line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
=head1 NAME |
2
|
|
|
|
|
|
|
|
3
|
|
|
|
|
|
|
Statistics::Smoothing::SGT - A Simple Good-Turing (SGT) smoothing implementation |
4
|
|
|
|
|
|
|
|
5
|
|
|
|
|
|
|
=head1 SYNOPSIS |
6
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
=head2 Basic Usage |
8
|
|
|
|
|
|
|
|
9
|
|
|
|
|
|
|
use Statistics::Smoothing::SGT |
10
|
|
|
|
|
|
|
my $sgt = new Statistics::Smoothing::SGT($frequencyClasses, $total); |
11
|
|
|
|
|
|
|
$sgt->calculateValues(); |
12
|
|
|
|
|
|
|
$probabilities = $sgt->getProbabilities(); |
13
|
|
|
|
|
|
|
$newFrequencies = $sgt->getNewFrequencies(); |
14
|
|
|
|
|
|
|
$nBar = $sgt->getNBar(); |
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
=head1 AUTHORS |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
Florian Doemges, florian@doemges.net |
19
|
|
|
|
|
|
|
|
20
|
|
|
|
|
|
|
Bjoern Wilmsmann, bjoern@wilmsmann.de |
21
|
|
|
|
|
|
|
|
22
|
|
|
|
|
|
|
=head1 COPYRIGHT |
23
|
|
|
|
|
|
|
|
24
|
|
|
|
|
|
|
Copyright (C) 2006, Florian Doemges and Bjoern Wilmsmann, |
25
|
|
|
|
|
|
|
Department of Linguistics, Ruhr-University, Bochum |
26
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
Partially based on the SGT module (Copyright (C) 2004) by Andre Halama |
28
|
|
|
|
|
|
|
(halama@linguistics.rub.de) and Tibor Kiss (tibor@linguistics.rub.de), |
29
|
|
|
|
|
|
|
Department of Linguistics, Ruhr-University, Bochum. |
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
This module in turn was based on the work (including an implementation |
32
|
|
|
|
|
|
|
of the algorithm in C) by |
33
|
|
|
|
|
|
|
Geoffrey Sampson, Department of Informatics, University of Sussex. |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify |
36
|
|
|
|
|
|
|
it under the terms of the GNU General Public License as published by |
37
|
|
|
|
|
|
|
the Free Software Foundation; either version 2 of the License, or (at |
38
|
|
|
|
|
|
|
your option) any later version. |
39
|
|
|
|
|
|
|
|
40
|
|
|
|
|
|
|
This program is distributed in the hope that it will be useful, but |
41
|
|
|
|
|
|
|
WITHOUT ANY WARRANTY; without even the implied warranty of |
42
|
|
|
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
43
|
|
|
|
|
|
|
General Public License for more details. |
44
|
|
|
|
|
|
|
|
45
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License |
46
|
|
|
|
|
|
|
along with this program; if not, write to |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
The Free Software Foundation, Inc., |
49
|
|
|
|
|
|
|
59 Temple Place - Suite 330, |
50
|
|
|
|
|
|
|
Boston, MA 02111-1307, USA. |
51
|
|
|
|
|
|
|
|
52
|
|
|
|
|
|
|
Note: a copy of the GNU General Public License is available on the web |
53
|
|
|
|
|
|
|
at L and is included in this |
54
|
|
|
|
|
|
|
distribution under the name GPL. |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
=head1 BUGS |
57
|
|
|
|
|
|
|
|
58
|
|
|
|
|
|
|
=head1 SEE ALSO |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
=head1 DESCRIPTION |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
This Perl module implements the Simple Good Turing (SGT) algorithm |
63
|
|
|
|
|
|
|
for smoothing of probabilistic values developed by William Gale and |
64
|
|
|
|
|
|
|
