num4anova 0.0.3-java → 0.0.5-java
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/CHANGELOG.md +11 -1
- data/ext/num4anova/MultiComp.java +4 -1
- data/ext/num4anova/OneWayLayout.java +140 -3
- data/ext/num4anova/TwoWayLayout.java +186 -0
- data/lib/num4anova.rb +103 -14
- metadata +3 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 7a089d52ead2c9726d9ca61db475f128925afddca7cbcf6f65f0da75e41f0b1b
|
4
|
+
data.tar.gz: 050e29e04bd0e8272da89d5aba231fe08294f50f090d261da7aa6822f5df5bc5
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: f0cc84d4cebb416b04ed0360ef4cf72799aaebcf70104b6e2d2dc2bc471400d3ac313b71f5885fd0854ae8b3ae22dd66c753066233b1ed2848aabf586cee2c05
|
7
|
+
data.tar.gz: 88a74fb295e7ce0e0bdba405a14239a37fac4f3115bd29f358ce7b54fcc3b71a1533ff095e5aea7e2b141d867051eca703d179874f1ed349aef92575d8d0a186
|
data/CHANGELOG.md
CHANGED
@@ -2,10 +2,20 @@
|
|
2
2
|
|
3
3
|
## Unreleased
|
4
4
|
|
5
|
+
## [0.0.5] - 2024-02-01
|
6
|
+
|
7
|
+
### add
|
8
|
+
- add function of twoway_anova.
|
9
|
+
|
10
|
+
## [0.0.4] - 2024-01-30
|
11
|
+
### add
|
12
|
+
- add function of replicate_plot
|
13
|
+
- add function of replicate_test
|
14
|
+
|
5
15
|
## [0.0.3] - 2024-01-26
|
6
16
|
|
7
17
|
### add
|
8
|
-
- add function of dunnet_test
|
18
|
+
- add function of dunnet_test
|
9
19
|
|
10
20
|
## [0.0.2] - 2024-01-23
|
11
21
|
|
@@ -249,6 +249,7 @@ public class MultiComp {
|
|
249
249
|
private double[] n = null;
|
250
250
|
protected int getK() { return k;}
|
251
251
|
protected int getV() { return v;}
|
252
|
+
|
252
253
|
public double[][] calcTestStatistic(double[][] xi) {
|
253
254
|
k = xi.length;
|
254
255
|
mean = new double[k];
|
@@ -267,8 +268,10 @@ public class MultiComp {
|
|
267
268
|
private double calcVe(double[][] xi) {
|
268
269
|
double sumSq = 0.0;
|
269
270
|
int sumN = 0;
|
271
|
+
|
270
272
|
for(int i = 0; i < k; i++) {
|
271
273
|
DescriptiveStatistics stat = new DescriptiveStatistics();
|
274
|
+
|
272
275
|
Arrays.stream(xi[i]).forEach(stat::addValue);
|
273
276
|
mean[i] = stat.getMean();
|
274
277
|
n[i] = stat.getN();
|
@@ -286,7 +289,7 @@ public class MultiComp {
|
|
286
289
|
int v = super.getV();
|
287
290
|
int k = super.getK();
|
288
291
|
double den = k - 1;
|
289
|
-
|
292
|
+
|
290
293
|
TDistribution tDist = new TDistribution(v);
|
291
294
|
double l_val = tDist.inverseCumulativeProbability(a / den);
|
292
295
|
double r_val = tDist.inverseCumulativeProbability(1.0 - a / den);
|
@@ -28,6 +28,7 @@ import java.util.ArrayList;
|
|
28
28
|
import org.apache.commons.math3.stat.inference.OneWayAnova;
|
29
29
|
import org.apache.commons.math3.stat.descriptive.DescriptiveStatistics;
|
30
30
|
import org.apache.commons.math3.distribution.ChiSquaredDistribution;
|
31
|
+
import org.apache.commons.math3.distribution.FDistribution;
|
31
32
|
import java.util.Map;
|
32
33
|
public class OneWayLayout {
|
33
34
|
private static OneWayLayout oneWay = new OneWayLayout();
|
@@ -57,7 +58,19 @@ public class OneWayLayout {
|
|
57
58
|
OneWayAnovaTest oneway = new BartletTest();
|
58
59
|
|
59
60
|
double statistic = oneway.calcTestStatistic(xi);
|
60
|
-
return oneway.
