Merge remote-tracking branch 'origin' into new_sim
# Conflicts: # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/HierarchicalSimulation.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/io/HierarchicalWriter.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/methods/HierarchicalFBASimulator.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/methods/HierarchicalMixedSimulator.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/methods/HierarchicalODERKSimulator.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/methods/HierarchicalSSADirectSimulator.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/model/HierarchicalModel.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/HierarchicalUtilities.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/comp/HierarchicalEventComparator.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/comp/ReplacementHandler.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/CompartmentSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/ConstraintSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/CoreSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/EventSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/InitAssignmentSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/ModelSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/ParameterSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/ReactionSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/ReplacementSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/RuleSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/SpeciesReferenceSetup.java # analysis/src/main/java/edu/utah/ece/async/analysis/simulation/hierarchical/util/setup/SpeciesSetup.java
This commit is contained in:
leandrohw
commit
ae01a4211d
350 changed files with 893 additions and 965 deletions
21
learn/pom.xml
Normal file
21
learn/pom.xml
Normal file
|
|
@ -0,0 +1,21 @@
|
|||
<?xml version="1.0" encoding="UTF-8"?>
|
||||
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
|
||||
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
|
||||
<modelVersion>4.0.0</modelVersion>
|
||||
<artifactId>iBioSim-learn</artifactId>
|
||||
<parent>
|
||||
<groupId>edu.utah.ece.async</groupId>
|
||||
<version>0.0.1-SNAPSHOT</version>
|
||||
<relativePath>../</relativePath>
|
||||
<artifactId>iBioSim</artifactId>
|
||||
</parent>
|
||||
<packaging>jar</packaging>
|
||||
|
||||
<dependencies>
|
||||
<dependency>
|
||||
<groupId>edu.utah.ece.async</groupId>
|
||||
<artifactId>iBioSim-analysis</artifactId>
|
||||
<version>0.0.1-SNAPSHOT</version>
|
||||
</dependency>
|
||||
</dependencies>
|
||||
</project>
|
||||
|
|
@ -0,0 +1,134 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.AMSModel;
|
||||
|
||||
import java.util.*;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class DMVCrun {
|
||||
|
||||
private static int numDMVCruns = 0;
|
||||
|
||||
//private int id;
|
||||
|
||||
private ArrayList<ArrayList<Double>> valueL;
|
||||
|
||||
private ArrayList<Integer> startPoints;
|
||||
|
||||
private ArrayList<Integer> endPoints;
|
||||
|
||||
private ArrayList<Double> avgVals;
|
||||
|
||||
public DMVCrun(){
|
||||
//this.id = DMVCrun.numDMVCruns; // required?? whenever required the absolute id would be different. here it remains the same for all the runs of the same variable.
|
||||
DMVCrun.numDMVCruns = DMVCrun.numDMVCruns + 1;
|
||||
valueL = new ArrayList<ArrayList<Double>>();
|
||||
startPoints = new ArrayList<Integer>();
|
||||
endPoints = new ArrayList<Integer>();
|
||||
avgVals = new ArrayList<Double>();
|
||||
}
|
||||
|
||||
public void addStartPoint(int i){
|
||||
startPoints.add(i);
|
||||
}
|
||||
|
||||
public void addEndPoint(int i){
|
||||
endPoints.add(i);
|
||||
calcAvg();
|
||||
}
|
||||
|
||||
public void calcAvg(){ //for the most recently added set of values by default
|
||||
//ArrayList<Double> vals = this.valueL.get(valueL.size() -1);
|
||||
Double total = 0.0;
|
||||
for (Double d:this.valueL.get(valueL.size() -1)){
|
||||
total+=d;
|
||||
}
|
||||
avgVals.add(total/this.valueL.get(valueL.size() -1).size());
|
||||
}
|
||||
|
||||
public Double[] getAvgVals() { // think.. if entire avgVals or just an indexed one
|
||||
return avgVals.toArray(new Double[avgVals.size()]);
|
||||
}
|
||||
|
||||
public int getNumPoints(){
|
||||
int numPoints = 0;
|
||||
for (int i = 0; i<startPoints.size(); i++){
|
||||
numPoints += ((endPoints.get(i) - startPoints.get(i))+1);
|
||||
}
|
||||
return numPoints;
|
||||
}
|
||||
|
||||
public int getEndPoint(int i) {
|
||||
return endPoints.get(i);
|
||||
}
|
||||
|
||||
public int getStartPoint(int i) {
|
||||
return startPoints.get(i);
|
||||
}
|
||||
|
||||
public void addValue(double d){
|
||||
if (valueL.size() == startPoints.size()){//(this.valueL.isEmpty()){
|
||||
this.valueL.add(valueL.size(),new ArrayList<Double>());
|
||||
}
|
||||
this.valueL.get(valueL.size() -1).add(d);
|
||||
}
|
||||
|
||||
public void removeValue(){ // removes the most recent set of dmv values
|
||||
if (valueL.size() > startPoints.size()){
|
||||
valueL.remove(valueL.size() -1);
|
||||
}
|
||||
}
|
||||
|
||||
public Double getLastValue(){ // removes the most recent set of dmv values
|
||||
Double total = 0.0;
|
||||
for (Double d:this.valueL.get(valueL.size() -1)){
|
||||
total+=d;
