statsample 0.6.1 → 0.6.2
Sign up to get free protection for your applications and to get access to all the features.
- data/History.txt +4 -0
- data/Manifest.txt +8 -19
- data/data/locale/es/LC_MESSAGES/statsample.mo +0 -0
- data/demo/dominance_analysis_bootstrap.rb +20 -0
- data/demo/dominanceanalysis.rb +11 -0
- data/demo/multiple_regression.rb +40 -0
- data/demo/polychoric.rb +13 -0
- data/demo/tetrachoric.rb +10 -0
- data/lib/distribution.rb +1 -0
- data/lib/distribution/normalbivariate.rb +100 -0
- data/lib/statsample.rb +4 -105
- data/lib/statsample/bivariate.rb +5 -1
- data/lib/statsample/bivariate/polychoric.rb +581 -0
- data/lib/statsample/bivariate/tetrachoric.rb +37 -5
- data/lib/statsample/converters.rb +11 -0
- data/lib/statsample/dominanceanalysis.rb +104 -90
- data/lib/statsample/dominanceanalysis/bootstrap.rb +160 -131
- data/lib/statsample/factor/pca.rb +1 -2
- data/lib/statsample/factor/principalaxis.rb +2 -2
- data/lib/statsample/graph/svghistogram.rb +170 -172
- data/lib/statsample/matrix.rb +79 -0
- data/lib/statsample/mle.rb +6 -4
- data/lib/statsample/mle/probit.rb +0 -1
- data/lib/statsample/regression/multiple/alglibengine.rb +23 -23
- data/lib/statsample/regression/multiple/baseengine.rb +112 -113
- data/lib/statsample/regression/multiple/gslengine.rb +91 -94
- data/lib/statsample/regression/multiple/rubyengine.rb +1 -1
- data/lib/statsample/srs.rb +1 -1
- data/lib/statsample/test.rb +0 -1
- data/lib/statsample/test/umannwhitney.rb +8 -5
- data/po/es/statsample.po +201 -39
- data/po/statsample.pot +184 -32
- data/test/test_bivariate.rb +21 -2
- data/test/test_distribution.rb +58 -40
- data/test/test_factor.rb +0 -1
- data/test/test_gsl.rb +13 -14
- data/test/test_regression.rb +1 -1
- data/test/test_statistics.rb +1 -4
- metadata +10 -21
- data/demo/benchmark.rb +0 -76
- data/demo/chi-square.rb +0 -44
- data/demo/crosstab.rb +0 -7
- data/demo/dice.rb +0 -13
- data/demo/distribution_t.rb +0 -95
- data/demo/graph.rb +0 -9
- data/demo/item_analysis.rb +0 -30
- data/demo/mean.rb +0 -81
- data/demo/nunnally_6.rb +0 -34
- data/demo/pca.rb +0 -29
- data/demo/proportion.rb +0 -57
- data/demo/regression.rb +0 -82
- data/demo/sample_test.csv +0 -113
- data/demo/spss_matrix.rb +0 -3
- data/demo/strata_proportion.rb +0 -152
- data/demo/stratum.rb +0 -141
- data/demo/t-student.rb +0 -17
- data/demo/umann.rb +0 -8
- data/lib/matrix_extension.rb +0 -92
data/History.txt
CHANGED
@@ -1,3 +1,7 @@
|
|
1
|
+
=== 0.6.2 / 2010-02-11
|
2
|
+
* New Statsample::Bivariate::Polychoric. For implement: X2 and G2
|
3
|
+
* New matrix.rb, for faster development of Contingence Tables and Correlation Matrix
|
4
|
+
|
1
5
|
=== 0.6.1 / 2010-02-08
|
2
6
|
* Bug fix on DominanceAnalysis summary for Ruby1.9
|
3
7
|
* Some extra documentation
|
data/Manifest.txt
CHANGED
@@ -9,35 +9,23 @@ data/repeated_fields.csv
|
|
9
9
|
data/test_binomial.csv
|
10
10
|
data/tetmat_matrix.txt
|
11
11
|
data/tetmat_test.txt
|
12
|
-
demo/
|
13
|
-
demo/
|
14
|
-
demo/
|
15
|
-
demo/
|
16
|
-
demo/
|
17
|
-
demo/graph.rb
|
18
|
-
demo/item_analysis.rb
|
19
|
-
demo/mean.rb
|
20
|
-
demo/nunnally_6.rb
|
21
|
-
demo/pca.rb
|
22
|
-
demo/proportion.rb
|
23
|
-
demo/regression.rb
|
24
|
-
demo/sample_test.csv
|
25
|
-
demo/spss_matrix.rb
|
26
|
-
demo/strata_proportion.rb
|
27
|
-
demo/stratum.rb
|
28
|
-
demo/t-student.rb
|
29
|
-
demo/umann.rb
|
12
|
+
demo/dominance_analysis_bootstrap.rb
|
13
|
+
demo/dominanceanalysis.rb
|
14
|
+
demo/multiple_regression.rb
|
15
|
+
demo/polychoric.rb
|
16
|
+
demo/tetrachoric.rb
|
30
17
|
lib/distribution.rb
|
31
18
|
lib/distribution/chisquare.rb
|
32
19
|
lib/distribution/f.rb
|
33
20
|
lib/distribution/normal.rb
|
21
|
+
lib/distribution/normalbivariate.rb
|
34
22
|
lib/distribution/t.rb
|
35
|
-
lib/matrix_extension.rb
|
36
23
|
lib/spss.rb
|
37
24
|
lib/statistics2.rb
|
38
25
|
lib/statsample.rb
|
39
26
|
lib/statsample/anova.rb
|
40
27
|
lib/statsample/bivariate.rb
|
28
|
+
lib/statsample/bivariate/polychoric.rb
|
41
29
|
lib/statsample/bivariate/tetrachoric.rb
|
42
30
|
lib/statsample/codification.rb
|
43
31
|
lib/statsample/combination.rb
|
@@ -60,6 +48,7 @@ lib/statsample/graph/svghistogram.rb
|
|
60
48
|
lib/statsample/graph/svgscatterplot.rb
|
61
49
|
lib/statsample/histogram.rb
|
62
50
|
lib/statsample/htmlreport.rb
|
51
|
+
lib/statsample/matrix.rb
|
63
52
|
lib/statsample/mle.rb
|
64
53
|
lib/statsample/mle/logit.rb
|
65
54
|
lib/statsample/mle/normal.rb
|
Binary file
|
@@ -0,0 +1,20 @@
|
|
1
|
+
#!/usr/bin/ruby
|
2
|
+