|
61
|
+
return oneway.execute_test(statistic, a);
|
62
|
+
}
|
63
|
+
public void replicatePlot(String dname, Map<String, double[]> vals) {
|
64
|
+
ChartPlot plot = new ReplicateChartPlot();
|
65
|
+
|
66
|
+
JFreeChart chart = plot.createChart("反復測定", dname, vals);
|
67
|
+
plot.writeJPEG("replicate.jpeg", chart, 800, 500);
|
68
|
+
}
|
69
|
+
public boolean replicateTest(double[][] xi, double a) {
|
70
|
+
OneWayAnovaTest oneway = new ReplicateTest();
|
71
|
+
|
72
|
+
double statistic = oneway.calcTestStatistic(xi);
|
73
|
+
return oneway.execute_test(statistic, a);
|
61
74
|
}
|
62
75
|
/*********************************/
|
63
76
|
/* interface define */
|
@@ -78,7 +91,7 @@ public class OneWayLayout {
|
|
78
91
|
}
|
79
92
|
private interface OneWayAnovaTest {
|
80
93
|
double calcTestStatistic(double[][] xi);
|
81
|
-
boolean
|
94
|
+
boolean execute_test(double statistic, double a);
|
82
95
|
}
|
83
96
|
/*********************************/
|
84
97
|
/* class define */
|
@@ -196,6 +209,7 @@ public class OneWayLayout {
|
|
196
209
|
}
|
197
210
|
}
|
198
211
|
}
|
212
|
+
// バートレット検定
|
199
213
|
private class BartletTest implements OneWayAnovaTest {
|
200
214
|
private int n = 0;
|
201
215
|
public double calcTestStatistic(double[][] xi) {
|
@@ -226,6 +240,7 @@ public class OneWayLayout {
|
|
226
240
|
double invSumN = 0.0;
|
227
241
|
int sumN = 0;
|
228
242
|
DescriptiveStatistics stat = new DescriptiveStatistics();
|
243
|
+
|
229
244
|
for(int i = 0; i < n; i++) {
|
230
245
|
Arrays.stream(xi[i]).forEach(stat::addValue);
|
231
246
|
invSumN += 1.0 / (stat.getN() - 1.0);
|
@@ -236,12 +251,134 @@ public class OneWayLayout {
|
|
236
251
|
* (invSumN - 1.0 / (sumN - n));
|
237
252
|
return ln2L / deno;
|
238
253
|
}
|
239
|
-
public boolean
|
254
|
+
public boolean execute_test(double statistic, double a) {
|
240
255
|
ChiSquaredDistribution chi2Dist = new ChiSquaredDistribution(n - 1);
|
241
256
|
double r_val = chi2Dist.inverseCumulativeProbability(1.0 - a);
|
242
257
|
|
243
258
|
return (r_val < statistic) ? true : false;
|
244
259
|
}
|
245
260
|
}
|
261
|
+
// 反復測定Plot
|
262
|
+
private class ReplicateChartPlot implements ChartPlot {
|
263
|
+
public JFreeChart createChart(String title, String dname, Map<String, double[]> vals) {
|
264
|
+
CategoryPlot plot = createPlot(dname, vals);
|
265
|
+
ChartFactory.setChartTheme(StandardChartTheme.createLegacyTheme());
|
266
|
+
JFreeChart chart = new JFreeChart(title, plot);
|
267
|
+
|
268
|
+
ChartUtils.applyCurrentTheme(chart);
|
269
|
+
return chart;
|
270
|
+
}
|
271
|
+