|
||||
}
|
||||
return (total/this.valueL.get(valueL.size() -1).size());
|
||||
}
|
||||
|
||||
public void clearAll(){
|
||||
startPoints.clear();
|
||||
endPoints.clear();
|
||||
valueL.clear();
|
||||
avgVals.clear();
|
||||
}
|
||||
|
||||
public Double getConstVal(){ //for the most recently added set of values by default
|
||||
//ArrayList<Double> vals = this.valueL.get(valueL.size() -1);
|
||||
Double total = 0.0;
|
||||
for (Double d:this.valueL.get(valueL.size() -1)){
|
||||
total+=d;
|
||||
}
|
||||
return (total/this.valueL.get(valueL.size() -1).size());
|
||||
}
|
||||
|
||||
/*
|
||||
public boolean belongsToRuns(int i){
|
||||
for (int j = 0; j<startPoints.size(); j++){
|
||||
if ((i>=startPoints.get(j)) && (i<=endPoints.get(j))){
|
||||
return true;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
*/
|
||||
}
|
||||
File diff suppressed because it is too large
Load diff
|
|
@ -0,0 +1,280 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.AMSModel;
|
||||
|
||||
|
||||
//import java.util.*;
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Variable implements Comparable<Variable>{
|
||||
|
||||
private String name;
|
||||
|
||||
private boolean dmvc;
|
||||
|
||||
private boolean input;
|
||||
|
||||
private boolean output;
|
||||
|
||||
private boolean care;
|
||||
|
||||
private boolean interpolate;
|
||||
|
||||
private DMVCrun runs; // tmp
|
||||
|
||||
private Double initValue_vMax,initRate_rMax;
|
||||
|
||||
private Double initValue_vMin,initRate_rMin;
|
||||
|
||||
private String forceType;
|
||||
|
||||
private Double epsilon;
|
||||
|
||||
private boolean destab;
|
||||
|
||||
public Variable(String name){
|
||||
this.name = name;
|
||||
this.runs = new DMVCrun();
|
||||
this.forceType = null;
|
||||
this.care = true;
|
||||
this.epsilon = null; //default
|
||||
this.destab = false;
|
||||
this.interpolate = false;
|
||||
}
|
||||
|
||||
public boolean isDmvc() {
|
||||
return dmvc;
|
||||
}
|
||||
|
||||
public void setDmvc(boolean dmvc) {
|
||||
this.dmvc = dmvc;
|
||||
}
|
||||
|
||||
public boolean isInput() {
|
||||
return input;
|
||||
}
|
||||
public void setInput(boolean input) {
|
||||
this.input = input;
|
||||
}
|
||||
|
||||
public boolean isCare() {
|
||||
return care;
|
||||
}
|
||||
|
||||
public void setCare(boolean care) {
|
||||
this.care = care;
|
||||
}
|
||||
|
||||
public boolean isInterpolate() {
|
||||
return interpolate;
|
||||
}
|
||||
|
||||
public void setInterpolate(boolean interpolate) {
|
||||
this.interpolate = interpolate;
|
||||
}
|
||||
|
||||
public String getName() {
|
||||
return name;
|
||||
}
|
||||
|
||||
// public void setName(String name) {
|
||||
// this.name = name;
|
||||
// }
|
||||
|
||||
public boolean isOutput() {
|
||||
return output;
|
||||
}
|
||||
|
||||
public void setOutput(boolean output) {
|
||||
this.output = output;
|
||||
}
|
||||
|
||||
public DMVCrun getRuns() {
|
||||
return runs;
|
||||
}
|
||||
|
||||
public void addInitValues(Double v){
|
||||
if ((initValue_vMin == null) && (initValue_vMax == null)){
|
||||
initValue_vMin = v;
|
||||
initValue_vMax = v;
|
||||
}
|
||||
else{
|
||||
if (v < initValue_vMin){
|
||||
initValue_vMin = v;
|
||||
}
|
||||
else if (v > initValue_vMax){
|
||||
initValue_vMax = v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
public void addInitRates(Double r){
|
||||
if ((initRate_rMin == null) && (initRate_rMax == null)){
|
||||
initRate_rMin = r;
|
||||
initRate_rMax = r;
|
||||
}
|
||||
else{
|
||||
if (r < initRate_rMin){
|
||||
initRate_rMin = r;
|
||||
}
|
||||
else if (r > initRate_rMax){
|
||||
initRate_rMax = r;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
public void reset(){
|
||||
initRate_rMax = null;
|
||||
initRate_rMin = null;
|
||||
initValue_vMin = null;
|
||||
initValue_vMax = null;
|
||||
this.runs = new DMVCrun();
|
||||
//this.epsilon = null ??
|
||||
//this.care = true ???
|
||||
}
|
||||
|
||||
public String getInitValue(){
|
||||
return ("["+(int)Math.floor(initValue_vMin)+","+(int)Math.ceil(initValue_vMax)+"]");
|
||||
}
|
||||
|
||||
public String getInitRate(){
|
||||
return ("["+(int)Math.floor(initRate_rMin)+","+(int)Math.ceil(initRate_rMax)+"]");
|
||||
}
|
||||
|
||||
public void scaleInitByDelay(Double dScaleFactor){
|
||||
if (!dmvc){
|
||||
initRate_rMin /= dScaleFactor;
|
||||
initRate_rMax /= dScaleFactor;
|
||||
}
|
||||
}
|
||||
|
||||
public void scaleInitByVar(Double vScaleFactor){
|
||||
if (!dmvc){
|
||||
initRate_rMin *= vScaleFactor;
|
||||
initRate_rMax *= vScaleFactor;
|
||||
}
|
||||
initValue_vMax *= vScaleFactor;
|
||||
initValue_vMin *= vScaleFactor;
|
||||
}
|
||||
|
||||
@Override
|
||||
public int compareTo(Variable o) {
|
||||
return (this.getName().compareToIgnoreCase(o.getName()));
|
||||
}
|
||||
|
||||
public void copy(Variable a){
|
||||
//Private fields in a. We shouldn't access like this?
|
||||
this.name = a.name;
|
||||
this.runs = a.runs;
|
||||
this.dmvc = a.dmvc;
|
||||
this.input = a.input;
|
||||
this.output = a.output;
|
||||
this.forceType = a.forceType;
|
||||
this.care = a.care;
|
||||
this.epsilon = a.epsilon;
|
||||
this.destab = a.destab;
|
||||
this.interpolate = a.interpolate;
|
||||
// NOT Copying the init values and rates bcoz they get normalized every time a learn is performed on them.
