mirror of https://github.com/jlizier/jidt
466 lines
16 KiB
Java
Executable File
466 lines
16 KiB
Java
Executable File
/*
|
|
* Java Information Dynamics Toolkit (JIDT)
|
|
* Copyright (C) 2012, Joseph T. Lizier
|
|
*
|
|
* This program is free software: you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License as published by
|
|
* the Free Software Foundation, either version 3 of the License, or
|
|
* (at your option) any later version.
|
|
*
|
|
* This program is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
* GNU General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public License
|
|
* along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
|
|
package infodynamics.measures.continuous;
|
|
|
|
import java.util.Vector;
|
|
|
|
import infodynamics.utils.EmpiricalMeasurementDistribution;
|
|
import infodynamics.utils.MatrixUtils;
|
|
|
|
/**
|
|
* A Predictive Information (PI) / Excess Entropy calculator
|
|
* (implementing {@link PredictiveInfoCalculator})
|
|
* which is affected using a
|
|
* given Mutual Information (MI) calculator (implementing
|
|
* {@link MutualInfoCalculatorMultiVariate}) to make the calculations.
|
|
*
|
|
* <p>Usage is as per the paradigm outlined for {@link PredictiveInfoCalculator},
|
|
* except that in the constructor(s) for this class the implementation for
|
|
* a {@link MutualInfoCalculatorMultiVariate} must be supplied.
|
|
* </p>
|
|
*
|
|
* <p>This class <i>may</i> be used directly, however users are advised that
|
|
* several child classes are available which already plug-in the various MI estimators
|
|
* to provide AIS calculators (taking specific caution associated with
|
|
* each type of estimator):</p>
|
|
* <ul>
|
|
* <li>{@link infodynamics.measures.continuous.gaussian.PredictiveInfoCalculatorGaussian}</li>
|
|
* <li>{@link infodynamics.measures.continuous.kernel.PredictiveInfoCalculatorKernel}</li>
|
|
* <li>{@link infodynamics.measures.continuous.kraskov.PredictiveInfoCalculatorKraskov}</li>
|
|
* </ul>
|
|
*
|
|
* <p><b>References:</b><br/>
|
|
* <ul>
|
|
* <li>Bialek, W., Nemenman, I., and Tishby, N.,
|
|
* <a href="http://dx.doi.org/10.1016/S0378-4371(01)00444-7">
|
|
* "Complexity through nonextensivity"</a>,
|
|
* Physica A, 302, 89-99. (2001).</li>
|
|
* <li>J. P. Crutchfield, D. P. Feldman,
|
|
* <a href="http://dx.doi.org/10.1063/1.1530990">
|
|
* "Regularities Unseen, Randomness Observed: Levels of Entropy Convergence"</a>,
|
|
* Chaos, Vol. 13, No. 1. (2003), pp. 25-54.</li>
|
|
* </ul>
|
|
*
|
|
* @author Joseph Lizier (<a href="joseph.lizier at gmail.com">email</a>,
|
|
* <a href="http://lizier.me/joseph/">www</a>)
|
|
* @see ActiveInfoStorageCalculator
|
|
*/
|
|
public class PredictiveInfoCalculatorViaMutualInfo implements
|
|
PredictiveInfoCalculator {
|
|
|
|
/**
|
|
* The underlying mutual information calculator
|
|
*/
|
|
protected MutualInfoCalculatorMultiVariate miCalc;
|
|
/**
|
|
* Length of past and future vectors to consider (embedding length)
|
|
*/
|
|
protected int k = 1;
|
|
/**
|
|
* Embedding delay to use between elements of the embedding vector.
|
|
* We're hard-coding a delay of 1 between the history vector and the future
|
|
* vector however.
|
|
*/
|
|
protected int tau = 1;
|
|
/**
|
|
* Whether debug mode is on
|
|
*/
|
|
protected boolean debug = false;
|
|
|
|
/**
|
|
* Construct using an instantiation of the named MI calculator
|
|
*
|
|
* @param miCalculatorClassName fully qualified class name of the MI calculator to instantiate
|
|
* @throws InstantiationException
|
|
* @throws IllegalAccessException
|
|
* @throws ClassNotFoundException
|
|
*/
|
|
public PredictiveInfoCalculatorViaMutualInfo(String miCalculatorClassName)
|
|
throws InstantiationException, IllegalAccessException, ClassNotFoundException {
|
|
@SuppressWarnings("unchecked")
|
|
Class<MutualInfoCalculatorMultiVariate> miClass =
|
|
(Class<MutualInfoCalculatorMultiVariate>) Class.forName(miCalculatorClassName);
|
|
MutualInfoCalculatorMultiVariate miCalc = miClass.newInstance();
|
|
construct(miCalc);
|
|
}
|
|
|
|
/**
|
|
* Construct using an instantiation of the given MI class
|
|
*
|
|
* @param miCalcClass Class of the MI calculator to instantiate and use
|
|
* @throws InstantiationException
|
|
* @throws IllegalAccessException
|
|
*/
|
|
protected PredictiveInfoCalculatorViaMutualInfo(Class<MutualInfoCalculatorMultiVariate> miCalcClass)
|
|
throws InstantiationException, IllegalAccessException {
|
|
MutualInfoCalculatorMultiVariate miCalc = miCalcClass.newInstance();
|
|
construct(miCalc);
|
|
}
|
|
|
|
/**
|
|
* Construct using the given (constructed but not initialised)
|
|
* MI calculator.