$:.unshift(File.dirname(__FILE__)+'/../lib/')
|
3
|
+
require 'statsample'
|
4
|
+
|
5
|
+
|
6
|
+
a=100.times.collect {rand}.to_scale
|
7
|
+
b=100.times.collect {rand}.to_scale
|
8
|
+
c=100.times.collect {rand}.to_scale
|
9
|
+
d=100.times.collect {rand}.to_scale
|
10
|
+
|
11
|
+
ds={'a'=>a,'b'=>b,'c'=>c,'d'=>d}.to_dataset
|
12
|
+
|
13
|
+
ds['y']=ds.collect{|row| row['a']*5+row['b']*2+row['c']*2+row['d']*2+10*rand()}
|
14
|
+
dab=Statsample::DominanceAnalysis::Bootstrap.new(ds, 'y')
|
15
|
+
if HAS_GSL
|
16
|
+
# Use Gsl if available (faster calculation)
|
17
|
+
dab.regression_class=Statsample::Regression::Multiple::GslEngine
|
18
|
+
end
|
19
|
+
dab.bootstrap(100,nil,true)
|
20
|
+
puts dab.summary
|
@@ -0,0 +1,11 @@
|
|
1
|
+
#!/usr/bin/ruby
|
2
|
+
$:.unshift(File.dirname(__FILE__)+'/../lib/')
|
3
|
+
|
4
|
+
require 'statsample'
|
5
|
+
a=1000.times.collect {rand}.to_scale
|
6
|
+
b=1000.times.collect {rand}.to_scale
|
7
|
+
c=1000.times.collect {rand}.to_scale
|
8
|
+
ds={'a'=>a,'b'=>b,'c'=>c}.to_dataset
|
9
|
+
ds['y']=ds.collect{|row| row['a']*5+row['b']*3+row['c']*2+rand()}
|
10
|
+
da=Statsample::DominanceAnalysis.new(ds,'y')
|
11
|
+
puts da.summary
|
@@ -0,0 +1,40 @@
|
|
1
|
+
#!/usr/bin/ruby
|
2
|
+
$:.unshift(File.dirname(__FILE__)+'/../lib/')
|
3
|
+
|
4
|
+
require 'statsample'
|
5
|
+
require 'benchmark'
|
6
|
+
samples=10000
|
7
|
+
a=samples.times.collect {rand}.to_scale
|
8
|
+
b=samples.times.collect {rand}.to_scale
|
9
|
+
c=samples.times.collect {rand}.to_scale
|
10
|
+
d=samples.times.collect {rand}.to_scale
|
11
|
+
|
12
|
+
ds={'a'=>a,'b'=>b,'c'=>c,'d'=>d}.to_dataset
|
13
|
+
ds['y']=ds.collect{|row| row['a']*5+row['b']*3+row['c']*2+row['d']*1+rand()}
|
14
|
+
|
15
|
+
Benchmark.bm(7) do |x|
|
16
|
+
|
17
|
+
|
18
|
+
rb=ReportBuilder.new("Multiple Regression Engines")
|
19
|
+
|
20
|
+
if HAS_GSL
|
21
|
+
x.report("GSL:") {
|
22
|
+
lr=Statsample::Regression::Multiple::GslEngine.new(ds,'y',:name=>"Multiple Regression using GSL")
|
23
|
+
rb.add(lr.summary)
|
24
|
+
}
|
25
|
+
end
|
26
|
+
|
27
|
+
|
28
|
+
if HAS_ALGIB
|
29
|
+
x.report("Alglib:") {
|
30
|
+
lr=Statsample::Regression::Multiple::AlglibEngine.new(ds,'y', :name=>"Multiple Regression using Alglib")
|
31
|
+
rb.add(lr.summary)
|
32
|
+
}
|
33
|
+
end
|
34
|
+
x.report("Ruby:") {
|
35
|
+
lr=Statsample::Regression::Multiple::RubyEngine.new(ds,'y',:name=>"Multiple Regression using RubyEngine")
|
36
|
+
rb.add(lr.summary)
|
37
|
+
}
|
38
|
+
|
39
|
+
puts rb.to_text
|
40
|
+
end
|
data/demo/polychoric.rb
ADDED
@@ -0,0 +1,13 @@
|
|
1
|
+
#!/usr/bin/ruby
|
2
|
+
$:.unshift(File.dirname(__FILE__)+'/../lib/')
|
3
|
+
|
4
|
+
require 'statsample'
|
5
|
+
#ct=Matrix[[58,52,1],[26,58,3],[8,12,9]]
|
6
|
+
|
7
|
+
ct=Matrix[[30,1,0,0,0,0],[0,10,2,0,0,0], [0,4,8,3,1,0], [0,3,3,37,9,0], [0,0,1, 25, 71, 49], [ 0,0,0,2, 20, 181]]
|
8
|
+
poly=Statsample::Bivariate::Polychoric.new(ct)
|
9
|
+
|
10
|
+
puts poly.summary
|
11
|
+
puts poly.chi_square_independence
|
12
|
+
puts poly.chi_square_model
|
13
|
+
puts poly.chi_square_independence
|
data/demo/tetrachoric.rb
ADDED
data/lib/distribution.rb
CHANGED
@@ -0,0 +1,100 @@
|
|
1
|
+
module Distribution
|
2
|
+
# Calculate pdf and cdf for bivariate normal distribution
|
3
|
+
module NormalBivariate
|
4
|
+
|
5
|
+
class << self
|
6
|
+
SIDE=0.1
|
7
|
+
LIMIT=5
|
8
|
+
# Probability density function
|
9
|
+
# Source: http://en.wikipedia.org/wiki/Multivariate_normal_distribution
|
10
|
+
def pdf(x,y, rho, sigma1=1.0, sigma2=1.0)
|
11
|
+
(1.quo(2 * Math::PI * sigma1*sigma2 * Math::sqrt( 1 - rho**2 ))) *
|
12
|
+
Math::exp(-(1.quo(2*(1-rho**2))) *
|
13
|
+
((x**2/sigma1) + (y**2/sigma2) - (2*rho*x*y).quo(sigma1*sigma2) ))
|
14
|
+
end
|
15
|
+
def f(x,y,aprime,bprime,rho)
|
16
|
+
r=aprime*(2*x-aprime)+bprime*(2*y-bprime)+2*rho*(x-aprime)*(y-bprime)
|
17
|
+
Math::exp(r)
|
18
|
+
end
|
19
|
+
def cdf(a,b,rho)
|
20
|
+
cdf_math(a,b,rho)
|
21
|
+
end
|
22
|
+
def sgn(x)
|
23
|
+
if(x>=0)
|
24
|
+
1
|
25
|
+
else
|
26
|
+
-1
|
27
|
+
end
|
28
|
+
end
|
29
|
+
# As http://finance.bi.no/~bernt/gcc_prog/recipes/recipes/node23.html
|
30
|
+
def cdf_math(a,b,rho)
|
31
|
+
#puts "a:#{a} - b:#{b} - rho:#{rho}"
|
32
|
+
if (a<=0 and b<=0 and rho<=0)
|
33
|
+
# puts "ruta 1"
|
34
|
+
aprime=a.quo(Math::sqrt(2.0*(1.0-rho**2)))
|
35
|
+
bprime=b.quo(Math::sqrt(2.0*(1.0-rho**2)))
|
36
|
+
aa=[0.3253030, 0.4211071, 0.1334425, 0.006374323]
|
37
|
+
bb=[0.1337764, 0.6243247, 1.3425378, 2.2626645]
|
38
|
+
sum=0
|
39
|
+
4.times do |i|
|
40
|
+
4.times do |j|
