line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
=head1 NAME |
2
|
|
|
|
|
|
|
|
3
|
|
|
|
|
|
|
Text::NSP::Measures::2D::Fisher - Perl module that provides methods |
4
|
|
|
|
|
|
|
to compute the Fishers exact tests. |
5
|
|
|
|
|
|
|
|
6
|
|
|
|
|
|
|
=head1 SYNOPSIS |
7
|
|
|
|
|
|
|
|
8
|
|
|
|
|
|
|
=head3 Basic Usage |
9
|
|
|
|
|
|
|
|
10
|
|
|
|
|
|
|
use Text::NSP::Measures::2D::Fisher::left; |
11
|
|
|
|
|
|
|
|
12
|
|
|
|
|
|
|
my $npp = 60; my $n1p = 20; my $np1 = 20; my $n11 = 10; |
13
|
|
|
|
|
|
|
|
14
|
|
|
|
|
|
|
$left_value = calculateStatistic( n11=>$n11, |
15
|
|
|
|
|
|
|
n1p=>$n1p, |
16
|
|
|
|
|
|
|
np1=>$np1, |
17
|
|
|
|
|
|
|
npp=>$npp); |
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
if( ($errorCode = getErrorCode())) |
20
|
|
|
|
|
|
|
{ |
21
|
|
|
|
|
|
|
print STDERR $errorCode." - ".getErrorMessage(); |
22
|
|
|
|
|
|
|
} |
23
|
|
|
|
|
|
|
else |
24
|
|
|
|
|
|
|
{ |
25
|
|
|
|
|
|
|
print getStatisticName."value for bigram is ".$left_value; |
26
|
|
|
|
|
|
|
} |
27
|
|
|
|
|
|
|
|
28
|
|
|
|
|
|
|
|
29
|
|
|
|
|
|
|
=head1 DESCRIPTION |
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
Assume that the frequency count data associated with a bigram |
32
|
|
|
|
|
|
|
is stored in a 2x2 contingency table: |
33
|
|
|
|
|
|
|
|
34
|
|
|
|
|
|
|
word2 ~word2 |
35
|
|
|
|
|
|
|
word1 n11 n12 | n1p |
36
|
|
|
|
|
|
|
~word1 n21 n22 | n2p |
37
|
|
|
|
|
|
|
-------------- |
38
|
|
|
|
|
|
|
np1 np2 npp |
39
|
|
|
|
|
|
|
|
40
|
|
|
|
|
|
|
where n11 is the number of times occur together, and |
41
|
|
|
|
|
|
|
n12 is the number of times occurs with some word other than |
42
|
|
|
|
|
|
|
word2, and n1p is the number of times in total that word1 occurs as |
43
|
|
|
|
|
|
|
the first word in a bigram. |
44
|
|
|
|
|
|
|
|
45
|
|
|
|
|
|
|
The fishers exact tests are calculated by fixing the marginal totals |
46
|
|
|
|
|
|
|
and computing the hypergeometric probabilities for all the possible |
47
|
|
|
|
|
|
|
contingency tables, |
48
|
|
|
|
|
|
|
|
49
|
|
|
|
|
|
|
A left sided test is calculated by adding the probabilities of all |
50
|
|
|
|
|
|
|
the possible two by two contingency tables formed by fixing the |
51
|
|
|
|
|
|
|
marginal totals and changing the value of n11 to less than the given |
52
|
|
|
|
|
|
|
value. A left sided Fisher's Exact Test tells us how likely it is to |
53
|
|
|
|
|
|
|
randomly sample a table where n11 is less than observed. In other words, |
54
|
|
|
|
|
|
|
it tells us how likely it is to sample an observation where the two words |
55
|
|
|
|
|
|
|
are less dependent than currently observed. |
56
|
|
|
|
|
|
|
|
57
|
|
|
|
|
|
|
A right sided test is calculated by adding the probabilities of all |
58
|
|
|
|
|
|
|
the possible two by two contingency tables formed by fixing the |
59
|
|
|
|
|
|
|
marginal totals and changing the value of n11 to greater than or |
60
|
|
|
|
|
|
|
equal to the given value. A right sided Fisher's Exact Test tells us |