|
|
*
|
|
* @param miCalc MI calculator which is already constructed but
|
|
* there has not been a call to its {@link MutualInfoCalculatorMultiVariate#initialise()}
|
|
* method yet
|
|
*/
|
|
protected PredictiveInfoCalculatorViaMutualInfo(MutualInfoCalculatorMultiVariate miCalc) {
|
|
construct(miCalc);
|
|
}
|
|
|
|
/**
|
|
* Internal routine to execute common code for constructing an instance
|
|
*
|
|
* @param miCalc
|
|
*/
|
|
protected void construct(MutualInfoCalculatorMultiVariate miCalc) {
|
|
this.miCalc = miCalc;
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#initialise()
|
|
*/
|
|
@Override
|
|
public void initialise() throws Exception {
|
|
initialise(k, tau); // Initialise with current value of k
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#initialise(int)
|
|
*/
|
|
@Override
|
|
public void initialise(int k) throws Exception {
|
|
initialise(k, tau);
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
*
|
|
* <p>All child classes <b>must</b> call this routine on this as the super class
|
|
* once they have finished executing their specialised code
|
|
* for their {@link #initialise()} implementations.
|
|
* </p>
|
|
*
|
|
*/
|
|
@Override
|
|
public void initialise(int k, int tau) throws Exception {
|
|
this.k = k;
|
|
this.tau = tau;
|
|
miCalc.initialise(k, k);
|
|
}
|
|
|
|
/**
|
|
* Set properties for the underlying calculator implementation.
|
|
* New property values are not guaranteed to take effect until the next call
|
|
* to an initialise method.
|
|
*
|
|
* <p>Allowable property names include:</p>
|
|
* <ul>
|
|
* <li>Those defined for the {@link PredictiveInfoCalculator} interface
|
|
* (i.e. {@link #K_PROP_NAME} or {@link K_EMBEDDING_PROP_NAME}
|
|
* or {@link #TAU_PROP_NAME})</li>
|
|
* <li>Any properties defined for the underlying
|
|
* {@link MutualInfoCalculatorMultiVariate#setProperty(String, String)} implementation,
|
|
* <b>however</b> the user is <b>not</b> allowed to set the property
|
|
* {@link MutualInfoCalculatorMultiVariate#PROP_TIME_DIFF} here.
|
|
* This would set a time difference from the history vector to the next
|
|
* step, which we currently do not allow.
|
|
* (If we change our mind one day and allow it, we could implement
|
|
* it simply by letting the time diff property be set here).</li>
|
|
* </ul>
|
|
*
|
|
* <p>Note that implementing classes may defined additional properties.</p>
|
|
*
|
|
* @param propertyName name of the property
|
|
* @param propertyValue value of the property
|
|
* @throws Exception for invalid property values
|
|
*/
|
|
@Override
|
|
public void setProperty(String propertyName, String propertyValue)
|
|
throws Exception {
|
|
if (propertyName.equalsIgnoreCase(MutualInfoCalculatorMultiVariate.PROP_TIME_DIFF)) {
|
|
throw new Exception("Cannot set " + MutualInfoCalculatorMultiVariate.PROP_TIME_DIFF
|
|
+ " property on the PredictiveInfoCalculator");
|
|
}
|
|
boolean propertySet = true;
|
|
if (propertyName.equalsIgnoreCase(K_PROP_NAME) ||
|
|
propertyName.equalsIgnoreCase(K_EMBEDDING_PROP_NAME)) {
|
|
k = Integer.parseInt(propertyValue);
|
|
} else if (propertyName.equalsIgnoreCase(TAU_PROP_NAME)) {
|
|
tau = Integer.parseInt(propertyValue);
|
|
} else {
|
|
// No property was set on this class, assume it is for the underlying
|
|
// MI calculator
|
|
miCalc.setProperty(propertyName, propertyValue);
|
|
propertySet = false;
|
|
}
|
|
if (debug && propertySet) {
|
|
System.out.println(this.getClass().getSimpleName() + ": Set property " + propertyName +
|
|
" to " + propertyValue);
|
|
}
|
|
}
|
|
|
|
@Override
|
|
public String getProperty(String propertyName) throws Exception {
|
|
if (propertyName.equalsIgnoreCase(K_PROP_NAME) ||
|
|
propertyName.equalsIgnoreCase(K_EMBEDDING_PROP_NAME)) {
|
|
return Integer.toString(k);
|
|
} else if (propertyName.equalsIgnoreCase(TAU_PROP_NAME)) {
|
|
return Integer.toString(tau);
|
|
} else {
|
|
// No property was set on this class, assume it is for the underlying
|
|
// MI calculator, even if it is for
|
|
// MutualInfoCalculatorMultiVariate.PROP_TIME_DIFF which
|
|
// is not a valid property for the PI calculator:
|
|
return miCalc.getProperty(propertyName);
|
|
}
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#setObservations(double[])
|
|
*/
|
|
@Override
|
|
public void setObservations(double[] observations) throws Exception {
|
|
if (observations.length - 2*(k-1)*tau - 1 <= 0) {