private CategoryPlot createPlot( String dname, Map<String, double[]> vals) {
|
272
|
+
CreatePlot plotImpl = new ReplicatePlot();
|
273
|
+
|
274
|
+
return plotImpl.createPlot(dname, vals);
|
275
|
+
}
|
276
|
+
private class ReplicatePlot implements CreatePlot {
|
277
|
+
public CategoryPlot createPlot(String dname, Map<String, double[]> vals) {
|
278
|
+
LineAndShapeRenderer renderer = new LineAndShapeRenderer(true, true);
|
279
|
+
|
280
|
+
renderer.setDefaultToolTipGenerator(
|
281
|
+
new StandardCategoryToolTipGenerator()
|
282
|
+
);
|
283
|
+
CategoryPlot plot = new CategoryPlot();
|
284
|
+
|
285
|
+
plot.setOrientation(PlotOrientation.VERTICAL);
|
286
|
+
plot.mapDatasetToRangeAxis(0,0);
|
287
|
+
plot.setDatasetRenderingOrder(DatasetRenderingOrder.FORWARD);
|
288
|
+
|
289
|
+
/*--- 横軸 ---*/
|
290
|
+
CategoryAxis categoryAxis = new CategoryAxis("因子");
|
291
|
+
plot.setDomainAxis(categoryAxis);
|
292
|
+
|
293
|
+
/*--- 縦軸 ---*/
|
294
|
+
NumberAxis valueAxis0 = new NumberAxis("推定周辺平均");
|
295
|
+
plot.setRangeAxis(valueAxis0);
|
296
|
+
|
297
|
+
plot.setRenderer(0, renderer);
|
298
|
+
plot.setDataset(0, createDataset(dname, vals));
|
299
|
+
return plot;
|
300
|
+
}
|
301
|
+
private CategoryDataset createDataset(String dname, Map<String, double[]> vals) {
|
302
|
+
DefaultCategoryDataset data = new DefaultCategoryDataset();
|
303
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
304
|
+
|
305
|
+
for(Map.Entry<String, double[]> entry : vals.entrySet()) {
|
306
|
+
double[] v = entry.getValue();
|
307
|
+
|
308
|
+
Arrays.stream(v).forEach(stat::addValue);
|
309
|
+
data.addValue(stat.getMean(), dname, entry.getKey());
|
310
|
+
stat.clear();
|
311
|
+
}
|
312
|
+
return data;
|
313
|
+
}
|
314
|
+
}
|
315
|
+
}
|
316
|
+
private class ReplicateTest implements OneWayAnovaTest {
|
317
|
+
private int a1 = 0;
|
318
|
+
private int b1 = 0;
|
319
|
+
public double calcTestStatistic(double[][] xi) {
|
320
|
+
b1 = xi[0].length;
|
321
|
+
a1 = xi.length;
|
322
|
+
double st = calcSt(xi, a1, b1);
|
323
|
+
double sa = calcSa(xi, a1, b1);
|
324
|
+
double sb = calcSb(xi, a1, b1);
|
325
|
+
double se = st - sa - sb;
|
326
|
+
double meanSa = sa / (a1 - 1);
|
327
|
+
double meanSe = se / ((a1 - 1) * (b1 - 1));
|
328
|
+
|
329
|
+
return meanSa / meanSe ;
|
330
|
+
}
|
331
|
+
private double calcSt(double[][] xi, int a, int b) {
|
332
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
333
|
+
double sumSt1 = 0.0;
|
334
|
+
double sumSt2 = 0.0;
|
335
|
+
|
336
|
+
for(int i = 0; i < a; i++) {
|
337
|
+
Arrays.stream(xi[i]).forEach(stat::addValue);
|
338
|
+
