0
|
1
|
|
my($self) = @_; |
284
|
|
|
|
|
|
|
|
285
|
|
|
|
|
|
|
# |
286
|
|
|
|
|
|
|
# First, read data and save it to a file, |
287
|
|
|
|
|
|
|
# and send each relevant column off to get stats. |
288
|
|
|
|
|
|
|
# |
289
|
0
|
|
|
|
|
|
$self->{_copy_filename} = Fsdb::Support::NamedTmpfile::alloc($self->{_tmpdir}); |
290
|
|
|
|
|
|
|
my $copy_writer = new Fsdb::IO::Writer(-file => $self->{_copy_filename}, |
291
|
0
|
|
|
|
|
|
-clone => $self->{_in}); |
292
|
|
|
|
|
|
|
|
293
|
0
|
|
|
|
|
|
my $read_fastpath_sub = $self->{_in}->fastpath_sub(); |
294
|
0
|
|
|
|
|
|
my $copy_fastpath_sub = $copy_writer->fastpath_sub(); |
295
|
|
|
|
|
|
|
|
296
|
|
|
|
|
|
|
# and take stats |
297
|
0
|
|
|
|
|
|
my(@stats_source_queues); |
298
|
|
|
|
|
|
|
my(@stats_sinks); |
299
|
0
|
|
|
|
|
|
my(@stats_threads); |
300
|
0
|
|
|
|
|
|
my $columns_aref = $self->{_columns}; |
301
|
0
|
|
|
|
|
|
my $colis_aref = $self->{_colis}; |
302
|
0
|
|
|
|
|
|
foreach (0..$#$columns_aref) { |
303
|
|
|
|
|
|
|
($stats_source_queues[$_], $stats_sinks[$_], $stats_threads[$_]) = |
304
|
0
|
0
|
|
|
|
|
dbpipeline_open2([-cols => [qw(data)]], dbcolstats(($self->{_sample} ? '--sample' : '--nosample'), 'data')); |
305
|
|
|
|
|
|
|
} |
306
|
0
|
|
|
|
|
|
my $fref; |
307
|
0
|
|
|
|
|
|
while ($fref = &$read_fastpath_sub()) { |
308
|
|
|
|
|
|
|
# copy and send to stats |
309
|
0
|
|
|
|
|
|
$copy_writer->write_rowobj($fref); |
310
|
|
|
|
|
|
|
# with forking we have to close output explicitly |
311
|
|
|
|
|
|
|
# otherwise we block on the pipe(2) to the subprocesses. |
312
|
0
|
|
|
|
|
|
foreach (0..$#$columns_aref) { |
313
|
0
|
|
|
|
|
|
$stats_sinks[$_]->write_row($fref->[$colis_aref->[$_]]); |
314
|
|
|
|
|
|
|
}; |
315
|
|
|
|
|
|
|
}; |
316
|
|
|
|
|
|
|
# close up both |
317
|
0
|
|
|
|
|
|
$copy_writer->close; |
318
|
0
|
|
|
|
|
|
foreach (0..$#$columns_aref) { |
319
|
0
|
|
|
|
|
|
$stats_sinks[$_]->close; |
320
|
0
|
|
|
|
|
|
$stats_sinks[$_] = undef; |
321
|
|
|
|
|
|
|
}; |
322
|
0
|
|
|
|
|
|
my @means; |
323
|
|
|
|
|
|
|
my @stddevs; |
324
|
0
|
|
|
|
|
|
foreach (0..$#$columns_aref) { |
325
|
0
|
|
|
|
|
|
my $stats_href = dbpipeline_close2_hash($stats_source_queues[$_], $stats_sinks[$_], $stats_threads[$_]); |
326
|
0
|
|
|
|
|
|
$means[$_] = $stats_href->{'mean'}; |
327
|
0
|
0
|
|
|
|
|
croak $self->{_prog} . ": column " . $columns_aref->[$_] . " does not have valid mean.\n" |
328
|
|
|
|
|
|
|
if (!defined($means[$_])); |
329
|
0
|
|
|
|
|
|
$stddevs[$_] = $stats_href->{'stddev'}; |
330
|
0
|
0
|
|
|
|
|
croak $self->{_prog} . ": column " . $columns_aref->[$_] . " does not have valid standard deviation.\n" |
331
|
|
|
|
|
|
|
if (!defined($stddevs[$_])); |
332
|
|
|
|
|
|
|
}; |
333
|
|
|
|
|
|
|
|
334
|
|
|
|
|
|
|
# |
335
|
|
|
|
|
|
|
# Now read back the data, |
336
|
|
|
|
|
|
|
# compute the z-scores on the fly, |
337
|
|
|
|
|
|
|
# and use that to compute the correlation. |
338