|
41
|
+
sum+=aa[i]*aa[j] * f(bb[i], bb[j], aprime, bprime,rho)
|
42
|
+
end
|
43
|
+
end
|
44
|
+
sum=sum*(Math::sqrt(1.0-rho**2).quo(Math::PI))
|
45
|
+
return sum
|
46
|
+
elsif(a*b*rho<=0.0)
|
47
|
+
|
48
|
+
#puts "ruta 2"
|
49
|
+
if(a<=0 and b>=0 and rho>=0)
|
50
|
+
return Distribution::Normal.cdf(a) - cdf(a,-b,-rho)
|
51
|
+
elsif (a>=0.0 and b<=0.0 and rho>=0)
|
52
|
+
return Distribution::Normal.cdf(b) - cdf(-a,b,-rho)
|
53
|
+
elsif (a>=0.0 and b>=0.0 and rho<=0)
|
54
|
+
return Distribution::Normal.cdf(a) + Distribution::Normal.cdf(b) - 1.0 + cdf(-a,-b,rho)
|
55
|
+
end
|
56
|
+
elsif (a*b*rho>=0.0)
|
57
|
+
#puts "ruta 3"
|
58
|
+
denum=Math::sqrt(a**2 - 2*rho*a*b + b**2)
|
59
|
+
rho1=((rho*a-b)*sgn(a)).quo(denum)
|
60
|
+
rho2=((rho*b-a)*sgn(b)).quo(denum)
|
61
|
+
delta=(1.0-sgn(a)*sgn(b)).quo(4)
|
62
|
+
#puts "#{rho1} - #{rho2}"
|
63
|
+
return cdf(a, 0.0, rho1) + cdf(b, 0.0, rho2) - delta
|
64
|
+
end
|
65
|
+
raise "Should'nt be here! #{a} - #{b} #{rho}"
|
66
|
+
end
|
67
|
+
# Cdf for a given x and y
|
68
|
+
# Reference:
|
69
|
+
# * Jantaravareerat, M. & Thomopoulos, N. (n/d). Tables for standard bivariate normal distribution
|
70
|
+
|
71
|
+
def cdf_iterate(x,y,rho,s1=1,s2=1)
|
72
|
+
# Special cases
|
73
|
+
return 1 if x>LIMIT and y>LIMIT
|
74
|
+
return 0 if x<-LIMIT or y<-LIMIT
|
75
|
+
return Distribution::Normal.cdf(y) if x>LIMIT
|
76
|
+
return Distribution::Normal.cdf(x) if y>LIMIT
|
77
|
+
|
78
|
+
#puts "x:#{x} - y:#{y}"
|
79
|
+
x=-LIMIT if x<-LIMIT
|
80
|
+
x=LIMIT if x>LIMIT
|
81
|
+
y=-LIMIT if y<-LIMIT
|
82
|
+
y=LIMIT if y>LIMIT
|
83
|
+
|
84
|
+
x_squares=((LIMIT+x) / SIDE).to_i
|
85
|
+
y_squares=((LIMIT+y) / SIDE).to_i
|
86
|
+
sum=0
|
87
|
+
x_squares.times do |i|
|
88
|
+
y_squares.times do |j|
|
89
|
+
z1=-LIMIT+(i+1)*SIDE
|
90
|
+
z2=-LIMIT+(j+1)*SIDE
|
91
|
+
#puts " #{z1}-#{z2}"
|
92
|
+
h=(pdf(z1,z2,rho,s1,s2)+pdf(z1-SIDE,z2,rho,s1,s2)+pdf(z1,z2-SIDE,rho,s1,s2) + pdf(z1-SIDE,z2-SIDE,rho,s1,s2)).quo(4)
|
93
|
+
sum+= (SIDE**2)*h # area
|
94
|
+
end
|
95
|
+
end
|
96
|
+
sum
|
97
|
+
end
|
98
|
+
end
|
99
|
+
end
|
100
|
+
end
|
data/lib/statsample.rb
CHANGED
@@ -23,6 +23,7 @@ $:.unshift(File.expand_path(File.dirname(__FILE__)+"/../ext"))
|
|
23
23
|
|
24
24
|
require 'matrix'
|
25
25
|
require 'distribution'
|
26
|
+
require 'reportbuilder'
|
26
27
|
|
27
28
|
class Numeric
|
28
29
|
def square ; self * self ; end
|
@@ -108,7 +109,7 @@ end
|
|
108
109
|
# * Dataset: An union of vectors.
|
109
110
|
#
|
110
111
|
module Statsample
|
111
|
-
VERSION = '0.6.
|
112
|
+
VERSION = '0.6.2'
|
112
113
|
SPLIT_TOKEN = ","
|
113
114
|
autoload(:Database, 'statsample/converters')
|
114
115
|
autoload(:Anova, 'statsample/anova')
|
@@ -135,6 +136,7 @@ module Statsample
|
|
135
136
|
autoload(:Regression, 'statsample/regression')
|
136
137
|
autoload(:Test, 'statsample/test')
|
137
138
|
autoload(:Factor, 'statsample/factor')
|
139
|
+
|
138
140
|
def self.load(filename)
|
139
141
|
if File.exists? filename
|
140
142
|
o=false
|
@@ -165,110 +167,6 @@ module Statsample
|
|
165
167
|
fp.close
|
166
168
|
end
|
167
169
|
end
|
168
|
-
module HtmlSummary
|
169
|
-
def add_line(n=nil)
|
170
|
-
self << "<hr />"
|
171
|
-
end
|
172
|
-
def nl
|
173
|
-
self << "<br />"
|
174
|
-
end
|
175
|
-
def add(text)
|
176
|
-
self << ("<p>"+text.gsub("\n","<br />")+"</p>")
|
177
|
-
end
|
178
|
-
def parse_table(table)
|
179
|
-
self << table.parse_html
|
180
|
-
end
|
181
|
-
end
|
182
|
-
module ConsoleSummary
|
183
|
-
def add_line(n=80)
|
184
|
-
self << "-"*n+"\n"
|
185
|
-
end
|
186
|
-
def nl
|
187
|
-
self << "\n"
|
188
|
-
end
|
189
|
-
def add(text)
|
190
|
-
self << text
|
191
|
-
end
|
192
|
-
def parse_table(table)
|
193
|
-
self << table.parse_console
|
194
|
-
end
|
195
|
-
end
|
196
|
-
class ReportTable
|
197
|
-
attr_reader :header
|
198
|
-
def initialize(h=[])
|
199
|
-
@rows=[]
|
200
|
-
@max_cols=[]
|
201
|
-
self.header=(h)
|
202
|
-
end
|
203
|
-
def add_row(row)
|
204
|
-
row.each_index{|i|
|
205
|
-
@max_cols[i]=row[i].to_s.size if @max_cols[i].nil? or row[i].to_s.size > @max_cols[i]
|
206
|
-
}
|
207
|
-
@rows.push(row)
|
208
|
-
end
|
209
|
-
def add_horizontal_line
|
210
|
-
@rows.push(:hr)
|
211
|
-
end
|
212
|
-
def header=(h)
|
213
|
-
h.each_index{|i|
|
214
|
-
@max_cols[i]=h[i].to_s.size if @max_cols[i].nil? or h[i].to_s.size>@max_cols[i]
|
215
|
-
}
|
216
|
-
@header=h
|
217
|
-
end
|
218
|
-
def parse_console_row(row)
|
219
|
-
out="| "
|
220
|
-
@max_cols.each_index{|i|
|
221
|
-
if row[i].nil?