61
|
|
|
|
|
|
|
how likely it is to randomly sample a table where n11 is greater |
62
|
|
|
|
|
|
|
than observed. In other words, it tells us how likely it is to sample |
63
|
|
|
|
|
|
|
an observation where the two words are more dependent than currently |
64
|
|
|
|
|
|
|
observed. |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
A two-tailed fishers test is calculated by adding the probabilities of |
67
|
|
|
|
|
|
|
all the contingency tables with probabilities less than the probability |
68
|
|
|
|
|
|
|
of the observed table. The two-tailed fishers test tells us how likely |
69
|
|
|
|
|
|
|
it would be to observe an contingency table which is less probable than |
70
|
|
|
|
|
|
|
the current table. |
71
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
=head2 Methods |
73
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
=over |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
=cut |
77
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
|
79
|
|
|
|
|
|
|
package Text::NSP::Measures::2D::Fisher; |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
|
82
|
4
|
|
|
4
|
|
2981
|
use Text::NSP::Measures::2D; |
|
4
|
|
|
|
|
10
|
|
|
4
|
|
|
|
|
775
|
|
83
|
4
|
|
|
4
|
|
18
|
use strict; |
|
4
|
|
|
|
|
6
|
|
|
4
|
|
|
|
|
76
|
|
84
|
4
|
|
|
4
|
|
15
|
use Carp; |
|
4
|
|
|
|
|
7
|
|
|
4
|
|
|
|
|
172
|
|
85
|
4
|
|
|
4
|
|
25
|
use warnings; |
|
4
|
|
|
|
|
14
|
|
|
4
|
|
|
|
|
3133
|
|
86
|
|
|
|
|
|
|
# use subs(calculateStatistic); |
87
|
|
|
|
|
|
|
require Exporter; |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
our ($VERSION, @EXPORT, @ISA); |
90
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
@ISA = qw(Exporter); |
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
@EXPORT = qw(initializeStatistic calculateStatistic |
94
|
|
|
|
|
|
|
getErrorCode getErrorMessage getStatisticName |
95
|
|
|
|
|
|
|
$n11 $n12 $n21 $n22 $m11 $m12 $m21 $m22 |
96
|
|
|
|
|
|
|
$npp $np1 $np2 $n2p $n1p $errorCodeNumber |
97
|
|
|
|
|
|
|
$errorMessage); |
98
|
|
|
|
|
|
|
|
99
|
|
|
|
|
|
|
$VERSION = '0.97'; |
100
|
|
|
|
|
|
|
|
101
|
|
|
|
|
|
|
|
102
|
|
|
|
|
|
|
=item getValues() -This method calls the |
103
|
|
|
|
|
|
|
computeObservedValues() and the computeExpectedValues() methods to |
104
|
|
|
|
|
|
|
compute the observed and marginal total values. It checks these values |
105
|
|
|
|
|
|
|
for any errors that might cause the Fishers Exact test measures to |
106
|
|
|
|
|
|
|
fail. |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
INPUT PARAMS : $count_values .. Reference of an array containing |
109
|
|
|
|
|
|
|
the count values computed by the |
110
|
|
|
|
|
|
|
count.pl program. |
111
|
|
|
|
|
|
|
|
112
|
|
|
|
|
|
|
RETURN VALUES : 1/undef ..returns '1' to indicate success |
113
|
|
|
|
|
|
|
and an undefined(NULL) value to indicate |
114
|
|
|
|
|
|
|
failure. |
115
|
|
|
|
|
|
|
|
116
|
|
|
|
|
|
|
=cut |
117
|
|
|
|
|
|
|
|