|
||||
// So the values get multiplied by the same factor multiple times which is wrong.
|
||||
// this.initValue_vMax = a.initValue_vMax;
|
||||
// this.initValue_vMin = a.initValue_vMin;
|
||||
// this.initRate_rMax = a.initRate_rMax;
|
||||
// this.initRate_rMin = a.initRate_rMin;
|
||||
}
|
||||
|
||||
public void forceDmvc(Boolean dmvc){
|
||||
if (dmvc == null){
|
||||
this.forceType = null;
|
||||
// this.dmvc = false;
|
||||
}
|
||||
else if (!dmvc){
|
||||
this.forceType = "Cont";
|
||||
this.dmvc = false;
|
||||
}
|
||||
else if (dmvc){
|
||||
this.forceType = "DMV";
|
||||
this.dmvc = true;
|
||||
}
|
||||
}
|
||||
|
||||
public boolean isForcedDmv(){
|
||||
if (forceType == null)
|
||||
return false;
|
||||
else if (forceType.equalsIgnoreCase("DMV"))
|
||||
return true;
|
||||
else
|
||||
return false;
|
||||
}
|
||||
|
||||
public boolean isForcedCont(){
|
||||
if (forceType == null)
|
||||
return false;
|
||||
else if (forceType.equalsIgnoreCase("Cont"))
|
||||
return true;
|
||||
else
|
||||
return false;
|
||||
}
|
||||
|
||||
public void setEpsilon(Double v) {
|
||||
this.epsilon = v;
|
||||
}
|
||||
|
||||
public Double getEpsilon() {
|
||||
return(epsilon);
|
||||
}
|
||||
|
||||
public boolean isDestab() {
|
||||
return destab;
|
||||
}
|
||||
|
||||
public void setDestab(boolean b) {
|
||||
destab = b;
|
||||
}
|
||||
|
||||
/*
|
||||
public void addInitValues(Double d, int i){
|
||||
if (initValues.isEmpty()){
|
||||
for (int j = 0; j < vars.length; j++){
|
||||
initValues.add(j, new ArrayList<Double>());
|
||||
}
|
||||
initValues.get(i).add(d);
|
||||
}
|
||||
else{
|
||||
initValues.get(i).add(d);
|
||||
}
|
||||
//this.initValues[j] = d;
|
||||
}
|
||||
|
||||
public void addInitRates(Double d, int i){
|
||||
if (initRates.isEmpty()){
|
||||
for (int j = 0; j < vars.length; j++){
|
||||
initRates.add(j, new ArrayList<Double>());
|
||||
}
|
||||
initRates.get(i).add(d);
|
||||
}
|
||||
else{
|
||||
initRates.get(i).add(d);
|
||||
}
|
||||
//this.initRates[j] = d;
|
||||
}
|
||||
*/
|
||||
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,174 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.HashMap;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Connection
|
||||
{
|
||||
|
||||
private double score;
|
||||
private String child;
|
||||
private List<String> parents;
|
||||
private Map<String, Type> parentToType;
|
||||
|
||||
public enum Type
|
||||
{
|
||||
ACTIVATOR("activator"), REPRESSOR("repressor"), UNKNOWN("unknown");
|
||||
|
||||
private final String name;
|
||||
|
||||
Type(String name)
|
||||
{
|
||||
this.name = name;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
public Connection(String child, double score, String type, String parent)
|
||||
{
|
||||
this.child = child;
|
||||
this.score = score;
|
||||
this.parents = new ArrayList<String>();
|
||||
this.parents.add(parent);
|
||||
this.parentToType = new HashMap<String, Type>();
|
||||
this.parentToType.put(parent, getType(type));
|
||||
}
|
||||
|
||||
public Connection(String child, double score, List<Connection> connections)
|
||||
{
|
||||
this.child = child;
|
||||
this.score = score;
|
||||
this.parents = new ArrayList<String>();
|
||||
this.parentToType = new HashMap<String, Type>();
|
||||
|
||||
for (Connection connection : connections)
|
||||
{
|
||||
this.parents.addAll(connection.getParents());
|
||||
|
||||
for (String parent : connection.getParents())
|
||||
{
|
||||
this.parentToType.put(parent, getType(connection.getParentType(parent)));
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
public Connection(String child, double score, Connection... connections)
|
||||
{
|
||||
this.child = child;
|
||||
this.score = score;
|
||||
this.parents = new ArrayList<String>();
|
||||
this.parentToType = new HashMap<String, Type>();
|
||||
|
||||
for (Connection connection : connections)
|
||||
{
|
||||
this.parents.addAll(connection.getParents());
|
||||
|
||||
for (String parent : connection.getParents())
|
||||
{
|
||||
this.parentToType.put(parent, getType(connection.getParentType(parent)));
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
public Type getType(String type)
|
||||
{
|
||||
type = type.toLowerCase();
|
||||
|
||||
if (type.equals("repressor"))
|
||||
{
|
||||
return Type.REPRESSOR;
|
||||
}
|
||||
else if (type.equals("activator"))
|
||||
{
|
||||
return Type.ACTIVATOR;
|
||||
}
|
||||
else
|
||||
{
|
||||
return Type.UNKNOWN;
|
||||
}
|
||||
}
|
||||
|
||||
public double getScore()
|
||||
{
|
||||
return score;
|
||||
}
|
||||
|
||||
public List<String> getParents()
|
||||
{
|
||||
return parents;
|
||||
}
|
||||
|
||||
public String getChild()
|
||||
{
|
||||
return child;
|
||||
}
|
||||
|
||||
public String getParentType(String parent)
|
||||
{
|
||||
if (parentToType.containsKey(parent))
|
||||
{
|
||||
return parentToType.get(parent).name();
|
||||
}
|
||||
else
|
||||
{
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