|
|
// There are no observations to add here
|
|
throw new Exception("Not enough observations to set here given k and tau");
|
|
}
|
|
double[][] currentPastVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(observations, k, tau, (k-1)*tau, observations.length - 2*(k-1)*tau - 1);
|
|
double[][] currentNextVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(observations, k, 2*(k-1)*tau + 1, observations.length - 2*(k-1)*tau - 1);
|
|
miCalc.setObservations(currentPastVectors, currentNextVectors);
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#startAddObservations()
|
|
*/
|
|
@Override
|
|
public void startAddObservations() {
|
|
miCalc.startAddObservations();
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#addObservations(double[])
|
|
*/
|
|
@Override
|
|
public void addObservations(double[] observations) throws Exception {
|
|
if (observations.length - 2*(k-1)*tau - 1 <= 0) {
|
|
// There are no observations to add here
|
|
// Don't throw an exception, do nothing since more observations
|
|
// can be added later.
|
|
return;
|
|
}
|
|
double[][] currentPastVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(observations, k, tau, (k-1)*tau, observations.length - 2*(k-1)*tau - 1);
|
|
double[][] currentNextVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(observations, 1, 2*(k-1)*tau + 1, observations.length - 2*(k-1)*tau - 1);
|
|
miCalc.addObservations(currentPastVectors, currentNextVectors);
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#addObservations(double[], int, int)
|
|
*/
|
|
@Override
|
|
public void addObservations(double[] observations, int startTime,
|
|
int numTimeSteps) throws Exception {
|
|
addObservations(MatrixUtils.select(observations, startTime, numTimeSteps));
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#finaliseAddObservations()
|
|
*/
|
|
@Override
|
|
public void finaliseAddObservations() throws Exception {
|
|
miCalc.finaliseAddObservations();
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#setObservations(double[], boolean[])
|
|
*/
|
|
@Override
|
|
public void setObservations(double[] observations, boolean[] valid)
|
|
throws Exception {
|
|
Vector<int[]> startAndEndTimePairs = computeStartAndEndTimePairs(valid);
|
|
|
|
// We've found the set of start and end times for this pair
|
|
startAddObservations();
|
|
for (int[] timePair : startAndEndTimePairs) {
|
|
int startTime = timePair[0];
|
|
int endTime = timePair[1];
|
|
addObservations(observations, startTime, endTime - startTime + 1);
|
|
}
|
|
finaliseAddObservations();
|
|
}
|
|
|
|
/**
|
|
* Compute a vector of start and end pairs of time points, between which we have
|
|
* valid series of observations.
|
|
*
|
|
* <p>This method is made public so it can be used if one wants to compute the number of
|
|
* observations prior to making a call to {@link #setObservations(double[], boolean[])}.</p>
|
|
*
|
|
* @param valid a time series (with indices the same as observations)
|
|
* indicating whether the entry in observations at that index is valid;
|
|
* we only take vectors as samples to add to the observation set where
|
|
* all points in the time series (even between points in
|
|
* the embedded k-vectors with embedding delays) are valid.
|
|
* @return a vector for start and end time pairs of valid series
|
|
* of observations (as defined by <code>valid</code>).
|
|
*/
|
|
public Vector<int[]> computeStartAndEndTimePairs(boolean[] valid) {
|
|
// Scan along the data avoiding invalid values
|
|
int startTime = 0;
|
|
int endTime = 0;
|
|
boolean lookingForStart = true;
|
|
Vector<int[]> startAndEndTimePairs = new Vector<int[]>();
|
|
for (int t = 0; t < valid.length; t++) {
|
|
if (lookingForStart) {
|
|
// Precondition: startTime holds a candidate start time
|
|
if (valid[t]) {
|
|
// This point is OK at the destination
|
|
if (t - startTime < 2*(k-1)*tau+1) {
|
|
// We're still checking the past history only, so
|
|
continue;
|
|
} else {
|
|
// We've got the full past history ok
|
|
// set a candidate endTime
|
|
endTime = t;
|
|
lookingForStart = false;
|
|
if (t == valid.length - 1) {
|
|
// we need to terminate now
|
|
int[] timePair = new int[2];
|
|
timePair[0] = startTime;
|
|
timePair[1] = endTime;
|
|
startAndEndTimePairs.add(timePair);
|
|
// System.out.printf("t_s=%d, t_e=%d\n", startTime, endTime);
|
|
}
|
|
}
|
|
} else {
|
|
// We need to keep looking.