118
|
|
|
|
|
|
|
sub getValues |
119
|
|
|
|
|
|
|
{ |
120
|
36
|
|
|
36
|
1
|
42
|
my $values = shift; |
121
|
|
|
|
|
|
|
|
122
|
|
|
|
|
|
|
# computes and returns the marginal totals from the frequency |
123
|
|
|
|
|
|
|
# combination values. returns undef if there is an error in |
124
|
|
|
|
|
|
|
# the computation or the values are inconsistent. |
125
|
36
|
100
|
|
|
|
88
|
if(!(Text::NSP::Measures::2D::computeMarginalTotals($values)) ){ |
126
|
15
|
|
|
|
|
43
|
return; |
127
|
|
|
|
|
|
|
} |
128
|
|
|
|
|
|
|
|
129
|
|
|
|
|
|
|
# computes and returns the observed and marginal values from |
130
|
|
|
|
|
|
|
# the frequency combination values. returns 0 if there is an |
131
|
|
|
|
|
|
|
# error in the computation or the values are inconsistent. |
132
|
21
|
100
|
|
|
|
53
|
if( !(Text::NSP::Measures::2D::computeObservedValues($values)) ) { |
133
|
15
|
|
|
|
|
41
|
return; |
134
|
|
|
|
|
|
|
} |
135
|
|
|
|
|
|
|
|
136
|
6
|
|
|
|
|
20
|
return 1; |
137
|
|
|
|
|
|
|
} |
138
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
=item computeDistribution() - This method calculates the probabilities |
141
|
|
|
|
|
|
|
for all the possible tables |
142
|
|
|
|
|
|
|
|
143
|
|
|
|
|
|
|
INPUT PARAMS : $n11_start .. the value for the cell 1,1 in the first contingency |
144
|
|
|
|
|
|
|
table |
145
|
|
|
|
|
|
|
$final_limit .. the value of cell 1,1 in the last contingency table |
146
|
|
|
|
|
|
|
for which we have to compute the probability. |
147
|
|
|
|
|
|
|
|
148
|
|
|
|
|
|
|
RETURN VALUES : $probability .. Reference to a hash containing hypergeometric |
149
|
|
|
|
|
|
|
probabilities for all the possible contingency |
150
|
|
|
|
|
|
|
tables |
151
|
|
|
|
|
|
|
|
152
|
|
|
|
|
|
|
=cut |
153
|
|
|
|
|
|
|
|
154
|
|
|
|
|
|
|
sub computeDistribution |
155
|
|
|
|
|
|
|
{ |
156
|
6
|
|
|
6
|
1
|
8
|
my $n11_start = shift @_; |
157
|
6
|
|
|
|
|
12
|
my $final_limit = shift @_; |
158
|
|
|
|
|
|
|
|
159
|
|
|
|
|
|
|
# first sort the numerator array in the descending order. |
160
|
6
|
|
|
|
|
70
|
my @numerator = sort { $b <=> $a } ($n1p, $np1, $n2p, $np2); |
|
24
|
|
|
|
|
45
|
|
161
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
# initialize the hash to store the probability distribution values. |
163
|
6
|
|
|
|
|
9
|
my %probability = (); |
164
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
# declare some temporary variables for use in loops and computing the values. |
166
|
6
|
|
|
|
|
13
|
my $i; |
167
|
6
|
|
|
|
|
10
|
my $j=0; |
168
|
|
|
|
|
|
|
|
169
|
|
|
|
|
|
|
# initialize the product variable to be used in the probability computation. |
170
|
6
|
|
|
|
|
7
|
my $product = 0; |
171
|
|
|
|
|
|
|
|
172
|
|
|
|
|
|
|
# set the values for the first contingency table. |
173
|
6
|
|
|
|
|
10
|
$n11 = $n11_start; |
174
|
6
|
|
|
|
|
11
|
$n12 = $n1p-$n11; |
175
|
6
|
|
|
|
|
9
|
$n21 = $np1-$n11; |
176
|
6
|
|
|
|
|
10
|
$n22 = $n2p - $n21; |
177
|
|
|
|
|
|
|
|
178