public boolean equalParents(List<String> parents)
|
||||
{
|
||||
for (String parent : parents)
|
||||
{
|
||||
if (!this.parents.contains(parent))
|
||||
{
|
||||
return false;
|
||||
}
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
@Override
|
||||
public String toString()
|
||||
{
|
||||
return "Connection [score=" + score + ", child=" + child + ", parents=" + getParents() + "]";
|
||||
}
|
||||
|
||||
public String getParentString()
|
||||
{
|
||||
String list = "";
|
||||
|
||||
for (String parent : parents)
|
||||
{
|
||||
list = list + parent + " ";
|
||||
}
|
||||
|
||||
return list;
|
||||
}
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,114 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.HashMap;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Encodings
|
||||
{
|
||||
|
||||
private int numBins;
|
||||
private Map<Integer, double[]> discreteSpecies;
|
||||
private List<List<List<Integer>>> levelAssignments;
|
||||
|
||||
public Encodings()
|
||||
{
|
||||
numBins = 3;
|
||||
discreteSpecies = new HashMap<Integer, double[]>();
|
||||
levelAssignments = new ArrayList<List<List<Integer>>>();
|
||||
}
|
||||
|
||||
public Encodings(int bin)
|
||||
{
|
||||
numBins = bin;
|
||||
discreteSpecies = new HashMap<Integer, double[]>();
|
||||
levelAssignments = new ArrayList<List<List<Integer>>>();
|
||||
}
|
||||
|
||||
public void addDiscreteSpecies(int col, double[] values)
|
||||
{
|
||||
discreteSpecies.put(col, values);
|
||||
}
|
||||
|
||||
public void addLevelAssignment(int experiment, int row, int col, double data)
|
||||
{
|
||||
while (levelAssignments.size() < experiment + 1)
|
||||
{
|
||||
levelAssignments.add(new ArrayList<List<Integer>>());
|
||||
}
|
||||
while (levelAssignments.get(experiment).size() < row + 1)
|
||||
{
|
||||
levelAssignments.get(experiment).add(new ArrayList<Integer>());
|
||||
}
|
||||
while (levelAssignments.get(experiment).get(row).size() < col + 1)
|
||||
{
|
||||
levelAssignments.get(experiment).get(row).add(0);
|
||||
}
|
||||
levelAssignments.get(experiment).get(row).set(col, getLevelAssignment(col, data));
|
||||
}
|
||||
|
||||
public int getLevelAssignment(int col, double data)
|
||||
{
|
||||
double[] discrete = discreteSpecies.get(col);
|
||||
|
||||
for (int i = 1; i < discrete.length; i++)
|
||||
{
|
||||
if (data <= discrete[i])
|
||||
{
|
||||
return i - 1;
|
||||
}
|
||||
}
|
||||
|
||||
return numBins - 1;
|
||||
}
|
||||
|
||||
public List<List<List<Integer>>> getLevelAssignments()
|
||||
{
|
||||
return levelAssignments;
|
||||
}
|
||||
|
||||
public int size()
|
||||
{
|
||||
return levelAssignments.size();
|
||||
}
|
||||
|
||||
public void print()
|
||||
{
|
||||
for (int i = 0; i < levelAssignments.size(); i++)
|
||||
{
|
||||
System.out.println("Experiment :" + i);
|
||||
for (int j = 0; j < levelAssignments.get(i).size(); j++)
|
||||
{
|
||||
for (int k = 1; k < levelAssignments.get(i).get(j).size(); k++)
|
||||
{
|
||||
System.out.print(levelAssignments.get(i).get(j).get(k) + " ");
|
||||
}
|
||||
System.out.println("");
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,73 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Experiments
|
||||
{
|
||||
|
||||
private List<List<List<Double>>> experiments;
|
||||
|
||||
public Experiments()
|
||||
{
|
||||
experiments = new ArrayList<List<List<Double>>>();
|
||||
}
|
||||
|
||||
public void addExperiment(int experiment, int row, int col, double data)
|
||||
{
|
||||
while (experiments.size() < experiment + 1)
|
||||
{
|
||||
experiments.add(new ArrayList<List<Double>>());
|
||||
}
|
||||
while (experiments.get(experiment).size() < row + 1)
|
||||
{
|
||||
experiments.get(experiment).add(new ArrayList<Double>());
|
||||
}
|
||||
while (experiments.get(experiment).get(row).size() < col + 1)
|
||||
{
|
||||
experiments.get(experiment).get(row).add(0.0);
|
||||
}
|
||||
experiments.get(experiment).get(row).set(col, data);
|
||||
}
|
||||
|
||||
public double getDataPoint(int experiment, int row, int col)
|
||||
{
|
||||
return experiments.get(experiment).get(row).get(col);
|
||||
}
|
||||
|
||||
public Experiments removeMutations(String s)
|
||||
{
|
||||
return this;
|
||||
}
|
||||
|
||||
public int getNumOfExperiments()
|
||||
{
|
||||
return experiments.size();
|
||||
}
|
||||
|
||||
public List<List<List<Double>>> getExperiments()
|
||||
{
|
||||
return experiments;
|
||||
}
|
||||
}
|
||||
710
learn/src/main/java/edu/utah/ece/async/learn/genenet/Learn.java
Normal file
710
learn/src/main/java/edu/utah/ece/async/learn/genenet/Learn.java
Normal file
|
|
@ -0,0 +1,710 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.io.File;
|
||||
import java.io.FileNotFoundException;
|
||||
import java.io.FileOutputStream;
|
||||
import java.io.IOException;
|
||||
import java.io.OutputStreamWriter;
|
||||
import java.util.ArrayList;
|
||||
import java.util.Arrays;
|
||||
import java.util.Collections;