sumSt1 += stat.getSumsq();
|
339
|
+
sumSt2 += stat.getSum();
|
340
|
+
stat.clear();
|
341
|
+
}
|
342
|
+
return sumSt1 - sumSt2 * sumSt2 / (a * b);
|
343
|
+
}
|
344
|
+
private double calcSa(double[][] xi, int a, int b) {
|
345
|
+
double sumSa1 = 0.0;
|
346
|
+
double sumSa2 = 0.0;
|
347
|
+
double[] an = new double[a];
|
348
|
+
|
349
|
+
for(int i = 0; i < a; i++) {
|
350
|
+
for(int j = 0; j < b; j++) {
|
351
|
+
an[i] += xi[i][j];
|
352
|
+
sumSa2 += xi[i][j];
|
353
|
+
}
|
354
|
+
}
|
355
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
356
|
+
Arrays.stream(an).forEach(stat::addValue);
|
357
|
+
sumSa1 = stat.getSumsq() / b;
|
358
|
+
return sumSa1 - sumSa2 * sumSa2 / (a * b);
|
359
|
+
}
|
360
|
+
private double calcSb(double[][] xi, int a, int b) {
|
361
|
+
double[] bn = new double[b];
|
362
|
+
double sumSb1 = 0.0;
|
363
|
+
double sumSb2 = 0.0;
|
364
|
+
|
365
|
+
for(int i = 0; i < a; i++) {
|
366
|
+
for(int j = 0; j < b; j++) {
|
367
|
+
bn[j] += xi[i][j];
|
368
|
+
sumSb2 += xi[i][j];
|
369
|
+
}
|
370
|
+
}
|
371
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
372
|
+
Arrays.stream(bn).forEach(stat::addValue);
|
373
|
+
sumSb1 = stat.getSumsq() / a;
|
374
|
+
return sumSb1 - sumSb2 * sumSb2 / (a * b);
|
375
|
+
}
|
376
|
+
public boolean execute_test(double statistic, double a) {
|
377
|
+
FDistribution fDist = new FDistribution(a1 - 1, (a1 - 1) * (b1 - 1));
|
378
|
+
double f = fDist.inverseCumulativeProbability(1.0 - a);
|
379
|
+
|
380
|
+
return (statistic >= f) ? true : false;
|
381
|
+
}
|
382
|
+
}
|
246
383
|
}
|
247
384
|
|
@@ -0,0 +1,186 @@
|
|
1
|
+
import org.apache.commons.math3.stat.descriptive.DescriptiveStatistics;
|
2
|
+
import org.apache.commons.math3.distribution.FDistribution;
|
3
|
+
import java.util.Arrays;
|
4
|
+
public class TwoWayLayout {
|
5
|
+
private static TwoWayLayout twoWay = new TwoWayLayout();
|
6
|
+
public static TwoWayLayout getInstance() {
|
7
|
+
return twoWay;
|
8
|
+
}
|
9
|
+
public boolean[] twowayAnova(double[][][] xij, double a) {
|
10
|
+
TwoWayAnovaTest twoway = new TwoWayAnova();
|
11
|
+
|
12
|
+
double[] statistic = twoway.calcTestStatistic(xij);
|
13
|
+
return twoway.execute_test(statistic, a);
|
14
|
+
}
|
15
|
+
/*********************************/
|
16
|
+
/* interface define */
|
17
|
+
/*********************************/
|
18
|
+
private interface TwoWayAnovaTest {
|
19
|
+
double[] calcTestStatistic(double[][][] xij);
|
20
|
+
boolean[] execute_test(double statistic[], double a);
|
21
|
+
}
|
22
|
+
/*********************************/
|
23
|
+
/* class define */
|
24
|
+