|
6
|
|
|
|
|
17
|
while($n22 < 0) |
179
|
|
|
|
|
|
|
{ |
180
|
0
|
|
|
|
|
0
|
$n11++; |
181
|
0
|
|
|
|
|
0
|
$n12 = $n1p - $n11; |
182
|
0
|
|
|
|
|
0
|
$n21 = $np1 - $n11; |
183
|
0
|
|
|
|
|
0
|
$n22 = $n2p - $n21; |
184
|
|
|
|
|
|
|
} |
185
|
|
|
|
|
|
|
|
186
|
|
|
|
|
|
|
# declare the denominator array. |
187
|
6
|
|
|
|
|
11
|
my @denominator = (); |
188
|
|
|
|
|
|
|
|
189
|
6
|
|
|
|
|
9
|
$product = 0; |
190
|
|
|
|
|
|
|
|
191
|
6
|
|
|
|
|
8
|
my $prob = 0; |
192
|
|
|
|
|
|
|
|
193
|
6
|
|
|
|
|
8
|
$i = $n11; |
194
|
6
|
|
|
|
|
10
|
$n12 = $n1p - $i; |
195
|
6
|
|
|
|
|
8
|
$n21 = $np1 - $i; |
196
|
6
|
|
|
|
|
8
|
$n22 = $n2p - $n21; |
197
|
|
|
|
|
|
|
|
198
|
|
|
|
|
|
|
# initialize the denominator array with values sorted in the descending order. |
199
|
6
|
|
|
|
|
17
|
@denominator = sort { $b <=> $a } ($npp, $n22, $n12, $n21, $i); |
|
51
|
|
|
|
|
72
|
|
200
|
|
|
|
|
|
|
|
201
|
|
|
|
|
|
|
#decalare other variables for use in computation. |
202
|
6
|
|
|
|
|
17
|
my @dLimits = (); |
203
|
6
|
|
|
|
|
8
|
my @nLimits = (); |
204
|
6
|
|
|
|
|
9
|
my $dIndex = 0; |
205
|
6
|
|
|
|
|
8
|
my $nIndex = 0; |
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
# set the dLimits and nLimits arrays to be used in the cancellation of factorials |
208
|
|
|
|
|
|
|
# and to be used in the computation of factorial. |
209
|
|
|
|
|
|
|
# the dLimits and the nLimits allow us to cancel out factorials in the numerator |
210
|
|
|
|
|
|
|
# and the denominator. for example: |
211
|
|
|
|
|
|
|
# 6! 1*2*3*4*5*6 |
212
|
|
|
|
|
|
|
# --- = --------------- = 5*6 |
213
|
|
|
|
|
|
|
# 4! 1*2*3*4 |
214
|
|
|
|
|
|
|
# |
215
|
|
|
|
|
|
|
# we achieve this by defining a range within which all the |
216
|
|
|
|
|
|
|
# nos must be multiplied. So every pair of entries in the nLimits array defines a range |
217
|
|
|
|
|
|
|
# so for the above case the entries would be: |
218
|
|
|
|
|
|
|
# 5,6 |
219
|
|
|
|
|
|
|
# |
220
|
6
|
|
|
|
|
19
|
for ( $j = 0; $j < 4; $j++ ) |
221
|
|
|
|
|
|
|
{ |
222
|
24
|
100
|
|
|
|
77
|
if ( $numerator[$j] > $denominator[$j] ) |
|
|
100
|
|
|
|
|
|
223
|
|
|
|
|
|
|
{ |
224
|
6
|
|
|
|
|
11
|
$nLimits[$nIndex] = $denominator[$j] + 1; |
225
|
6
|
|
|
|
|
31
|
$nLimits[$nIndex+1] = $numerator[$j]; |
226
|
6
|
|
|
|
|
17
|
$nIndex += 2; |
227
|
|
|
|
|
|
|
} |
228
|
|
|
|
|
|
|
elsif ( $denominator[$j] > $numerator[$j] ) |
229
|
|
|
|
|
|
|
{ |
230
|
6
|
|
|
|
|
13
|
$dLimits[$dIndex] = $numerator[$j] + 1; |
231
|
6
|
|
|
|
|
12
|
$dLimits[$dIndex+1] = $denominator[$j]; |
232
|
6
|
|
|
|
|
16
|
$dIndex += 2; |
233
|
|
|
|
|
|
|
} |
234
|
|
|
|
|
|
|
} |
235
|
6
|
|
|
|
|
13
|
$dLimits[$dIndex] = 1; |
236
|
6
|
|
|
|
|
12
|
$dLimits[$dIndex+1] = $denominator[4]; |
237
|
|
|
|
|
|
|
|
238
|
|
|
|
|
|
|
# since, all the variables have been initialized, we start the computations. |
239
|
6
|
|
|
|
|
27
|
$product = computeHyperGeometric(\@dLimits, \@nLimits); |
240
|
6
|
|