Geoffrey Sampson. |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
The algorithm is described in detail in Sampson's Empirical Linguistics |
67
|
|
|
|
|
|
|
(Continuum International, London and New York, 2001), chapter 7. |
68
|
|
|
|
|
|
|
An online version of this paper is available at Geoffrey Sampson's |
69
|
|
|
|
|
|
|
homepage under L. |
70
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
=head2 Error Codes |
72
|
|
|
|
|
|
|
|
73
|
|
|
|
|
|
|
=head2 Methods |
74
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
=over |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
=cut |
78
|
|
|
|
|
|
|
|
79
|
|
|
|
|
|
|
|
80
|
|
|
|
|
|
|
package Statistics::Smoothing::SGT; |
81
|
|
|
|
|
|
|
|
82
|
|
|
|
|
|
|
# use strict, as we do not our variables to go haywire |
83
|
1
|
|
|
1
|
|
302199
|
use strict; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
72
|
|
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
# use these for debugging purposes |
86
|
1
|
|
|
1
|
|
6
|
use warnings; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
36
|
|
87
|
1
|
|
|
1
|
|
1246
|
use Data::Dumper; |
|
1
|
|
|
|
|
13863
|
|
|
1
|
|
|
|
|
1514
|
|
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
our ($VERSION); |
90
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
$VERSION = '2.1.2'; |
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
# constructor |
94
|
|
|
|
|
|
|
sub new { |
95
|
1
|
|
|
1
|
0
|
24
|
my ($class, $frequencyClasses, $ngramTotal) = @_; |
96
|
1
|
|
|
|
|
3
|
my $rowsFound = scalar(keys(%{$frequencyClasses})); |
|
1
|
|
|
|
|
5
|
|
97
|
1
|
50
|
|
|
|
4
|
unless ($rowsFound >=5) { |
98
|
0
|
|
|
|
|
0
|
die("At least 5 m/V(m) pairs must be provided for SGT.\nThe hash contains only $rowsFound rows.\n"); |
99
|
|
|
|
|
|
|
} |
100
|
1
|
|
|
|
|
5
|
my $self = { |
101
|
|
|
|
|
|
|
frequencyClasses => $frequencyClasses, |
102
|
|
|
|
|
|
|
ngramTotal => $ngramTotal |
103
|
|
|
|
|
|
|
}; |
104
|
1
|
|
|
|
|
4
|
bless($self, $class); |
105
|
1
|
|
|
|
|
3
|
return $self; |
106
|
|
|
|
|
|
|
} |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
# method for calculating all Z(m) |
109
|
|
|
|
|
|
|
sub calculateZ() { |
110
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
111
|
0
|
|
|
|
|
|
my @zValues; |
112
|
|
|
|
|
|
|
my @equivalenceClasses; |
113
|
0
|
|
|
|
|
|
my @cardinalities; |
114
|
|
|
|
|
|
|
|
115
|
|
|
|
|
|
|
# iterate over frequencies for conversion into ascendingly ordered list |
116
|
0
|
|
|
|
|
|
foreach my $frequency (sort {$a <=> $b} keys(%{$self->{frequencyClasses}})) { |
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
# push frequency to array of equivalence classes |
118
|
0
|
|
|
|
|
|
push(@equivalenceClasses, $frequency); |
119
|
|
|
|
|
|
|
|
120
|
|
|
|
|
|
|
# push cardinality of this frequency class to array of cardinalities at corresponding |
121
|
|
|
|
|
|
|
# position |
122
|
0
|
|
|
|
|
|
push(@cardinalities, $self->{frequencyClasses}->{$frequency}); |
123
|
|
|
|
|
|
|
} |
124
|
|
|
|
|
|
|
|
125
|
|
|
|
|
|
|
# save arrays for processing in other functions |
126
|
0
|
|
|
|
|
|
$self->{equivalenceClasses} = \@equivalenceClasses; |
127
|
0
|
|
|
|
|
|
$self->{cardinalities} = \@cardinalities; |
128
|
|
|
|
|
|
|
|
129
|
|
|
|
|
|
|
# calculate z value for m = 1 |
130
