natas_hw
/
jidt
mirror of https://github.com/jlizier/jidt
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Added unit tests for AIS Multivariate Gaussian calculator, including on auto-embedding, and that it functions ok for univariate data.

This commit is contained in:
jlizier 2018-05-16 22:43:46 +10:00
parent dfe7ef0334
commit 4ff4e4bfc9
1 changed files with 123 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

123
java/unittests/infodynamics/measures/continuous/gaussian/ActiveInfoStorageMultiVariateGaussianTester.java Normal file
View File

@ -0,0 +1,123 @@
package infodynamics.measures.continuous.gaussian;
import infodynamics.measures.continuous.ActiveInfoStorageCalculatorViaMutualInfo;
import infodynamics.utils.ArrayFileReader;
import infodynamics.utils.MatrixUtils;
import infodynamics.utils.RandomGenerator;
import junit.framework.TestCase;
public class ActiveInfoStorageMultiVariateGaussianTester extends TestCase {
public void testBasicFunctionality() throws Exception {
ArrayFileReader afr = new ArrayFileReader("demos/data/2coupledRandomCols-1.txt");
double[][] data = afr.getDouble2DMatrix();
int[] ks = {1,2,3,4};
int[] taus = {1,2,3,4};
for (int kIndex = 0; kIndex < ks.length; kIndex++) {
int k = ks[kIndex];
for (int tauIndex = 0; tauIndex < taus.length; tauIndex++) {
int tau = taus[tauIndex];
ActiveInfoStorageCalculatorMultiVariateGaussian ais =
new ActiveInfoStorageCalculatorMultiVariateGaussian();
ais.initialise(data[0].length, k, tau);
ais.setObservations(data);
assertEquals(data.length - (k-1)*tau - 1,
ais.getNumObservations());
@SuppressWarnings("unused")
double result = ais.computeAverageLocalOfObservations();
}
}
}
public void testHandlesOneDimension() throws Exception {
ArrayFileReader afr = new ArrayFileReader("demos/data/2coupledRandomCols-1.txt");
double[][] data = afr.getDouble2DMatrix();
double[] column1 = MatrixUtils.selectColumn(data, 0);
double[][] column1In2D = MatrixUtils.selectColumns(data, new int[] {0});
int[] ks = {1,2,3,4};
int[] taus = {1,2,3,4};
for (int kIndex = 0; kIndex < ks.length; kIndex++) {
int k = ks[kIndex];
for (int tauIndex = 0; tauIndex < taus.length; tauIndex++) {
int tau = taus[tauIndex];
ActiveInfoStorageCalculatorMultiVariateGaussian ais =
new ActiveInfoStorageCalculatorMultiVariateGaussian();
ais.initialise(1, k, tau);
ais.setObservations(column1);
assertEquals(data.length - (k-1)*tau - 1,
ais.getNumObservations());
double resultFromColumn1 = ais.computeAverageLocalOfObservations();
ais.initialise(1, k, tau);
ais.setObservations(column1In2D);
assertEquals(data.length - (k-1)*tau - 1,
ais.getNumObservations());
double resultFrom2D = ais.computeAverageLocalOfObservations();
assertEquals(resultFromColumn1, resultFrom2D, 0.00000001);
}
}
}
public void testAutoEmbeddingAIS() throws Exception {
System.out.println("Start AIS multivariate autoembedding test (Gaussian).");
checkAutoEmbeddingLength(false, false);
checkAutoEmbeddingLength(false, true);
checkAutoEmbeddingLength(true, false);
checkAutoEmbeddingLength(true, true);
}
public void checkAutoEmbeddingLength(boolean source1Has3, boolean source2Has3) throws Exception {
// Generate multivariate data
int dataLength = 5000;
RandomGenerator rg = new RandomGenerator();
double[][] data = rg.generateNormalData(dataLength, 2, 0, 1);
for (int i=3; i < dataLength; i++) {
data[i][0] = 0.2*data[i-2][0] + 0.2*data[i-1][0] + data[i][0];
if (source1Has3) {
data[i][0] += 0.2*data[i-3][0];
}
data[i][1] = 0.2*data[i-2][1] + 0.2*data[i-1][1] + data[i][1];
if (source2Has3) {
data[i][1] += 0.2*data[i-3][1];
}
}
int correctK = (source1Has3 || source2Has3) ? 3 : 2;
// Instantiate calculator and set search bounds
ActiveInfoStorageCalculatorMultiVariateGaussian aisCalc =
new ActiveInfoStorageCalculatorMultiVariateGaussian();
// Can't search larger than the max we expect because fluctuations can make the selection larger
aisCalc.setProperty(ActiveInfoStorageCalculatorViaMutualInfo.PROP_AUTO_EMBED_METHOD,
ActiveInfoStorageCalculatorViaMutualInfo.AUTO_EMBED_METHOD_MAX_CORR_AIS);
// We only check up to the correct k -- selections beyond this are stochastic, so the algorithm
// could choose a larger value if we let it, and that's not wrong.
aisCalc.setProperty(ActiveInfoStorageCalculatorViaMutualInfo.PROP_K_SEARCH_MAX, Integer.toString(correctK));
aisCalc.setProperty(ActiveInfoStorageCalculatorViaMutualInfo.PROP_TAU_SEARCH_MAX, "1");
aisCalc.setDebug(true);
// Run optimisation
aisCalc.initialise(2);
aisCalc.setObservations(data);
int optimisedK = Integer.parseInt(aisCalc.getProperty(ActiveInfoStorageCalculatorViaMutualInfo.K_PROP_NAME));
// Test that answer was correct
assertEquals(correctK, optimisedK);
double estimate = aisCalc.computeAverageLocalOfObservations();
System.out.printf("Computed average was %.5f from %d samples (embedding length optimisedK = %d\n",
estimate, aisCalc.getNumObservations(), optimisedK);
}
}