/*********************************/
|
25
|
+
private class TwoWayAnova implements TwoWayAnovaTest {
|
26
|
+
private int a = 0;
|
27
|
+
private int b = 0;
|
28
|
+
private int n = 0;
|
29
|
+
private int an = 0;
|
30
|
+
private int bn = 0;
|
31
|
+
private int abn = 0;
|
32
|
+
private int en = 0;
|
33
|
+
public double[] calcTestStatistic(double[][][] xij) {
|
34
|
+
double statistic[] = new double[3];
|
35
|
+
a = xij.length;
|
36
|
+
b = xij[0].length;
|
37
|
+
n = xij[0][0].length;
|
38
|
+
an = a- 1;
|
39
|
+
bn = b - 1;
|
40
|
+
abn = (a- 1) * (b - 1);
|
41
|
+
en = a * b * (n - 1);
|
42
|
+
|
43
|
+
double[][] meanXij = calcMeanXij(xij);
|
44
|
+
double[] meanAn = calcMeanAn(meanXij);
|
45
|
+
double[] meanBn = calcMeanBn(meanXij);
|
46
|
+
double meanABn = calcMeanABn(meanAn);
|
47
|
+
|
48
|
+
double allDrift = calcAllDrift(xij, meanABn); // 全変動
|
49
|
+
double anDrift = calcAnDrift(meanAn, meanABn); // 水準Ai間変動
|
50
|
+
double bnDrift = calcBnDrift(meanBn, meanABn); // 水準Bj間変動
|
51
|
+
// 交互作用の変動
|
52
|
+
double interaDrift = calcInteraDrift(meanXij, meanAn, meanBn, meanABn);
|
53
|
+
double benchDrift = calcBenchDrift(xij, meanXij); // 水準内変動
|
54
|
+
double va = b * n * anDrift / an;
|
55
|
+
double vb = a * n * bnDrift / bn;
|
56
|
+
double vab = n * interaDrift / abn;
|
57
|
+
double ve = benchDrift / en;
|
58
|
+
|
59
|
+
statistic[0] = va / ve;
|
60
|
+
statistic[1] = vb / ve;
|
61
|
+
statistic[2] = vab/ ve;
|
62
|
+
return statistic;
|
63
|
+
}
|
64
|
+
private double[][] calcMeanXij(double[][][] xij) {
|
65
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
66
|
+
double[][] meanXij = new double[a][b];
|
67
|
+
|
68
|
+
for(int i = 0; i < a; i++) {
|
69
|
+
for(int j = 0; j < b; j++) {
|
70
|
+
Arrays.stream(xij[i][j]).forEach(stat::addValue);
|
71
|
+
meanXij[i][j] = stat.getMean();
|
72
|
+
stat.clear();
|
73
|
+
}
|
74
|
+
}
|
75
|
+
return meanXij;
|
76
|
+
}
|
77
|
+
private double[] calcMeanAn(double[][] meanXij) {
|
78
|
+
double[] an = new double[a];
|
79
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
80
|
+
|
81
|
+
for(int i = 0; i < a; i++) {
|
82
|
+
double sumSa = 0.0;
|
83
|
+
for(int j = 0; j < b; j++) {
|
84
|
+
sumSa += meanXij[i][j];
|
85
|
+
}
|
86
|
+
an[i] = sumSa / b;
|
87
|
+
}
|
88
|
+
return an;
|
89
|
+
}
|
90
|
+
private double[] calcMeanBn(double[][] meanXij) {
|
91
|
+
double[] bn = new double[b];
|
92
|
+
double[] sumA = new double[b];
|
93
|
+
|
94
|
+
for(int i = 0; i < a; i++) {
|
95
|
+
for(int j = 0; j < b; j++) {
|
96
|
+
bn[j] += meanXij[i][j] / a;
|
97
|
+
}
|
98
|
+
}
|
99
|
+
return bn;
|