|
||||
import java.util.HashMap;
|
||||
import java.util.HashSet;
|
||||
import java.util.LinkedList;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
import java.util.Set;
|
||||
|
||||
import edu.utah.ece.async.dataModels.util.GlobalConstants;
|
||||
import edu.utah.ece.async.dataModels.util.exceptions.BioSimException;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Learn
|
||||
{
|
||||
|
||||
private int bins;
|
||||
|
||||
public Learn(int bins)
|
||||
{
|
||||
this.bins = bins;
|
||||
}
|
||||
|
||||
public void learnBaselineNetwork(SpeciesCollection S, Experiments E, NetCon C)
|
||||
{
|
||||
for (int i = 0; i < S.size(); i++)
|
||||
{
|
||||
String s_0 = S.getInterestingSpecies(i);
|
||||
|
||||
for (int j = 0; j < S.size(); j++)
|
||||
{
|
||||
String s_1 = S.getInterestingSpecies(j);
|
||||
|
||||
if (!s_0.equals(s_1))
|
||||
{
|
||||
addParent(s_0, s_1, E, S, C);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private void addParent(String parent, String child, Experiments E, SpeciesCollection S, NetCon C)
|
||||
{
|
||||
int parent_col = S.getColumn(parent);
|
||||
int child_col = S.getColumn(child);
|
||||
int repression = 0;
|
||||
int activation = 0;
|
||||
int unknown = 0;
|
||||
double childPrevious, childCurr;
|
||||
double parentPrevious, parentCurr;
|
||||
double childRate, previousChildRate;
|
||||
double parentRate, previousParentRate;
|
||||
|
||||
for (int i = 0; i < E.getNumOfExperiments(); i++)
|
||||
{
|
||||
previousChildRate = Double.NaN;
|
||||
previousParentRate = Double.NaN;
|
||||
for (int j = 1; j < E.getExperiments().get(i).size(); j++)
|
||||
{
|
||||
childPrevious = E.getDataPoint(i, j - 1, child_col);
|
||||
childCurr = E.getDataPoint(i, j, child_col);
|
||||
parentPrevious = E.getDataPoint(i, j - 1, parent_col);
|
||||
parentCurr = E.getDataPoint(i, j, parent_col);
|
||||
childRate = childCurr - childPrevious;
|
||||
parentRate = parentCurr - parentPrevious;
|
||||
|
||||
if (j == 1)
|
||||
{
|
||||
previousChildRate = childRate;
|
||||
previousParentRate = parentRate;
|
||||
continue;
|
||||
}
|
||||
|
||||
if (childCurr > 10 && parentCurr > 10)
|
||||
{
|
||||
if (childRate < previousChildRate)
|
||||
{
|
||||
if (parentRate > previousParentRate)
|
||||
{
|
||||
repression++;
|
||||
}
|
||||
else if (parentRate < previousParentRate)
|
||||
{
|
||||
activation++;
|
||||
}
|
||||
else
|
||||
{
|
||||
unknown++;
|
||||
}
|
||||
}
|
||||
else if (childRate > previousChildRate)
|
||||
{
|
||||
if (parentRate > previousParentRate)
|
||||
{
|
||||
activation++;
|
||||
}
|
||||
else if (parentRate < previousParentRate)
|
||||
{
|
||||
repression++;
|
||||
}
|
||||
else
|
||||
{
|
||||
unknown++;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
previousParentRate = parentRate;
|
||||
previousChildRate = childRate;
|
||||
}
|
||||
}
|
||||
|
||||
if (repression > activation && repression >= unknown)
|
||||
{
|
||||
C.addConnection(child, "repressor", repression, parent);
|
||||
}
|
||||
else if (activation > repression && repression >= unknown)
|
||||
{
|
||||
C.addConnection(child, "activator", repression, parent);
|
||||
}
|
||||
}
|
||||
|
||||
public void learnNetwork(SpeciesCollection S, Experiments E, NetCon C, Thresholds T, Encodings L) throws BioSimException
|
||||
{
|
||||
EncodeExpts(S, E, C, T, L);
|
||||
|
||||
for (int i = 0; i < S.size(); i++)
|
||||
{
|
||||
String s = S.getInterestingSpecies(i);
|
||||
|
||||
if (!C.containEdge(s))
|
||||
{
|
||||
Experiments new_E = E.removeMutations(s);
|
||||
selectInitialParents(s, S, new_E, C, T, L);
|
||||
createMultipleParents(s, S, new_E, C, T, L);
|
||||
competeMultipleParents(s, S, new_E, C, T, L);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private void EncodeExpts(SpeciesCollection S, Experiments E, NetCon C, Thresholds T, Encodings L) throws BioSimException
|
||||
{
|
||||
for (int i = 0; i < S.size(); i++)
|
||||
{
|
||||
String species = S.getInterestingSpecies(i);
|
||||
if(!S.containsSpeciesData(species))
|
||||
{
|
||||
throw new BioSimException("There is no data for " + species + ". Check if species ids in the target model match the ones from the data.", "Error in learning.");
|
||||
}
|
||||
getDiscreteLevels(species, S, E, L, bins);
|
||||
getBinAssign(species, S, E, L, bins);
|
||||
}
|
||||
}
|
||||
|
||||
private void getBinAssign(String species, SpeciesCollection S, Experiments experiments, Encodings L, int bins)
|
||||
{
|
||||
int col = S.getColumn(species);
|
||||
for (int i = 0; i < experiments.getNumOfExperiments(); i++)
|
||||
{
|
||||
List<List<Double>> experiment = experiments.getExperiments().get(i);
|
||||
|
||||
for (int j = 0; j < experiment.size(); j++)
|
||||
{
|
||||
double value = experiment.get(j).get(col);
|
||||
|
||||
L.addLevelAssignment(i, j, col, value);
|
||||
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
private void getDiscreteLevels(String species, SpeciesCollection S, Experiments experiments, Encodings L, int levels)