|
222
|
-
out << " "*(@max_cols[i]+2)+"|"
|
223
|
-
else
|
224
|
-
t=row[i].to_s
|
225
|
-
out << " "+t+" "*(@max_cols[i]-t.size+1)+"|"
|
226
|
-
end
|
227
|
-
}
|
228
|
-
out << "\n"
|
229
|
-
out
|
230
|
-
end
|
231
|
-
def parse_console_hr
|
232
|
-
"-"*(@max_cols.inject(0){|a,v|a+v.size+3}+2)+"\n"
|
233
|
-
end
|
234
|
-
def parse_console
|
235
|
-
out="\n"
|
236
|
-
out << parse_console_hr
|
237
|
-
out << parse_console_row(header)
|
238
|
-
out << parse_console_hr
|
239
|
-
|
240
|
-
@rows.each{|row|
|
241
|
-
if row==:hr
|
242
|
-
out << parse_console_hr
|
243
|
-
else
|
244
|
-
out << parse_console_row(row)
|
245
|
-
end
|
246
|
-
}
|
247
|
-
out << parse_console_hr
|
248
|
-
|
249
|
-
out
|
250
|
-
end
|
251
|
-
def parse_html
|
252
|
-
out="<table>\n"
|
253
|
-
if header.size>0
|
254
|
-
out << "<thead><th>"+header.join("</th><th>")+"</thead><tbody>"
|
255
|
-
end
|
256
|
-
out << "<tbody>\n"
|
257
|
-
row_with_line=false
|
258
|
-
@rows.each{|row|
|
259
|
-
if row==:hr
|
260
|
-
row_with_line=true
|
261
|
-
else
|
262
|
-
out << "<tr class='"+(row_with_line ? 'line':'')+"'><td>"
|
263
|
-
out << row.join("</td><td>") +"</td>"
|
264
|
-
out << "</tr>\n"
|
265
|
-
row_with_line=false
|
266
|
-
end
|
267
|
-
}
|
268
|
-
out << "</tbody></table>\n"
|
269
|
-
out
|
270
|
-
end
|
271
|
-
end
|
272
170
|
|
273
171
|
module STATSAMPLE__ #:nodoc:
|
274
172
|
end
|
@@ -288,3 +186,4 @@ end
|
|
288
186
|
require 'statsample/vector'
|
289
187
|
require 'statsample/dataset'
|
290
188
|
require 'statsample/crosstab'
|
189
|
+
require 'statsample/matrix'
|
data/lib/statsample/bivariate.rb
CHANGED
@@ -1,4 +1,5 @@
|
|
1
1
|
require 'statsample/bivariate/tetrachoric'
|
2
|
+
require 'statsample/bivariate/polychoric'
|
2
3
|
module Statsample
|
3
4
|
# Diverse correlation methods
|
4
5
|
module Bivariate
|
@@ -132,7 +133,7 @@ module Statsample
|
|
132
133
|
# Order of rows and columns depends on Dataset#fields order
|
133
134
|
|
134
135
|
def correlation_matrix(ds)
|
135
|
-
ds.collect_matrix do |row,col|
|
136
|
+
cm=ds.collect_matrix do |row,col|
|
136
137
|
if row==col
|
137
138
|
1.0
|
138
139
|
elsif (ds[row].type!=:scale or ds[col].type!=:scale)
|
@@ -141,6 +142,9 @@ module Statsample
|
|
141
142
|
pearson(ds[row],ds[col])
|
142
143
|
end
|
143
144
|
end
|
145
|
+
cm.extend(Statsample::CorrelationMatrix)
|
146
|
+
cm.labels=ds.fields
|
147
|
+
cm
|
144
148
|
end
|
145
149
|
|
146
150
|
# Retrieves the n valid pairwise.
|
@@ -0,0 +1,581 @@
|
|
1
|
+
module Statsample
|
2
|
+
module Bivariate
|
3
|
+
# Calculate Polychoric correlation for two vectors.
|
4
|
+
def self.polychoric(v1,v2)
|
5
|
+
pc=Polychoric.new_with_vectors(v1,v2)
|
6
|
+
pc.r
|
7
|
+
end
|
8
|
+
|
9
|
+
# Polychoric correlation matrix.
|
10
|
+
# Order of rows and columns depends on Dataset#fields order
|
11
|
+
def self.polychoric_correlation_matrix(ds)
|
12
|
+
ds.collect_matrix do |row,col|
|
13
|
+
if row==col
|
14
|
+
1.0
|
15
|
+
else
|
16
|
+
begin
|
17
|
+
polychoric(ds[row],ds[col])
|
18
|
+
rescue RuntimeError
|
19
|
+
nil
|
20
|
+
end
|
21
|
+
end
|
22
|
+
end
|
23
|
+
end
|
24
|
+
# Compute polychoric correlation.
|
25
|
+
#
|
26
|
+
# The polychoric correlation estimate what the correlation between raters, who classified on a ordered category scale, would be if ratings were made on a continuous scale; they are, theoretically, invariant over changes in the number or "width" of rating categories.
|
27
|
+
# See extensive documentation on http://www.john-uebersax.com/stat/tetra.htm
|
28
|
+
|
29
|
+
class Polychoric
|
30
|
+
include GetText
|
31
|
+
bindtextdomain("statsample")
|
32
|
+
# Name of the analysis
|
33
|
+
attr_accessor :name
|
34
|
+
# Max number of iterations used on iterative methods. Default to 100
|
35
|
+
attr_accessor :max_iterations
|
36
|
+
# Debug algorithm (See iterations, for example)
|
37
|
+
attr_accessor :debug
|
38
|
+
# Minimizer type. Default GSL::Min::FMinimizer::BRENT
|
39
|
+
# See http://rb-gsl.rubyforge.org/min.html for reference.
|
40
|
+
attr_accessor :minimizer_type
|
41
|
+
# Method of calculation.
|
42
|
+
#
|
43
|
+
# Drasgow (1988, cited by Uebersax, 2002) describe two method: joint maximum likelihood (ML) approach and two-step ML estimation.