100
|
+
}
|
101
|
+
private double calcMeanABn(double[] meanAn) {
|
102
|
+
DescriptiveStatistics stat = new DescriptiveStatistics();
|
103
|
+
|
104
|
+
Arrays.stream(meanAn).forEach(stat::addValue);
|
105
|
+
return stat.getMean();
|
106
|
+
}
|
107
|
+
// 全変動
|
108
|
+
private double calcAllDrift(double[][][] xij, double meanABn) {
|
109
|
+
double sumDrift = 0.0;
|
110
|
+
|
111
|
+
for(int i = 0; i < a; i++) {
|
112
|
+
for(int j = 0; j < b; j++) {
|
113
|
+
for(int k = 0; k < xij[i][j].length; k++) {
|
114
|
+
double diffXijk = xij[i][j][k] - meanABn;
|
115
|
+
sumDrift += diffXijk * diffXijk;
|
116
|
+
}
|
117
|
+
}
|
118
|
+
}
|
119
|
+
return sumDrift;
|
120
|
+
}
|
121
|
+
// 水準Ai間変動
|
122
|
+
private double calcAnDrift(double[] meanAn, double meanABn) {
|
123
|
+
double sumDrift = 0.0;
|
124
|
+
|
125
|
+
for(int i =0; i < meanAn.length; i++) {
|
126
|
+
double diffXi = meanAn[i] - meanABn;
|
127
|
+
|
128
|
+
sumDrift += diffXi * diffXi;
|
129
|
+
}
|
130
|
+
return sumDrift;
|
131
|
+
}
|
132
|
+
// 水準Bj間変動
|
133
|
+
private double calcBnDrift(double[] meanBn, double meanABn) {
|
134
|
+
double sumDrift = 0.0;
|
135
|
+
|
136
|
+
for(int j = 0; j < meanBn.length; j++) {
|
137
|
+
double diffXj = meanBn[j] - meanABn;
|
138
|
+
|
139
|
+
sumDrift += diffXj * diffXj;
|
140
|
+
}
|
141
|
+
return sumDrift;
|
142
|
+
}
|
143
|
+
// 交互作用の変動
|
144
|
+
private double calcInteraDrift(double[][] meanXij, double[] meanAn, double[] meanBn, double meanABn) {
|
145
|
+
double sumDrift = 0.0;
|
146
|
+
|
147
|
+
for(int i = 0; i< a; i++) {
|
148
|
+
for(int j = 0; j < b; j++) {
|
149
|
+
double diffXj = meanXij[i][j] - meanAn[i] - meanBn[j] + meanABn;
|
150
|
+
|
151
|
+
sumDrift += diffXj * diffXj;
|
152
|
+
}
|
153
|
+
}
|
154
|
+
return sumDrift;
|
155
|
+
}
|
156
|
+
// 水準内変動
|
157
|
+
private double calcBenchDrift(double[][][] xij, double[][] meanXij) {
|
158
|
+
double sumDrift = 0.0;
|
159
|
+
|
160
|
+
for(int i = 0; i < a; i++) {
|
161
|
+
for(int j = 0; j < b; j++) {
|
162
|
+
for(int k = 0; k < xij[i][j].length; k++) {
|
163
|
+
double diffXj = xij[i][j][k] - meanXij[i][j];
|
164
|
+
|
165
|
+
sumDrift += diffXj * diffXj;
|
166
|
+
}
|
167
|
+
}
|
168
|
+
}
|
169
|
+
return sumDrift;
|
170
|
+
}
|
171
|
+
public boolean[] execute_test(double statistic[], double a) {
|
172
|
+
boolean[] ret = new boolean[3];
|
173
|
+
|
174
|
+
ret[0] = evaluation(new FDistribution(an, en), statistic[0], a);
|
175
|
+
ret[1] = evaluation(new FDistribution(bn, en), statistic[1], a);
|
176
|
+
ret[2] = evaluation(new FDistribution(abn, en), statistic[2], a);
|
177
|