|
|
// Move the potential start time to the next point
|
|
startTime = t + 1;
|
|
}
|
|
} else {
|
|
// Precondition: startTime holds the start time for this set,
|
|
// endTime holds a candidate end time
|
|
// Check if we can include the current time step
|
|
boolean terminateSequence = false;
|
|
if (valid[t]) {
|
|
// We can extend
|
|
endTime = t;
|
|
} else {
|
|
terminateSequence = true;
|
|
}
|
|
if (t == valid.length - 1) {
|
|
// we need to terminate the sequence anyway
|
|
terminateSequence = true;
|
|
}
|
|
if (terminateSequence) {
|
|
// This section is done
|
|
int[] timePair = new int[2];
|
|
timePair[0] = startTime;
|
|
timePair[1] = endTime;
|
|
startAndEndTimePairs.add(timePair);
|
|
// System.out.printf("t_s=%d, t_e=%d\n", startTime, endTime);
|
|
lookingForStart = true;
|
|
startTime = t + 1;
|
|
}
|
|
}
|
|
}
|
|
return startAndEndTimePairs;
|
|
}
|
|
|
|
@Override
|
|
public double computeAverageLocalOfObservations() throws Exception {
|
|
return miCalc.computeAverageLocalOfObservations();
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#computeLocalOfPreviousObservations()
|
|
*/
|
|
@Override
|
|
public double[] computeLocalOfPreviousObservations() throws Exception {
|
|
double[] local = miCalc.computeLocalOfPreviousObservations();
|
|
if (!miCalc.getAddedMoreThanOneObservationSet()) {
|
|
double[] localsToReturn = new double[local.length + 2*(k-1)*tau + 1];
|
|
System.arraycopy(local, 0, localsToReturn, (k-1)*tau + 1, local.length);
|
|
return localsToReturn;
|
|
} else {
|
|
return local;
|
|
}
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.measures.continuous.PredictiveInfoCalculator#computeLocalUsingPreviousObservations(double[])
|
|
*/
|
|
@Override
|
|
public double[] computeLocalUsingPreviousObservations(double[] newObservations) throws Exception {
|
|
if (newObservations.length - 2*(k-1)*tau - 1 <= 0) {
|
|
// There are no observations to compute for here
|
|
return new double[newObservations.length];
|
|
}
|
|
double[][] newPastVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(newObservations, k, tau, (k-1)*tau, newObservations.length - 2*(k-1)*tau - 1);
|
|
double[][] newNextVectors =
|
|
MatrixUtils.makeDelayEmbeddingVector(newObservations, 1, 2*(k-1)*tau + 1, newObservations.length - 2*(k-1)*tau - 1);
|
|
double[] local = miCalc.computeLocalUsingPreviousObservations(newPastVectors, newNextVectors);
|
|
// Pad the front and back of the array with zeros where local PI isn't defined:
|
|
double[] localsToReturn = new double[local.length + 2*(k-1)*tau + 1];
|
|
System.arraycopy(local, 0, localsToReturn, (k-1)*tau + 1, local.length);
|
|
return localsToReturn;
|
|
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.utils.EmpiricalNullDistributionComputer#computeSignificance(int)
|
|
*/
|
|
@Override
|
|
public EmpiricalMeasurementDistribution computeSignificance(
|
|
int numPermutationsToCheck) throws Exception {
|
|
return miCalc.computeSignificance(numPermutationsToCheck);
|
|
}
|
|
|
|
/* (non-Javadoc)
|
|
* @see infodynamics.utils.EmpiricalNullDistributionComputer#computeSignificance(int[][])
|
|
*/
|
|
@Override
|
|
public EmpiricalMeasurementDistribution computeSignificance(
|
|
int[][] newOrderings) throws Exception {
|
|
return miCalc.computeSignificance(newOrderings);
|
|
}
|
|
|
|
@Override
|
|
public void setDebug(boolean debug) {
|
|
this.debug = debug;
|
|
miCalc.setDebug(debug);
|
|
}
|
|
|
|
@Override
|
|
public double getLastAverage() {
|
|
return miCalc.getLastAverage();
|
|
}
|
|
|
|
@Override
|
|
public int getNumObservations() throws Exception {
|
|
return miCalc.getNumObservations();
|
|
}
|
|
}
|