|
||||
{
|
||||
int col = S.getColumn(species);
|
||||
double[] level = new double[levels + 1];
|
||||
double v = Double.NaN, vp = Double.NaN;
|
||||
int steps = 0;
|
||||
level[0] = 0;
|
||||
level[levels] = Double.POSITIVE_INFINITY;
|
||||
|
||||
LinkedList<Double> values = new LinkedList<Double>();
|
||||
|
||||
for (List<List<Double>> experiment : experiments.getExperiments())
|
||||
{
|
||||
for (int i = 0; i < experiment.size(); i++)
|
||||
{
|
||||
double value = experiment.get(i).get(col);
|
||||
|
||||
values.add(value);
|
||||
}
|
||||
}
|
||||
|
||||
Collections.sort(values);
|
||||
|
||||
for (int j = 1; j < levels - 1; j++)
|
||||
{
|
||||
int count = 0;
|
||||
level[j] = Double.POSITIVE_INFINITY;
|
||||
|
||||
steps = values.size() / (levels - j + 1);
|
||||
|
||||
while (values.size() > 0 && count < steps)
|
||||
{
|
||||
v = values.pop();
|
||||
|
||||
count++;
|
||||
}
|
||||
|
||||
do
|
||||
{
|
||||
vp = values.pop();
|
||||
}
|
||||
while (values.size() > 0 && v == vp);
|
||||
|
||||
if (values.size() > 0)
|
||||
{
|
||||
level[j] = v;
|
||||
values.addFirst(vp);
|
||||
}
|
||||
}
|
||||
|
||||
L.addDiscreteSpecies(col, level);
|
||||
}
|
||||
|
||||
private static double scoreParents(String s, SpeciesCollection S, Set<String> P, Set<String> G, Experiments E, Thresholds T, Encodings L)
|
||||
{
|
||||
int votes_a = 0, votes_r = 0, votes_u = 0;
|
||||
int totalDen = 0, den = 0, totalNum = 0, num = 0;
|
||||
int targetCol = S.getColumn(s);
|
||||
double probRatio = 0;
|
||||
List<Integer> levelAssignmentsP = getLevelAssignments(P, S, L);
|
||||
Collections.sort(levelAssignmentsP);
|
||||
List<Integer> levelAssignmentsG = getLevelAssignments(G, S, L);
|
||||
Collections.sort(levelAssignmentsP);
|
||||
int l0 = levelAssignmentsP.remove(0);
|
||||
|
||||
for (int l : levelAssignmentsG)
|
||||
{
|
||||
for (int lp : levelAssignmentsP)
|
||||
{
|
||||
for (int i = 0; i < E.getNumOfExperiments(); i++)
|
||||
{
|
||||
for (int j = 0; j < E.getExperiments().get(i).size(); j++)
|
||||
{
|
||||
|
||||
boolean g = checkSameLevel(G, i, j, l, S, L);
|
||||
boolean p = checkSameLevel(P, i, j, lp, S, L);
|
||||
boolean p0 = checkSameLevel(P, i, j, l0, S, L);
|
||||
|
||||
if (g && p)
|
||||
{
|
||||
if (isIncreasing(i, j, targetCol, E))
|
||||
{
|
||||
num++;
|
||||
}
|
||||
totalNum++;
|
||||
}
|
||||
else if (g && p0)
|
||||
{
|
||||
if (isIncreasing(i, j, targetCol, E))
|
||||
{
|
||||
den++;
|
||||
}
|
||||
totalDen++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (num == 0 || den == 0)
|
||||
{
|
||||
probRatio = 1;
|
||||
}
|
||||
else
|
||||
{
|
||||
probRatio = (1.0 * num / totalNum) / (1.0 * den / totalDen);
|
||||
}
|
||||
|
||||
if (probRatio > T.getTa())
|
||||
{
|
||||
votes_a++;
|
||||
}
|
||||
else if (probRatio < T.getTr())
|
||||
{
|
||||
votes_r++;
|
||||
}
|
||||
else
|
||||
{
|
||||
votes_u++;
|
||||
}
|
||||
}
|
||||
|
||||
return 1.0 * (votes_a - votes_r) / (votes_a + votes_r + votes_u);
|
||||
}
|
||||
|
||||
private static boolean checkSameLevel(Set<String> species, int experiment, int row, int bin, SpeciesCollection S, Encodings L)
|
||||
{
|
||||
for (String s : species)
|
||||
{
|
||||
int col = S.getColumn(s);
|
||||
|
||||
if (L.getLevelAssignments().get(experiment).get(row).get(col) != bin)
|
||||
{
|
||||
return false;
|
||||
}
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
private static boolean isIncreasing(int experiment, int row, int targetCol, Experiments E)
|
||||
{
|
||||
List<List<Double>> list = E.getExperiments().get(experiment);
|
||||
|
||||
if (row + 1 < list.size())
|
||||
{
|
||||
if (list.get(row).get(targetCol) < list.get(row + 1).get(targetCol))
|
||||
{
|
||||
return true;
|
||||
}
|
||||
}
|
||||
|
||||
if (row - 1 >= 0)
|
||||
{
|
||||
if (list.get(row - 1).get(targetCol) < list.get(row).get(targetCol))
|
||||
{
|
||||
return true;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
private static List<Integer> getLevelAssignments(Set<String> P, SpeciesCollection S, Encodings L)
|
||||
{
|
||||
List<Integer> set = new ArrayList<Integer>();
|
||||
|
||||
for (String species : P)
|
||||
{
|
||||
int col = S.getColumn(species);
|
||||
|
||||
for (int i = 0; i < L.size(); i++)
|
||||
{
|
||||
List<List<Integer>> encodings = L.getLevelAssignments().get(i);
|
||||
|
||||
for (int j = 0; j < encodings.size(); j++)
|
||||
{
|
||||
int value = encodings.get(j).get(col);
|
||||
if (!set.contains(value))
|
||||
{
|
||||
set.add(encodings.get(j).get(col));
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return set;
|
||||
}
|
||||
|
||||
private static void selectInitialParents(String s, SpeciesCollection S, Experiments E, NetCon C, Thresholds T, Encodings L)
|
||||
{
|
||||
List<String> interestingSpecies = new ArrayList<String>(S.getInterestingSpecies());
|
||||
interestingSpecies.remove(s);
|
||||
|
||||
double score;
|
||||
|
||||