|
44
|
+
# For now, only implemented two-step ML (:two_step), with algorithm
|
45
|
+
# based on Drasgow(1986, cited by Gegenfurtner, 1992)
|
46
|
+
#
|
47
|
+
attr_accessor :method
|
48
|
+
# Absolute error for iteration. Default to 0.001
|
49
|
+
attr_accessor :epsilon
|
50
|
+
|
51
|
+
# Number of iterations
|
52
|
+
attr_reader :iteration
|
53
|
+
|
54
|
+
# Log of algorithm
|
55
|
+
attr_reader :log
|
56
|
+
attr_reader :loglike
|
57
|
+
MAX_ITERATIONS=100
|
58
|
+
EPSILON=0.001
|
59
|
+
MINIMIZER_TYPE=GSL::Min::FMinimizer::BRENT
|
60
|
+
def new_with_vectors(v1,v2)
|
61
|
+
Polychoric.new(Crosstab.new(v1,v2).to_matrix)
|
62
|
+
end
|
63
|
+
|
64
|
+
def initialize(matrix, opts=Hash.new)
|
65
|
+
@matrix=matrix
|
66
|
+
@n=matrix.column_size
|
67
|
+
@m=matrix.row_size
|
68
|
+
raise "row size <1" if @m<=1
|
69
|
+
raise "column size <1" if @n<=1
|
70
|
+
|
71
|
+
@method=:two_step
|
72
|
+
@name="Polychoric correlation"
|
73
|
+
@max_iterations=MAX_ITERATIONS
|
74
|
+
@epsilon=EPSILON
|
75
|
+
@minimizer_type=GSL::Min::FMinimizer::BRENT
|
76
|
+
@debug=false
|
77
|
+
@iteration=nil
|
78
|
+
opts.each{|k,v|
|
79
|
+
self.send("#{k}=",v) if self.respond_to? k
|
80
|
+
}
|
81
|
+
@r=nil
|
82
|
+
end
|
83
|
+
def r
|
84
|
+
if @r.nil?
|
85
|
+
compute
|
86
|
+
end
|
87
|
+
@r
|
88
|
+
end
|
89
|
+
|
90
|
+
def threshold_x
|
91
|
+
if @alpha.nil?
|
92
|
+
compute
|
93
|
+
end
|
94
|
+
@alpha[0,@alpha.size-1]
|
95
|
+
end
|
96
|
+
|
97
|
+
def threshold_y
|
98
|
+
if @beta.nil?
|
99
|
+
compute
|
100
|
+
end
|
101
|
+
@beta[0,@beta.size-1]
|
102
|
+
end
|
103
|
+
|
104
|
+
|
105
|
+
|
106
|
+
def compute
|
107
|
+
if @method==:two_step
|
108
|
+
compute_two_step_mle_drasgow
|
109
|
+
elsif @method==:as87
|
110
|
+
compute_two_step_as87
|
111
|
+
else
|
112
|
+
raise "Not implemented"
|
113
|
+
end
|
114
|
+
end
|
115
|
+
# *Computation of polychoric correlation usign two-step ML estimation.*
|
116
|
+
#
|
117
|
+
# Two-step ML estimation "first estimates the thresholds from the one-way marginal frequencies, then estimates rho, conditional on these thresholds, via maximum likelihood" (Uebersax, 2006).
|
118
|
+
#
|
119
|
+
# The algorithm is based on Drasgow(1986, cited by Gegenfurtner (1992)
|
120
|
+
# References:
|
121
|
+
# * Gegenfurtner, K. (1992). PRAXIS: Brent's algorithm for function minimization. Behavior Research Methods, Instruments & Computers, 24(4), 560-564. Available on http://www.allpsych.uni-giessen.de/karl/pdf/03.praxis.pdf
|
122
|
+
# * Uebersax, J.S. (2006). The tetrachoric and polychoric correlation coefficients. Statistical Methods for Rater Agreement web site. 2006. Available at: http://john-uebersax.com/stat/tetra.htm . Accessed February, 11, 2010
|
123
|
+
#
|
124
|
+
def compute_two_step_mle_drasgow
|
125
|
+
@nr=@matrix.row_size
|
126
|
+
@nc=@matrix.column_size
|
127
|
+
@sumr=[0]*@matrix.row_size
|
128
|
+
@sumrac=[0]*@matrix.row_size
|
129
|
+
@sumc=[0]*@matrix.column_size
|
130
|
+
@sumcac=[0]*@matrix.column_size
|
131
|
+
@alpha=[0]*@matrix.row_size
|
132
|
+
@beta=[0]*@matrix.row_size
|
133
|
+
@total=0
|
134
|
+
@nr.times do |i|
|
135
|
+
@nc.times do |j|
|
136
|
+
@sumr[i]+=@matrix[i,j]
|
137
|
+
@sumc[j]+=@matrix[i,j]
|
138
|
+
@total+=@matrix[i,j]
|
139
|
+
end
|
140
|
+
end
|
141
|
+
ac=0
|
142
|
+
(@nr-1).times do |i|
|
143
|
+
@sumrac[i]=@sumr[i]+ac
|
144
|
+
@alpha[i]=Distribution::Normal.p_value(@sumrac[i] / @total.to_f)
|
145
|
+
ac=@sumrac[i]
|
146
|
+
end
|
147
|
+
ac=0
|
148
|
+
(@nc-1).times do |i|
|
149
|
+
@sumcac[i]=@sumc[i]+ac
|
150
|
+
@beta[i]=Distribution::Normal.p_value(@sumcac[i] / @total.to_f)
|
151
|
+
ac=@sumcac[i]
|
152
|
+
end
|
153
|
+
@alpha[@nr-1]=10
|
154
|
+
@beta[@nc-1]=10
|
155
|
+
fn1=GSL::Function.alloc {|x|
|
156
|
+
loglike=0
|
157
|
+
pd=@nr.times.collect{ [0]*@nc}
|
158
|
+
pc=@nr.times.collect{ [0]*@nc}
|
159
|
+
|
160
|
+
@nr.times { |i|
|
161
|
+
@nc.times { |j|
|
162
|
+
pd[i][j]=Distribution::NormalBivariate.cdf(@alpha[i], @beta[j], x)