|
|
|
|
|
|
|
# |
339
|
0
|
|
|
|
|
|
$self->{_in}->close; |
340
|
|
|
|
|
|
|
$self->{_in} = new Fsdb::IO::Reader(-file => $self->{_copy_filename}, |
341
|
0
|
|
|
|
|
|
-comment_handler => $self->create_pass_comments_sub); |
342
|
0
|
|
|
|
|
|
my $sum_zs; |
343
|
|
|
|
|
|
|
my $ns; |
344
|
0
|
|
|
|
|
|
foreach my $i (0..$#$columns_aref) { |
345
|
0
|
|
|
|
|
|
foreach my $j (0..$#$columns_aref) { |
346
|
0
|
0
|
|
|
|
|
next if ($i >= $j); |
347
|
0
|
|
|
|
|
|
$sum_zs->[$i][$j] = 0; |
348
|
0
|
|
|
|
|
|
$ns->[$i][$j] = 0; |
349
|
|
|
|
|
|
|
}; |
350
|
|
|
|
|
|
|
}; |
351
|
|
|
|
|
|
|
|
352
|
0
|
|
|
|
|
|
$read_fastpath_sub = $self->{_in}->fastpath_sub(); # regenerate with copy stream |
353
|
0
|
|
|
|
|
|
while ($fref = &$read_fastpath_sub()) { |
354
|
0
|
|
|
|
|
|
my @zs; |
355
|
|
|
|
|
|
|
|
356
|
|
|
|
|
|
|
# compute z-score |
357
|
0
|
|
|
|
|
|
foreach (0..$#$columns_aref) { |
358
|
0
|
|
|
|
|
|
my $x = $fref->[$colis_aref->[$_]]; |
359
|
0
|
0
|
|
|
|
|
if ($x !~ /$is_numeric_regexp/) { |
360
|
0
|
|
|
|
|
|
$zs[$_] = undef; |
361
|
|
|
|
|
|
|
} else { |
362
|
0
|
|
|
|
|
|
$zs[$_] = ($x - $means[$_]) / $stddevs[$_]; |
363
|
|
|
|
|
|
|
}; |
364
|
|
|
|
|
|
|
}; |
365
|
|
|
|
|
|
|
|
366
|
|
|
|
|
|
|
# and compute running stats for correlations |
367
|
0
|
|
|
|
|
|
foreach my $i (0..$#$columns_aref) { |
368
|
0
|
|
|
|
|
|
foreach my $j (0..$#$columns_aref) { |
369
|
0
|
0
|
0
|
|
|
|
next if ($i >= $j ||!defined($zs[$i]) || !defined($zs[$j])); |
|
|
|
0
|
|
|
|
|
370
|
0
|
|
|
|
|
|
$sum_zs->[$i][$j] += $zs[$i] * $zs[$j]; |
371
|
0
|
|
|
|
|
|
($ns->[$i][$j])++; |
372
|
|
|
|
|
|
|
}; |
373
|
|
|
|
|
|
|
}; |
374
|
|
|
|
|
|
|
}; |
375
|
|
|
|
|
|
|
|
376
|
|
|
|
|
|
|
# |
377
|
|
|
|
|
|
|
# Finally output the results. |
378
|
|
|
|
|
|
|
# |
379
|
0
|
|
|
|
|
|
my(@correlations); |
380
|
0
|
|
|
|
|
|
foreach my $i (0..$#$columns_aref) { |
381
|
0
|
|
|
|
|
|
foreach my $j (0..$#$columns_aref) { |
382
|
0
|
0
|
|
|
|
|
next if ($i >= $j); |
383
|
0
|
0
|
|
|
|
|
if ($ns->[$i][$j] == 0) { |
384
|
0
|
|
|
|
|
|
push(@correlations, $self->{_empty}); |
385
|
|
|
|
|
|
|
} else { |
386
|
0
|
|
|
|
|
|
my $c = $sum_zs->[$i][$j] / $ns->[$i][$j]; |
387
|
0
|
|
|
|
|
|
push(@correlations, $self->numeric_formatting($c)); |
388
|
|
|
|
|
|
|
}; |
389
|
|
|
|
|
|
|
}; |
390
|
|
|
|
|
|
|
}; |
391
|
0
|
|
|
|
|
|
$self->{_out}->write_row_from_aref(\@correlations); |
392
|
|
|
|
|
|
|
}; |
393
|
|
|
|
|
|
|
|
394
|
|
|
|
|
|
|
=head1 AUTHOR and COPYRIGHT |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
Copyright (C) 1991-2015 by John Heidemann |
397
|
|
|
|
|
|
|
|
398
|
|
|
|
|
|
|
This program is distributed under terms of the GNU general |
399
|
|
|
|
|
|
|
public license, version 2. See the file COPYING |
400
|
|
|
|
|
|
|
with the distribution for details. |
401
|
|
|
|
|
|
|
|
402
|
|
|
|
|
|
|
=cut |
403
|
|
|
|
|
|
|
|
404
|
|
|
|
|
|
|
1; |