|
0
|
|
|
|
|
|
push(@zValues, 2 * $self->{cardinalities}->[0] / ($self->{equivalenceClasses}->[1])); |
131
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
# iterate over equivalence classes and their respective cardinalities |
133
|
|
|
|
|
|
|
# for calculating z values for all m > 1 and m < max |
134
|
0
|
|
|
|
|
|
for (my $i = 1; $i < @{$self->{equivalenceClasses}} - 1; $i++) { |
|
0
|
|
|
|
|
|
|
135
|
0
|
|
|
|
|
|
push(@zValues, 2 * $self->{cardinalities}->[$i] / ($self->{equivalenceClasses}->[$i + 1] - $self->{equivalenceClasses}->[$i - 1])); |
136
|
|
|
|
|
|
|
} |
137
|
|
|
|
|
|
|
|
138
|
|
|
|
|
|
|
# calculate z value for m = max |
139
|
0
|
|
|
|
|
|
push(@zValues, 2 * $self->{cardinalities}->[@{$self->{cardinalities}} - 1] / ($self->{equivalenceClasses}->[@{$self->{equivalenceClasses}} - 1] - $self->{equivalenceClasses}->[@{$self->{equivalenceClasses}} - 2])); |
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
# return |
142
|
0
|
|
|
|
|
|
return \@zValues; |
143
|
|
|
|
|
|
|
} |
144
|
|
|
|
|
|
|
|
145
|
|
|
|
|
|
|
# method for getting first gap between frequencies |
146
|
|
|
|
|
|
|
sub getGap { |
147
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
148
|
0
|
|
|
|
|
|
my $gapAt; |
149
|
0
|
|
|
|
|
|
my $buffer = 0; |
150
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
# iterate over equivalence classes |
152
|
0
|
|
|
|
|
|
for (my $i = 0; $i <= @{$self->{equivalenceClasses}}; $i++) { |
|
0
|
|
|
|
|
|
|
153
|
|
|
|
|
|
|
# check if gap between current value and previous |
154
|
|
|
|
|
|
|
# one is bigger than 1 |
155
|
0
|
0
|
0
|
|
|
|
if (!defined $self->{equivalenceClasses}->[$i] || $buffer + 1 < $self->{equivalenceClasses}->[$i]) { |
156
|
|
|
|
|
|
|
# if so, we have found our gap and can skip the |
157
|
|
|
|
|
|
|
# remaining iterations |
158
|
0
|
|
|
|
|
|
$gapAt = $buffer; |
159
|
0
|
|
|
|
|
|
last; |
160
|
|
|
|
|
|
|
} |
161
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
# write current value to buffer for checking next value |
163
|
0
|
|
|
|
|
|
$buffer = $self->{equivalenceClasses}->[$i]; |
164
|
|
|
|
|
|
|
} |
165
|
|
|
|
|
|
|
|
166
|
|
|
|
|
|
|
# return gap |
167
|
0
|
|
|
|
|
|
return $gapAt; |
168
|
|
|
|
|
|
|
} |
169
|
|
|
|
|
|
|
|
170
|
|
|
|
|
|
|
# method for deducing slope and intersection of a logarithmised |
171
|
|
|
|
|
|
|
# exponential function with its best fitting linear function |
172
|
|
|
|
|
|
|
# (i.e. linear regression) |
173
|
|
|
|
|
|
|
sub logBestFit { |
174
|
0
|
|
|
0
|
0
|
|
my ($self, $Xaxis, $Yaxis) = @_; |
175
|
0
|
|
|
|
|
|
my $XYs = 0; |
176
|
0
|
|
|
|
|
|
my $Xsquares = 0; |
177
|
0
|
|
|
|
|
|
my $meanX = 0; |
178
|
0
|
|
|
|
|
|
my $meanY = 0; |
179
|
0
|
|
|
|
|
|
my $rows = scalar @{$Xaxis}; |
|
0
|
|
|
|
|
|
|
180
|
|
|
|
|
|
|
|
181
|
|
|
|
|
|
|
# calculate log mean values for x and y |
182
|
0
|
|
|
|
|
|
for (my $i = 0; $i < $rows; $i++) { |
183
|
0
|
|
|
|
|
|
$meanX += log($Xaxis->[$i]); |
184
|
0
|
|
|
|
|
|
$meanY += log($Yaxis->[$i]); |
185
|
|
|
|
|
|
|
} |
186
|
0
|
|
|
|
|
|
$meanX /= $rows; |
187
|
0
|
|
|
|
|
|
$meanY /= $rows; |
188
|
|
|
|
|
|
|
|
189
|
|
|
|
|
|
|
# calculate slope and intersection |
190
|
0
|
|
|
|
|
|
for (my $i = 0; $i < $rows; $i++) { |
191
|
0