|
|
|
18
|
$probability{$i} = $product; |
241
|
6
|
|
|
|
|
10
|
$prob = $probability{$i}; |
242
|
|
|
|
|
|
|
|
243
|
|
|
|
|
|
|
# to reduce the no. of computations and the make the measure more efficient |
244
|
|
|
|
|
|
|
# we use the previous tables probabilities to compute the new tables probabilities |
245
|
|
|
|
|
|
|
# we can do this because the counts in the table will change by only a factor of 1 |
246
|
|
|
|
|
|
|
# thus instead of repeating all those multiplications we have to perform only |
247
|
|
|
|
|
|
|
# 4 multiplications. |
248
|
6
|
|
|
|
|
8
|
my $subproduct = 0; |
249
|
|
|
|
|
|
|
|
250
|
6
|
|
|
|
|
17
|
for ($i = $n11+1; $i <= $final_limit; $i++ ) |
251
|
|
|
|
|
|
|
{ |
252
|
49
|
|
|
|
|
61
|
$subproduct += log $n12; |
253
|
49
|
|
|
|
|
55
|
$n22++; |
254
|
49
|
|
|
|
|
55
|
$subproduct -= log $n22; |
255
|
49
|
|
|
|
|
59
|
$subproduct += log $n21; |
256
|
49
|
|
|
|
|
44
|
$n12--; |
257
|
49
|
|
|
|
|
46
|
$n21--; |
258
|
49
|
|
|
|
|
58
|
$subproduct -= log $i; |
259
|
49
|
|
|
|
|
98
|
$probability{$i} = $product+$subproduct; |
260
|
49
|
50
|
|
|
|
101
|
if($probability{$i} != 0) |
261
|
|
|
|
|
|
|
{ |
262
|
49
|
|
|
|
|
48
|
$product = $product+$subproduct; |
263
|
49
|
|
|
|
|
102
|
$subproduct=0; |
264
|
|
|
|
|
|
|
} |
265
|
|
|
|
|
|
|
} |
266
|
|
|
|
|
|
|
|
267
|
|
|
|
|
|
|
|
268
|
6
|
|
|
|
|
30
|
return (\%probability); |
269
|
|
|
|
|
|
|
} |
270
|
|
|
|
|
|
|
|
271
|
|
|
|
|
|
|
|
272
|
|
|
|
|
|
|
|
273
|
|
|
|
|
|
|
sub computeHyperGeometric |
274
|
|
|
|
|
|
|
{ |
275
|
6
|
|
|
6
|
0
|
10
|
my $dLimits = shift @_; |
276
|
6
|
|
|
|
|
10
|
my $nLimits = shift @_; |
277
|
6
|
|
|
|
|
9
|
my $product = 0; |
278
|
|
|
|
|
|
|
|
279
|
|
|
|
|
|
|
# compute the probability now, since all the variables have been initialized. |
280
|
6
|
|
|
|
|
21
|
while ( defined ( $nLimits->[0] ) ) |
281
|
|
|
|
|
|
|
{ |
282
|
6
|
|
|
|
|
23
|
while ( defined ( $nLimits->[0] ) ) |
283
|
|
|
|
|
|
|
{ |
284
|
90
|
|
|
|
|
128
|
$product += log $nLimits->[0]; |
285
|
90
|
|
|
|
|
87
|
$nLimits->[0]++; |
286
|
90
|
100
|
|
|
|
224
|
if ( $nLimits->[0] > $nLimits->[1] ) |
287
|
|
|
|
|
|
|
{ |
288
|
6
|
|
|
|
|
7
|
shift @{$nLimits}; |
|
6
|
|
|
|
|
49
|
|
289
|
6
|
|
|
|
|
9
|
shift @{$nLimits}; |
|
6
|
|
|
|
|
19
|
|
290
|
|
|
|
|
|
|
} |
291
|
|
|
|
|
|
|
} |
292
|
6
|
|
|
|
|
14
|
while ( defined ( $dLimits->[0] ) ) |
293
|
|
|
|
|
|
|
{ |
294
|
96
|
|
|
|
|
138
|
$product -= log $dLimits->[0]; |
295
|
96
|
|
|
|
|
96
|
$dLimits->[0]++; |
296
|
96
|
100
|
|
|
|
223
|
if ( $dLimits->[0] > $dLimits->[1] ) |
297
|
|
|
|
|
|
|
{ |
298
|
12
|
|
|
|
|
43
|
shift @{$dLimits}; |
|
12
|
|
|
|
|
49
|
|
299
|
12
|
|
|
|
|
18
|
shift @{$dLimits}; |
|
12
|
|
|
|
|
36
|
|
300
|
|
|
|
|
|
|
} |
301
|
|
|
|
|
|
|
} |
302
|
|
|
|
|
|
|
} |
303
|
6
|
|
|
|
|
13
|
return $product; |
304
|
|
|
|
|
|
|
} |
305
|
|
|
|
|
|
|
|
306
|
|
|
|
|
|
|
|
307
|
|
|
|
|
|
|
1; |
308
|
|
|
|
|
|
|
__END__ |