for (String p : interestingSpecies)
|
||||
{
|
||||
score = scoreParents(s, S, new HashSet<String>(Arrays.asList(p)), new HashSet<String>(Arrays.asList(s)), E, T, L);
|
||||
|
||||
if (score >= T.getTv())
|
||||
{
|
||||
C.addConnection(s, "activator", score, p);
|
||||
}
|
||||
else if (score <= -T.getTv())
|
||||
{
|
||||
C.addConnection(s, "repressor", score, p);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private static void createMultipleParents(String s, SpeciesCollection S, Experiments E, NetCon C, Thresholds T, Encodings L)
|
||||
{
|
||||
|
||||
double score1, score2, scoreb;
|
||||
|
||||
List<Connection> connectionsToDelete = new ArrayList<Connection>();
|
||||
List<Connection> multipleParents = new ArrayList<Connection>();
|
||||
List<Connection> singleParent = C.getListOfConnections(s);
|
||||
|
||||
if (singleParent == null)
|
||||
{
|
||||
return;
|
||||
}
|
||||
|
||||
for (int i = 0; i < singleParent.size(); i++)
|
||||
{
|
||||
Connection connection1 = singleParent.get(i);
|
||||
|
||||
score1 = connection1.getScore();
|
||||
|
||||
for (int j = i + 1; j < singleParent.size(); j++)
|
||||
{
|
||||
Connection connection2 = singleParent.get(j);
|
||||
score2 = connection2.getScore();
|
||||
|
||||
Set<String> parents = new HashSet<String>();
|
||||
parents.addAll(connection1.getParents());
|
||||
parents.addAll(connection2.getParents());
|
||||
|
||||
scoreb = scoreParents(s, S, parents, new HashSet<String>(Arrays.asList(s)), E, T, L);
|
||||
if (Math.abs(scoreb) >= Math.abs(score1) && Math.abs(scoreb) >= Math.abs(score2))
|
||||
{
|
||||
multipleParents.add(new Connection(s, scoreb, connection1, connection2));
|
||||
connectionsToDelete.add(connection1);
|
||||
connectionsToDelete.add(connection2);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
for (Connection connection : connectionsToDelete)
|
||||
{
|
||||
C.removeConnection(s, connection);
|
||||
}
|
||||
|
||||
for (Connection connection : multipleParents)
|
||||
{
|
||||
C.addConnection(s, connection);
|
||||
}
|
||||
}
|
||||
|
||||
private static void competeMultipleParents(String s, SpeciesCollection S, Experiments E, NetCon C, Thresholds T, Encodings L)
|
||||
{
|
||||
List<Connection> potentialParents = C.getListOfConnections(s);
|
||||
double scoreq, smallestScore;
|
||||
|
||||
if (potentialParents == null)
|
||||
{
|
||||
return;
|
||||
}
|
||||
|
||||
double Ta = T.getTa();
|
||||
double Tr = T.getTr();
|
||||
|
||||
while (potentialParents.size() > 1)
|
||||
{
|
||||
List<List<String>> Q = getContenders(s, C);
|
||||
List<Double> scores = new ArrayList<Double>();
|
||||
smallestScore = Double.POSITIVE_INFINITY;
|
||||
|
||||
for (List<String> q : Q)
|
||||
{
|
||||
HashSet<String> P = new HashSet<String>(q);
|
||||
HashSet<String> G = new HashSet<String>();
|
||||
|
||||
for (List<String> qp : Q)
|
||||
{
|
||||
if (qp != q)
|
||||
{
|
||||
G.addAll(qp);
|
||||
}
|
||||
}
|
||||
|
||||
G.add(s);
|
||||
scoreq = scoreParents(s, S, P, G, E, T, L);
|
||||
scores.add(Math.abs(scoreq));
|
||||
|
||||
if (Math.abs(scoreq) < smallestScore)
|
||||
{
|
||||
smallestScore = Math.abs(scoreq);
|
||||
}
|
||||
}
|
||||
|
||||
boolean isScoreTied = isTied(smallestScore, Q, scores);
|
||||
|
||||
if (isScoreTied)
|
||||
{
|
||||
if (T.getTa() > 4 || T.getTr() < 0)
|
||||
{
|
||||
removeLosers(s, C, Q);
|
||||
}
|
||||
else
|
||||
{
|
||||
T.setTa(T.getTa() + T.getTt());
|
||||
T.setTr(T.getTr() - T.getTt());
|
||||
}
|
||||
}
|
||||
else
|
||||
{
|
||||
removeLosers(s, C, Q, scores, smallestScore);
|
||||
}
|
||||
}
|
||||
|
||||
T.setTa(Ta);
|
||||
T.setTr(Tr);
|
||||
}
|
||||
|
||||
private static boolean checkNumParents(List<List<String>> Q)
|
||||
{
|
||||
boolean sameNum = true;
|
||||
|
||||
if (Q.size() > 0)
|
||||
{
|
||||
int size = Q.get(0).size();
|
||||
|
||||
for (int i = 1; i < Q.size(); i++)
|
||||
{
|
||||
if (Q.get(i).size() != size)
|
||||
{
|
||||
sameNum = false;
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return sameNum;
|
||||
}
|
||||
|
||||
private static boolean isTied(double smallestScore, List<List<String>> Q, List<Double> scores)
|
||||
{
|
||||
boolean isScoreTied = true;
|
||||
|
||||
for (int i = 0; i < Q.size(); i++)
|
||||
{
|
||||
if (scores.get(i) != smallestScore)
|
||||
{
|
||||
isScoreTied = false;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
return isScoreTied;
|
||||
}
|
||||
|
||||
private static void removeLosers(String s, NetCon C, List<List<String>> Q)
|
||||
{
|
||||
if (checkNumParents(Q))
|
||||
{
|
||||
int minParents = Integer.MAX_VALUE;
|
||||
int index = -1;
|
||||
for (int i = 0; i < Q.size(); i++)
|
||||
{
|
||||
if (Q.get(i).size() < minParents)
|
||||
{
|
||||
minParents = Q.get(i).size();
|
||||
index = i;
|
||||
}
|
||||
}
|
||||
C.removeConnectionByParent(s, Q.get(index));
|
||||
}
|
||||
else
|
||||
{
|
||||
C.removeConnection(s, 0);
|
||||
}
|
||||
}
|
||||
|
||||