|
|
|
|
|
|
$XYs += ((log($Xaxis->[$i]) - $meanX) * (log($Yaxis->[$i]) - $meanY)); |
192
|
0
|
|
|
|
|
|
$Xsquares += (log($Xaxis->[$i]) - $meanX) ** 2; |
193
|
|
|
|
|
|
|
} |
194
|
0
|
|
|
|
|
|
my $slope = $XYs / $Xsquares; |
195
|
0
|
|
|
|
|
|
my $intersection = $meanY - $slope * $meanX; |
196
|
|
|
|
|
|
|
|
197
|
|
|
|
|
|
|
# return slope and intersection |
198
|
0
|
|
|
|
|
|
return ($slope, $intersection); |
199
|
|
|
|
|
|
|
} |
200
|
|
|
|
|
|
|
|
201
|
|
|
|
|
|
|
# public method for calculating SGT-smoothed values |
202
|
|
|
|
|
|
|
sub calculateValues() { |
203
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
204
|
0
|
|
|
|
|
|
my $gapAt; |
205
|
|
|
|
|
|
|
my $slope; |
206
|
0
|
|
|
|
|
|
my $intersection; |
207
|
0
|
|
|
|
|
|
my %xValues; |
208
|
0
|
|
|
|
|
|
my %yValues; |
209
|
0
|
|
|
|
|
|
my $zValues; |
210
|
0
|
|
|
|
|
|
my %mFound; |
211
|
0
|
|
|
|
|
|
my %mStar; |
212
|
0
|
|
|
|
|
|
my $newFrequencies; |
213
|
0
|
|
|
|
|
|
my $m; |
214
|
0
|
|
|
|
|
|
my $xFlag; |
215
|
0
|
|
|
|
|
|
my $diff; |
216
|
0
|
|
|
|
|
|
my $i; |
217
|
|
|
|
|
|
|
|
218
|
|
|
|
|
|
|
# get z values |
219
|
0
|
|
|
|
|
|
$zValues = $self->calculateZ(); |
220
|
|
|
|
|
|
|
|
221
|
|
|
|
|
|
|
# get first frequency gap |
222
|
0
|
|
|
|
|
|
$gapAt = $self->getGap(); |
223
|
|
|
|
|
|
|
|
224
|
|
|
|
|
|
|
# perform linear regression |
225
|
0
|
|
|
|
|
|
($slope, $intersection) = $self->logBestFit($self->{equivalenceClasses}, $zValues); |
226
|
|
|
|
|
|
|
|
227
|
|
|
|
|
|
|
# calculcate y function values |
228
|
0
|
|
|
|
|
|
for ($i = 0; $i <= $self->{equivalenceClasses}->[@{$self->{equivalenceClasses}} - 1]; $i++) { |
|
0
|
|
|
|
|
|
|
229
|
0
|
|
|
|
|
|
$yValues{$i} = ($i + 1) * exp($slope * (log($i + 2) - log($i + 1))); |
230
|
|
|
|
|
|
|
} |
231
|
|
|
|
|
|
|
|
232
|
|
|
|
|
|
|
# calculate x value only, if last element has not yet |
233
|
|
|
|
|
|
|
# been reached, x values for large classes do not matter |
234
|
|
|
|
|
|
|
# anyway |
235
|
0
|
|
|
|
|
|
for($i = 0; $i < @{$self->{equivalenceClasses}} - 2; $i++) { |
|
0
|
|
|
|
|
|
|
236
|
0
|
|
|
|
|
|
$xValues{$self->{equivalenceClasses}->[$i]} = $self->{equivalenceClasses}->[$i + 1] * $self->{cardinalities}->[$i + 1] / $self->{cardinalities}->[$i]; |
237
|
|
|
|
|
|
|
} |
238
|
|
|
|
|
|
|
|
239
|
|
|
|
|
|
|
# build hash for all existing m (=equivalence class) and V(m) (= cardinality) |
240
|
0
|
|
|
|
|
|
for ($i = 0; $i < @{$self->{equivalenceClasses}}; $i++){ |
|
0
|
|
|
|
|
|
|
241
|
0
|
|
|
|
|
|
$mFound{$self->{equivalenceClasses}->[$i]} = $self->{cardinalities}->[$i]; |
242
|
|
|
|
|
|
|
} |
243
|
|
|
|
|
|
|
|
244
|
|
|
|
|
|
|
# iterate over equivalence classes in order to determine |
245
|
|
|
|
|
|
|
# which function to use |
246
|
0
|
|
|
|
|
|
for ($i = 1; $i <= $self->{equivalenceClasses}->[@{$self->{equivalenceClasses}} - 1]; $i++) { |
|
0
|
|
|
|
|
|
|
247
|
|
|
|
|
|
|
# check, if m (=equivalence class) and V(m) (= cardinality) exist for this index |
248
|
0
|
0
|
|
|
|
|
unless ($mFound{$i}) { |
249
|
0
|
|
|
|
|
|
next; |
250
|
|
|
|
|
|
|
} |
251
|
|
|
|
|
|
|
|
252
|
|
|
|
|
|
|
# if y function has not been used yet and first gap between frequencies |
253
|
|
|
|
|
|
|
# has not yet been reached |
254
|
0
|
0
|
0
|
|