|
163
|
+
pc[i][j] = pd[i][j]
|
164
|
+
pd[i][j] = pd[i][j] - pc[i-1][j] if i>0
|
165
|
+
pd[i][j] = pd[i][j] - pc[i][j-1] if j>0
|
166
|
+
pd[i][j] = pd[i][j] + pc[i-1][j-1] if (i>0 and j>0)
|
167
|
+
res= pd[i][j]
|
168
|
+
|
169
|
+
if res==0.0
|
170
|
+
res=1e-15
|
171
|
+
end
|
172
|
+
|
173
|
+
# puts "i:#{i} | j:#{j} | ac: #{sprintf("%0.4f", pc[i][j])} | pd: #{sprintf("%0.4f", pd[i][j])} | res:#{sprintf("%0.4f", res)}"
|
174
|
+
loglike+= @matrix[i,j] * Math::log( res )
|
175
|
+
}
|
176
|
+
}
|
177
|
+
# p pd
|
178
|
+
@loglike=loglike
|
179
|
+
@pd=pd
|
180
|
+
-loglike
|
181
|
+
}
|
182
|
+
@iteration = 0
|
183
|
+
max_iter = @max_iterations
|
184
|
+
m = 0 # initial guess
|
185
|
+
m_expected = 0.5
|
186
|
+
a=-0.99999
|
187
|
+
b=+0.99999
|
188
|
+
gmf = GSL::Min::FMinimizer.alloc(@minimizer_type)
|
189
|
+
gmf.set(fn1, m, a, b)
|
190
|
+
header=sprintf("using %s method\n", gmf.name)
|
191
|
+
header+=sprintf("%5s [%9s, %9s] %9s %10s %9s\n", "iter", "lower", "upper", "min",
|
192
|
+
"err", "err(est)")
|
193
|
+
|
194
|
+
header+=sprintf("%5d [%.7f, %.7f] %.7f %+.7f %.7f\n", @iteration, a, b, m, m - m_expected, b - a)
|
195
|
+
@log=header
|
196
|
+
puts header if @debug
|
197
|
+
begin
|
198
|
+
@iteration += 1
|
199
|
+
status = gmf.iterate
|
200
|
+
status = gmf.test_interval(0.001, 0.0)
|
201
|
+
|
202
|
+
if status == GSL::SUCCESS
|
203
|
+
@log+="Converged:"
|
204
|
+
puts "Converged:" if @debug
|
205
|
+
end
|
206
|
+
a = gmf.x_lower
|
207
|
+
b = gmf.x_upper
|
208
|
+
m = gmf.x_minimum
|
209
|
+
message=sprintf("%5d [%.7f, %.7f] %.7f %+.7f %.7f\n",
|
210
|
+
@iteration, a, b, m, m - m_expected, b - a);
|
211
|
+
@log+=message
|
212
|
+
puts message if @debug
|
213
|
+
end while status == GSL::CONTINUE and @iteration < @max_iterations
|
214
|
+
@r=gmf.x_minimum
|
215
|
+
end
|
216
|
+
# Chi-square to test r=0
|
217
|
+
def chi_square_independence
|
218
|
+
Statsample::Test::chi_square(@matrix, expected)
|
219
|
+
end
|
220
|
+
# Chi-square to test model==independence
|
221
|
+
|
222
|
+
def chi_square_model_expected
|
223
|
+
calculate if @r.nil?
|
224
|
+
model=Matrix.rows(@pd).collect {|c| c*@total}
|
225
|
+
Statsample::Test::chi_square(model, expected)
|
226
|
+
|
227
|
+
end
|
228
|
+
# Chi-square to test real == calculated with rho
|
229
|
+
def chi_square_model
|
230
|
+
calculate if @r.nil?
|
231
|
+
e=Matrix.rows(@pd).collect {|c| c*@total}
|
232
|
+
Statsample::Test::chi_square(@matrix, e)
|
233
|
+
end
|
234
|
+
def matrix_for_rho(rho)
|
235
|
+
pd=@nr.times.collect{ [0]*@nc}
|
236
|
+
pc=@nr.times.collect{ [0]*@nc}
|
237
|
+
@nr.times { |i|
|
238
|
+
@nc.times { |j|
|
239
|
+
pd[i][j]=Distribution::NormalBivariate.cdf(@alpha[i], @beta[j], rho)
|
240
|
+
pc[i][j] = pd[i][j]
|
241
|
+
pd[i][j] = pd[i][j] - pc[i-1][j] if i>0
|
242
|
+
pd[i][j] = pd[i][j] - pc[i][j-1] if j>0
|
243
|
+
pd[i][j] = pd[i][j] + pc[i-1][j-1] if (i>0 and j>0)
|
244
|
+
res= pd[i][j]
|
245
|
+
}
|
246
|
+
}
|
247
|
+
Matrix.rows(pc)
|
248
|
+
end
|
249
|
+
def g2
|
250
|
+
raise "Doesn't work"
|
251
|
+
e=expected
|
252
|
+
no_r_likehood=0
|
253
|
+
@nr.times {|i|
|
254
|
+
@nc.times {|j|
|
255
|
+
#p @matrix[i,j]
|
256
|
+
if @matrix[i,j]!=0
|
257
|
+
no_r_likehood+= @matrix[i,j]*Math::log(e[i,j])
|
258
|
+
end
|
259
|
+
}
|
260
|
+
}
|
261
|
+
p no_r_likehood
|
262
|
+
model=Matrix.rows(@pd).collect {|c| c*@total}
|
263
|
+
|
264
|
+
model_likehood=0
|
265
|
+
@nr.times {|i|
|
266
|
+
@nc.times {|j|
|
267
|
+
#p @matrix[i,j]
|
268
|
+
if @matrix[i,j]!=0
|
269
|
+
model_likehood+= @matrix[i,j] * Math::log(model[i,j])
|
270
|
+
end
|
271
|
+
}
|
272
|
+
}
|
273
|
+
|
274
|
+
p model_likehood
|
275
|
+
|
276
|
+
-2*(no_r_likehood-model_likehood)
|
277
|
+
|
278
|
+
end
|
279
|
+
def expected
|
280
|
+
rt=[]
|
281
|
+
ct=[]
|
282
|
+
t=0
|
283
|
+
@matrix.row_size.times {|i|
|
284
|
+
@matrix.column_size.times {|j|
|
285
|
+
rt[i]=0 if rt[i].nil?
|
286
|
+
ct[j]=0 if ct[j].nil?