+
return ret;
|
178
|
+
}
|
179
|
+
private boolean evaluation(FDistribution fDist, double statistic, double a) {
|
180
|
+
double r_val = fDist.inverseCumulativeProbability(1.0 - a);
|
181
|
+
|
182
|
+
return (statistic < r_val) ? false : true;
|
183
|
+
}
|
184
|
+
}
|
185
|
+
}
|
186
|
+
|
data/lib/num4anova.rb
CHANGED
@@ -4,6 +4,7 @@ require 'jfreechart-1.5.4.jar'
|
|
4
4
|
require 'commons-math3-3.6.1.jar'
|
5
5
|
|
6
6
|
java_import 'OneWayLayout'
|
7
|
+
java_import 'TwoWayLayout'
|
7
8
|
java_import 'java.util.HashMap'
|
8
9
|
# 分散分析を行う
|
9
10
|
# (Apache commoms math3使用)
|
@@ -72,13 +73,13 @@ module Num4AnovaLib
|
|
72
73
|
# @param [double] a 有意水準
|
73
74
|
# @return [boolean] 検定結果(true:棄却域内 false:棄却域外)
|
74
75
|
# @example
|
75
|
-
#
|
76
|
-
#
|
77
|
-
#
|
78
|
-
#
|
79
|
-
#
|
80
|
-
#
|
81
|
-
#
|
76
|
+
# xi = [
|
77
|
+
# [12.2, 18.8, 18.2],
|
78
|
+
# [22.2, 20.5, 14.6],
|
79
|
+
# [20.8, 19.5, 26.3],
|
80
|
+
# [26.4, 32.5, 31.3],
|
81
|
+
# [24.5, 21.2, 22.4],
|
82
|
+
# ]
|
82
83
|
# oneWay = Num4AnovaLib::OneWayLayoutLib.new
|
83
84
|
# oneWay.oneWay.oneway_anova(xi, 0.05)
|
84
85
|
# => true
|
@@ -92,19 +93,107 @@ module Num4AnovaLib
|
|
92
93
|
# @param [double] a 有意水準
|
93
94
|
# @return [boolean] 検定結果(true:棄却域内 false:棄却域外)
|
94
95
|
# @example
|
95
|
-
#
|
96
|
-
#
|
97
|
-
#
|
98
|
-
#
|
99
|
-
#
|
100
|
-
#
|
101
|
-
#
|
96
|
+
# xi = [
|
97
|
+
# [12.2, 18.8, 18.2],
|
98
|
+
# [22.2, 20.5, 14.6],
|
99
|
+
# [20.8, 19.5, 26.3],
|
100
|
+
# [26.4, 32.5, 31.3],
|
101
|
+
# [24.5, 21.2, 22.4],
|
102
|
+
# ]
|
102
103
|
# oneWay = Num4AnovaLib::OneWayLayoutLib.new
|
103
104
|
# oneWay.bartlet(xi, 0.05)
|
104
105
|
# => true
|
105
106
|
def bartlet(xi, a)
|
106
107
|
return @oneWay.bartletTest(xi.to_java(Java::double[]), a)
|
107
108
|
end
|
109
|
+
# 反復測定Plot
|
110
|
+
#
|
111
|
+
# @overload replicate_plot(dname, vals)
|
112
|
+
# @param [String] dname データ名
|
113
|
+
# @param [Hash] vals Hash(String, double[])
|
114
|
+
# @return [void] replicate.jpegファイルを出力
|
115
|
+
# @example
|
116
|
+
# vals = {
|
117
|
+
# "stageA1" => [27, 52, 18, 21, 32],
|
118
|
+
# "stageA2" => [52, 72, 31, 50, 45],
|
119
|
+
# "stageA3" => [47, 54, 29, 43, 32],
|
120
|
+
# "stageA4" => [28, 50, 22, 26, 29],
|
121
|
+
# }
|
122
|
+
# oneWay = Num4AnovaLib::OneWayLayoutLib.new
|
123
|
+
# oneWay.replicate_plot("LDH", vals)
|
124
|
+
# => replicate.jpeg
|
125
|
+
# @note
|
126
|
+
# グラフは、jfreechartを使用
|
127
|
+
def replicate_plot(dname, vals)
|
128
|
+
o = HashMap.new
|
129
|
+
vals.each{|k, v|
|
130
|
+
o[k] = v.to_java(Java::double)
|
131
|
+
}
|
132
|
+