|
|
if (!$xFlag && $i < $gapAt) { |
255
|
|
|
|
|
|
|
# calculate distance value between x and y function for which y |
256
|
|
|
|
|
|
|
# function is to be used |
257
|
0
|
|
|
|
|
|
$diff = (sqrt (($self->{equivalenceClasses}->[$i + 1] ** 2) * |
258
|
|
|
|
|
|
|
($self->{cardinalities}->[$i + 1] / $self->{cardinalities}->[$i] ** 2) * |
259
|
|
|
|
|
|
|
(1 + $self->{cardinalities}->[$i + 1] / $self->{cardinalities}->[$i]))) * 1.65; |
260
|
|
|
|
|
|
|
|
261
|
|
|
|
|
|
|
# if difference between x and y value is smaller than the difference |
262
|
|
|
|
|
|
|
# value calculated above |
263
|
0
|
0
|
|
|
|
|
if (abs($xValues{$i} - $yValues{$i}) < $diff) { |
264
|
|
|
|
|
|
|
# use y function from now on |
265
|
0
|
|
|
|
|
|
$mStar{$i} = $yValues{$i}; |
266
|
0
|
|
|
|
|
|
$xFlag = 1; |
267
|
|
|
|
|
|
|
} else { |
268
|
|
|
|
|
|
|
# use x function |
269
|
0
|
|
|
|
|
|
$mStar{$i} = $xValues{$i}; |
270
|
|
|
|
|
|
|
} |
271
|
|
|
|
|
|
|
} else { |
272
|
|
|
|
|
|
|
# use y function from now on |
273
|
0
|
|
|
|
|
|
$mStar{$i} = $yValues{$i}; |
274
|
0
|
|
|
|
|
|
$xFlag = 1; |
275
|
|
|
|
|
|
|
} |
276
|
|
|
|
|
|
|
|
277
|
|
|
|
|
|
|
# calculate new total (i.e. nBar); |
278
|
0
|
|
|
|
|
|
$self->{nBar} += $mStar{$i} * $mFound{$i}; |
279
|
|
|
|
|
|
|
} |
280
|
|
|
|
|
|
|
|
281
|
|
|
|
|
|
|
# save new frequencies |
282
|
0
|
|
|
|
|
|
$self->{newFrequencies} = \%mStar; |
283
|
|
|
|
|
|
|
|
284
|
|
|
|
|
|
|
# calculate the probability for unseen events (i.e. pZero) |
285
|
0
|
|
|
|
|
|
$self->{probabilities}->{0} = $self->{cardinalities}->[0] / $self->{ngramTotal}; |
286
|
|
|
|
|
|
|
|
287
|
|
|
|
|
|
|
# calculate all other probabilities |
288
|
0
|
|
|
|
|
|
foreach $m (keys(%mStar)) { |
289
|
0
|
|
|
|
|
|
$self->{probabilities}->{$m} = (1 - $self->{probabilities}->{0}) * ($mStar{$m} / $self->{nBar}); |
290
|
|
|
|
|
|
|
} |
291
|
|
|
|
|
|
|
} |
292
|
|
|
|
|
|
|
|
293
|
|
|
|
|
|
|
# public method for getting new frequencies |
294
|
|
|
|
|
|
|
sub getNewFrequencies() { |
295
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
296
|
|
|
|
|
|
|
|
297
|
|
|
|
|
|
|
# return equivalence classes |
298
|
0
|
|
|
|
|
|
return $self->{newFrequencies}; |
299
|
|
|
|
|
|
|
} |
300
|
|
|
|
|
|
|
|
301
|
|
|
|
|
|
|
# public method for getting equivalence classes |
302
|
|
|
|
|
|
|
sub getEquivalenceClasses() { |
303
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
304
|
|
|
|
|
|
|
|
305
|
|
|
|
|
|
|
# return equivalence classes |
306
|
0
|
|
|
|
|
|
return $self->{equivalenceClasses}; |
307
|
|
|
|
|
|
|
} |
308
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
# public method for getting probabilities |
310
|
|
|
|
|
|
|
sub getProbabilities() { |
311
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
312
|
|
|
|
|
|
|
|
313
|
|
|
|
|
|
|
# return probabilities |
314
|
0
|
|
|
|
|
|
return $self->{probabilities}; |
315
|
|
|
|
|
|
|
} |
316
|
|
|
|
|
|
|
|
317
|
|
|
|
|
|
|
# public method for getting nBar |
318
|
|
|
|
|
|
|
sub getNBar() { |
319
|
0
|
|
|
0
|
0
|
|
my ($self) = @_; |
320
|
|
|
|
|
|
|
|
321
|
|
|
|
|
|
|
# return nBar |
322
|
0
|
|
|
|
|
|
return $self->{nBar}; |
323
|
|
|
|
|
|
|
} |
324
|
|
|
|
|
|
|
|
325
|
|
|
|
|
|
|
1; |
326
|
|
|
|
|
|
|
|
327
|
|
|
|
|
|
|
__END__ |