|
287
|
+
rt[i]+=@matrix[i,j]
|
288
|
+
ct[j]+=@matrix[i,j]
|
289
|
+
t+=@matrix[i,j]
|
290
|
+
}
|
291
|
+
}
|
292
|
+
m=[]
|
293
|
+
@matrix.row_size.times {|i|
|
294
|
+
row=[]
|
295
|
+
@matrix.column_size.times {|j|
|
296
|
+
row[j]=(rt[i]*ct[j]).quo(t)
|
297
|
+
}
|
298
|
+
m.push(row)
|
299
|
+
}
|
300
|
+
|
301
|
+
Matrix.rows(m)
|
302
|
+
end
|
303
|
+
# Compute polychoric using AS87.
|
304
|
+
# Doesn't work for now! I can't find the error :(
|
305
|
+
|
306
|
+
def compute_two_step_as87
|
307
|
+
@nn=@n-1
|
308
|
+
@mm=@m-1
|
309
|
+
@nn7=7*@nn
|
310
|
+
@mm7=7*@mm
|
311
|
+
@mn=@n*@m
|
312
|
+
@cont=[nil]
|
313
|
+
@n.times {|j|
|
314
|
+
@m.times {|i|
|
315
|
+
@cont.push(@matrix[i,j])
|
316
|
+
}
|
317
|
+
}
|
318
|
+
|
319
|
+
pcorl=0
|
320
|
+
cont=@cont
|
321
|
+
xmean=0.0
|
322
|
+
sum=0.0
|
323
|
+
row=[]
|
324
|
+
colmn=[]
|
325
|
+
(1..@m).each do |i|
|
326
|
+
row[i]=0.0
|
327
|
+
l=i
|
328
|
+
(1..@n).each do |j|
|
329
|
+
row[i]=row[i]+cont[l]
|
330
|
+
l+=@m
|
331
|
+
end
|
332
|
+
raise "Should not be empty rows" if(row[i]==0.0)
|
333
|
+
xmean=xmean+row[i]*i.to_f
|
334
|
+
sum+=row[i]
|
335
|
+
end
|
336
|
+
xmean=xmean/sum.to_f
|
337
|
+
ymean=0.0
|
338
|
+
(1..@n).each do |j|
|
339
|
+
colmn[j]=0.0
|
340
|
+
l=(j-1)*@m
|
341
|
+
(1..@m).each do |i|
|
342
|
+
l=l+1
|
343
|
+
colmn[j]=colmn[j]+cont[l] #12
|
344
|
+
end
|
345
|
+
raise "Should not be empty cols" if colmn[j]==0
|
346
|
+
ymean=ymean+colmn[j]*j.to_f
|
347
|
+
end
|
348
|
+
ymean=ymean/sum.to_f
|
349
|
+
covxy=0.0
|
350
|
+
(1..@m).each do |i|
|
351
|
+
l=i
|
352
|
+
(1..@n).each do |j|
|
353
|
+
conxy=covxy+cont[l]*(i.to_f-xmean)*(j.to_f-ymean)
|
354
|
+
l=l+@m
|
355
|
+
end
|
356
|
+
end
|
357
|
+
|
358
|
+
chisq=0.0
|
359
|
+
(1..@m).each do |i|
|
360
|
+
l=i
|
361
|
+
(1..@n).each do |j|
|
362
|
+
chisq=chisq+((cont[l]**2).quo(row[i]*colmn[j]))
|
363
|
+
l=l+@m
|
364
|
+
end
|
365
|
+
end
|
366
|
+
|
367
|
+
phisq=chisq-1.0-(@mm*@nn).to_f / sum.to_f
|
368
|
+
phisq=0 if(phisq<0.0)
|
369
|
+
# Compute cumulative sum of columns and rows
|
370
|
+
sumc=[]
|
371
|
+
sumr=[]
|
372
|
+
sumc[1]=colmn[1]
|
373
|
+
sumr[1]=row[1]
|
374
|
+
cum=0
|
375
|
+
(1..@nn).each do |i| # goto 17 r20
|
376
|
+
cum=cum+colmn[i]
|
377
|
+
sumc[i]=cum
|
378
|
+
end
|
379
|
+
cum=0
|
380
|
+
(1..@mm).each do |i|
|
381
|
+
cum=cum+row[i]
|
382
|
+
sumr[i]=cum
|
383
|
+
end
|
384
|
+
alpha=[]
|
385
|
+
beta=[]
|
386
|
+
# Compute points of polytomy
|
387
|
+
(1..@mm).each do |i| #do 21
|
388
|
+
alpha[i]=Distribution::Normal.p_value(sumr[i] / sum.to_f)
|
389
|
+
end # 21
|
390
|
+
(1..@nn).each do |i| #do 22
|
391
|
+
beta[i]=Distribution::Normal.p_value(sumc[i] / sum.to_f)
|
392
|
+
end # 21
|
393
|
+
@alpha=alpha[1,alpha.size] << nil
|
394
|
+
@beta=beta[1,beta.size] << nil
|
395
|
+
@sumr=sumr
|
396
|
+
@sumc=sumc
|
397
|
+
@total=sum
|
398
|
+
|
399
|
+
# Compute Fourier coefficients a and b. Verified
|
400
|
+
h=hermit(alpha,@mm)
|
401
|
+
hh=hermit(beta,@nn)
|
402
|
+
a=[]
|
403
|
+
b=[]
|
404
|
+
if @m!=2 # goto 24
|
405
|
+
mmm=@m-2
|
406
|
+
(1..mmm).each do |i| #do 23
|
407
|
+
a1=sum.quo(row[i+1] * sumr[i] * sumr[i+1])
|
408
|
+
a2=sumr[i] * xnorm(alpha[i+1])
|
409
|
+
a3=sumr[i+1] * xnorm(alpha[i])
|
410
|
+
l=i
|
411
|
+
(1..7).each do |j| #do 23
|
412
|
+
a[l]=Math::sqrt(a1.quo(j))*(h[l+1] * a2 - h[l] * a3)
|
413
|
+
l=l+@mm
|
414
|
+
end
|
415
|
+
end #23
|
416
|
+
end
|
417
|
+
# 24
|
418
|
+
|
419
|
+
|
420
|
+
if @n!=2 # goto 26
|
421
|
+
nnn=@n-2
|
422
|
+
(1..nnn).each do |i| #do 25
|
423
|
+
a1=sum.quo(colmn[i+1] * sumc[i] * sumc[i+1])
|
424
|
+
a2=sumc[i] * xnorm(beta[i+1])
|
425
|
+
a3=sumc[i+1] * xnorm(beta[i])
|
426
|
+
l=i
|
427
|
+
(1..7).each do |j| #do 25
|
428
|
+
b[l]=Math::sqrt(a1.quo(j))*(a2 * hh[l+1] - a3*hh[l])
|
429
|
+
l=l+@nn
|
430
|
+
end # 25
|
431
|
+
end # 25
|
432
|
+
end
|
433
|
+
#26 r20
|
434
|
+
l = @mm
|
435
|
+
a1 = -sum * xnorm(alpha[@mm])
|
436
|
+
a2 = row[@m] * sumr[@mm]
|
437
|
+
(1..7).each do |j| # do 27
|
438
|
+
a[l]=a1 * h[l].quo(Math::sqrt(j*a2))
|
439
|
+
l=l+@mm
|
440
|
+
end # 27
|
441
|
+
|
442
|
+
l = @nn