return @oneWay.replicatePlot(dname, o)
|
133
|
+
end
|
134
|
+
# 反復測定検定
|
135
|
+
#
|
136
|
+
# @overload replicate_test(xi, a)
|
137
|
+
# @param [array] xi データ(double[][])
|
138
|
+
# @param [double] a 有意水準
|
139
|
+
# @return [boolean] 検定結果(true:棄却域内 false:棄却域外)
|
140
|
+
# @example
|
141
|
+
# xi = [
|
142
|
+
# [27, 52, 18, 21, 32],
|
143
|
+
# [52, 72, 31, 50, 45],
|
144
|
+
# [47, 54, 29, 43, 32],
|
145
|
+
# [28, 50, 22, 26, 29],
|
146
|
+
# ]
|
147
|
+
# oneWay = Num4AnovaLib::OneWayLayoutLib.new
|
148
|
+
# oneWay.replicate_test(xi, 0.05)
|
149
|
+
# => true
|
150
|
+
def replicate_test(xi, a)
|
151
|
+
return @oneWay.replicateTest(xi.to_java(Java::double[]), a)
|
152
|
+
end
|
153
|
+
end
|
154
|
+
|
155
|
+
# 二元配置の分散分析
|
156
|
+
class TwoWayLayoutLib
|
157
|
+
def initialize
|
158
|
+
@twoWay = TwoWayLayout.getInstance()
|
159
|
+
end
|
160
|
+
# 二元配置の分散分析
|
161
|
+
#
|
162
|
+
# @overload twoway_anova(xij, a)
|
163
|
+
# @param [array] xij データ(double[][][])
|
164
|
+
# @param [double] a 有意水準
|
165
|
+
# @return [Array] 検定結果(boolean[] true:棄却域内 false:棄却域外)
|
166
|
+
# @example
|
167
|
+
# xij = [
|
168
|
+
# [
|
169
|
+
# [13.2, 15.7, 11.9],
|
170
|
+
# [16.1, 15.7, 15.1],
|
171
|
+
# [9.1, 10.3, 8.2],
|
172
|
+
# ],
|
173
|
+
# [
|
174
|
+
# [22.8, 25.7, 18.5],
|
175
|
+
# [24.5, 21.2, 24.2],
|
176
|
+
# [11.9, 14.3, 13.7],
|
177
|
+
# ],
|
178
|
+
# [
|
179
|
+
# [21.8, 26.3, 32.1],
|
180
|
+
# [26.9, 31.3, 28.3],
|
181
|
+
# [15.1, 13.6, 16.2],
|
182
|
+
# ],
|
183
|
+
# [
|
184
|
+
# [25.7, 28.8, 29.5],
|
185
|
+
# [30.1, 33.8, 29.6],
|
186
|
+
# [15.2, 17.3, 14.8],
|
187
|
+
# ],
|
188
|
+
# ]
|
189
|
+
# twoWay = Num4AnovaLib::TwoWayLayoutLib.new
|
190
|
+
# twoWay.twoway_anova(xij, 0.05)
|
191
|
+
# =>
|
192
|
+
# [true, true, true]
|
193
|
+
def twoway_anova(xij, a)
|
194
|
+
ret = @twoWay.twowayAnova(xij.to_java(Java::double[][]), a)
|
195
|
+
return ret.to_a
|
196
|
+
end
|
108
197
|
end
|
109
198
|
end
|
110
199
|
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: num4anova
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0.
|
4
|
+
version: 0.0.5
|
5
5
|
platform: java
|
6
6
|
authors:
|
7
7
|
- siranovel
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2024-01
|
11
|
+
date: 2024-02-01 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rake
|
@@ -63,6 +63,7 @@ files:
|
|
63
63
|
- Rakefile
|
64
64
|
- ext/num4anova/MultiComp.java
|
65
65
|
- ext/num4anova/OneWayLayout.java
|
66
|
+
- ext/num4anova/TwoWayLayout.java
|
66
67
|
- lib/commons-math3-3.6.1.jar
|
67
68
|
- lib/dunnet.rb
|
68
69
|
- lib/jcommon-1.0.23.jar
|