|
443
|
+
a1 = -sum * xnorm(beta[@nn])
|
444
|
+
a2 = colmn[@n] * sumc[@nn]
|
445
|
+
|
446
|
+
(1..7).each do |j| # do 28
|
447
|
+
b[l]=a1 * hh[l].quo(Math::sqrt(j*a2))
|
448
|
+
l = l + @nn
|
449
|
+
end # 28
|
450
|
+
rcof=[]
|
451
|
+
# compute coefficients rcof of polynomial of order 8
|
452
|
+
rcof[1]=-phisq
|
453
|
+
(2..9).each do |i| # do 30
|
454
|
+
rcof[i]=0.0
|
455
|
+
end #30
|
456
|
+
m1=@mm
|
457
|
+
(1..@mm).each do |i| # do 31
|
458
|
+
m1=m1+1
|
459
|
+
m2=m1+@mm
|
460
|
+
m3=m2+@mm
|
461
|
+
m4=m3+@mm
|
462
|
+
m5=m4+@mm
|
463
|
+
m6=m5+@mm
|
464
|
+
n1=@nn
|
465
|
+
(1..@nn).each do |j| # do 31
|
466
|
+
n1=n1+1
|
467
|
+
n2=n1+@nn
|
468
|
+
n3=n2+@nn
|
469
|
+
n4=n3+@nn
|
470
|
+
n5=n4+@nn
|
471
|
+
n6=n5+@nn
|
472
|
+
|
473
|
+
rcof[3] = rcof[3] + a[i]**2 * b[j]**2
|
474
|
+
|
475
|
+
rcof[4] = rcof[4] + 2.0 * a[i] * a[m1] * b[j] * b[n1]
|
476
|
+
|
477
|
+
rcof[5] = rcof[5] + a[m1]**2 * b[n1]**2 +
|
478
|
+
2.0 * a[i] * a[m2] * b[j] * b[n2]
|
479
|
+
|
480
|
+
rcof[6] = rcof[6] + 2.0 * (a[i] * a[m3] * b[j] *
|
481
|
+
b[n3] + a[m1] * a[m2] * b[n1] * b[n2])
|
482
|
+
|
483
|
+
rcof[7] = rcof[7] + a[m2]**2 * b[n2]**2 +
|
484
|
+
2.0 * (a[i] * a[m4] * b[j] * b[n4] + a[m1] * a[m3] *
|
485
|
+
b[n1] * b[n3])
|
486
|
+
|
487
|
+
rcof[8] = rcof[8] + 2.0 * (a[i] * a[m5] * b[j] * b[n5] +
|
488
|
+
a[m1] * a[m4] * b[n1] * b[n4] + a[m2] * a[m3] * b[n2] * b[n3])
|
489
|
+
|
490
|
+
rcof[9] = rcof[9] + a[m3]**2 * b[n3]**2 +
|
491
|
+
2.0 * (a[i] * a[m6] * b[j] * b[n6] + a[m1] * a[m5] * b[n1] *
|
492
|
+
b[n5] + (a[m2] * a[m4] * b[n2] * b[n4]))
|
493
|
+
end # 31
|
494
|
+
end # 31
|
495
|
+
|
496
|
+
rcof=rcof[1,rcof.size]
|
497
|
+
poly = GSL::Poly.alloc(rcof)
|
498
|
+
roots=poly.solve
|
499
|
+
rootr=[nil]
|
500
|
+
rooti=[nil]
|
501
|
+
roots.each {|c|
|
502
|
+
rootr.push(c.real)
|
503
|
+
rooti.push(c.im)
|
504
|
+
}
|
505
|
+
@rootr=rootr
|
506
|
+
@rooti=rooti
|
507
|
+
|
508
|
+
norts=0
|
509
|
+
(1..7).each do |i| # do 43
|
510
|
+
|
511
|
+
next if rooti[i]!=0.0
|
512
|
+
if (covxy>=0.0)
|
513
|
+
next if(rootr[i]<0.0 or rootr[i]>1.0)
|
514
|
+
pcorl=rootr[i]
|
515
|
+
norts=norts+1
|
516
|
+
else
|
517
|
+
if (rootr[i]>=-1.0 and rootr[i]<0.0)
|
518
|
+
pcorl=rootr[i]
|
519
|
+
norts=norts+1
|
520
|
+
end
|
521
|
+
end
|
522
|
+
end # 43
|
523
|
+
raise "Error" if norts==0
|
524
|
+
@r=pcorl
|
525
|
+
end
|
526
|
+
#Computes vector h(mm7) of orthogonal hermite...
|
527
|
+
def hermit(s,k)
|
528
|
+
h=[]
|
529
|
+
(1..k).each do |i| # do 14
|
530
|
+
l=i
|
531
|
+
ll=i+k
|
532
|
+
lll=ll+k
|
533
|
+
h[i]=1.0
|
534
|
+
h[ll]=s[i]
|
535
|
+
v=1.0
|
536
|
+
(2..6).each do |j| #do 14
|
537
|
+
w=Math::sqrt(j)
|
538
|
+
h[lll]=(s[i]*h[ll] - v*h[l]).quo(w)
|
539
|
+
v=w
|
540
|
+
l=l+k
|
541
|
+
ll=ll+k
|
542
|
+
lll=lll+k
|
543
|
+
end
|
544
|
+
end
|
545
|
+
h
|
546
|
+
end
|
547
|
+
def xnorm(t)
|
548
|
+
Math::exp(-0.5 * t **2) * (1.0/Math::sqrt(2*Math::PI))
|
549
|
+
end
|
550
|
+
|
551
|
+
def summary
|
552
|
+
rp=ReportBuilder.new()
|
553
|
+
rp.add(self)
|
554
|
+
rp.to_text
|
555
|
+
end
|
556
|
+
|
557
|
+
def to_reportbuilder(generator)
|
558
|
+
compute if @r.nil?
|
559
|
+
section=ReportBuilder::Section.new(:name=>@name)
|
560
|
+
t=ReportBuilder::Table.new(:name=>_("Contingence Table"),:header=>[""]+(@n.times.collect {|i| "Y=#{i}"})+["Total"])
|
561
|
+
@m.times do |i|
|
562
|
+
t.add_row(["X = #{i}"]+(@n.times.collect {|j| @matrix[i,j]}) + [@sumr[i]])
|
563
|
+
end
|
564
|
+
t.add_hr
|
565
|
+
t.add_row(["T"]+(@n.times.collect {|j| @sumc[j]})+[@total])
|
566
|
+
section.add(t)
|
567
|
+
#generator.parse_element(t)
|
568
|
+
section.add(sprintf("r: %0.4f",r))
|
569
|
+
t=ReportBuilder::Table.new(:name=>_("Thresholds"), :header=>["","Value"])
|
570
|
+
threshold_x.each_with_index {|val,i|
|
571
|
+
t.add_row(["Threshold X #{i}", sprintf("%0.4f", val)])
|
572
|
+
}
|
573
|
+
threshold_y.each_with_index {|val,i|
|
574
|
+
t.add_row(["Threshold Y #{i}", sprintf("%0.4f", val)])
|
575
|
+
}
|
576
|
+
section.add(t)
|
577
|
+
generator.parse_element(section)
|
578
|
+
end
|
579
|
+
end
|
580
|
+
end
|
581
|
+
end
|