private static void removeLosers(String s, NetCon C, List<List<String>> Q, List<Double> scores, double smallestScore)
|
||||
{
|
||||
|
||||
for (int i = 0; i < Q.size(); i++)
|
||||
{
|
||||
if (scores.get(i) == smallestScore)
|
||||
{
|
||||
C.removeConnectionByParent(s, Q.get(i));
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private static List<List<String>> getContenders(String s, NetCon C)
|
||||
{
|
||||
List<List<String>> contenders = new ArrayList<List<String>>();
|
||||
List<Connection> potentialParents = C.getListOfConnections(s);
|
||||
|
||||
double maxScore = Double.MIN_VALUE;
|
||||
double minScore = Double.MAX_VALUE;
|
||||
|
||||
if (potentialParents != null)
|
||||
{
|
||||
for (Connection connection : potentialParents)
|
||||
{
|
||||
if (connection.getScore() > maxScore)
|
||||
{
|
||||
maxScore = connection.getScore();
|
||||
}
|
||||
else if (connection.getScore() < minScore)
|
||||
{
|
||||
minScore = connection.getScore();
|
||||
}
|
||||
}
|
||||
|
||||
for (Connection connection : potentialParents)
|
||||
{
|
||||
if (connection.getScore() == maxScore || connection.getScore() == minScore)
|
||||
{
|
||||
contenders.add(connection.getParents());
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return contenders;
|
||||
|
||||
}
|
||||
|
||||
public void getDotFile(String filename, String directory, SpeciesCollection collection, NetCon network) throws BioSimException
|
||||
{
|
||||
Map<String, String> speciesToNode;
|
||||
File fout = new File(directory + GlobalConstants.separator + filename);
|
||||
FileOutputStream fos = null;
|
||||
OutputStreamWriter osw = null;
|
||||
try
|
||||
{
|
||||
int index = 0;
|
||||
speciesToNode = new HashMap<String, String>();
|
||||
fos = new FileOutputStream(fout);
|
||||
osw = new OutputStreamWriter(fos);
|
||||
osw.write("digraph G { \n");
|
||||
|
||||
for (String s : collection.getInterestingSpecies())
|
||||
{
|
||||
String node = "s" + ++index;
|
||||
osw.write(node + " [shape=ellipse,color=black,label=\"" + s + "\"];\n");
|
||||
speciesToNode.put(s, node);
|
||||
}
|
||||
for (String s : network.getConnections().keySet())
|
||||
{
|
||||
for (Connection connection : network.getConnections().get(s))
|
||||
{
|
||||
for (String parent : connection.getParents())
|
||||
{
|
||||
String type = connection.getParentType(parent).toLowerCase();
|
||||
|
||||
if (type.equals("activator"))
|
||||
{
|
||||
osw.write(speciesToNode.get(parent) + " -> " + speciesToNode.get(s) + " [color=\"blue4\",arrowhead=vee]\n");
|
||||
}
|
||||
else if (type.equals("repressor"))
|
||||
{
|
||||
osw.write(speciesToNode.get(parent) + " -> " + speciesToNode.get(s) + " [color=\"firebrick4\",arrowhead=tee]\n");
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
osw.write("} \n");
|
||||
}
|
||||
catch (FileNotFoundException e)
|
||||
{
|
||||
throw new BioSimException("Could not write dot file. File could not be found.", "Error in Learning");
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
throw new BioSimException("Error when writing dot file.", "Error in Learning");
|
||||
}
|
||||
finally
|
||||
{
|
||||
try
|
||||
{
|
||||
if (osw != null)
|
||||
{
|
||||
osw.close();
|
||||
}
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
throw new BioSimException("Failed to close writer", "Error in Learning");
|
||||
}
|
||||
|
||||
try
|
||||
{
|
||||
if (fos != null)
|
||||
{
|
||||
fos.close();
|
||||
}
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
throw new BioSimException("Failed to close outputstream", "Error in Learning");
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
111
learn/src/main/java/edu/utah/ece/async/learn/genenet/NetCon.java
Normal file
111
learn/src/main/java/edu/utah/ece/async/learn/genenet/NetCon.java
Normal file
|
|
@ -0,0 +1,111 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.HashMap;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class NetCon
|
||||
{
|
||||
|
||||
private Map<String, List<Connection>> connections;
|
||||
|
||||
public NetCon()
|
||||
{
|
||||
connections = new HashMap<String, List<Connection>>();
|
||||
}
|
||||
|
||||
public boolean containEdge(String s)
|
||||
{
|
||||
return false;
|
||||
}
|
||||
|
||||
public void addConnection(String child, String type, double score, String parent)
|
||||
{
|
||||
if (!connections.containsKey(child))
|
||||
{
|
||||
connections.put(child, new ArrayList<Connection>());
|
||||
}
|
||||
connections.get(child).add(new Connection(child, score, type, parent));
|
||||
}
|
||||
|
||||
public void addConnection(String child, Connection connection)
|
||||
{
|
||||
if (!connections.containsKey(child))
|
||||
{
|
||||
connections.put(child, new ArrayList<Connection>());
|
||||
}
|
||||
connections.get(child).add(connection);
|
||||
}
|
||||
|
||||
public void addConnection(String child, double score, Connection... collections)
|
||||
{
|
||||
if (!connections.containsKey(child))
|
||||
{
|
||||
connections.put(child, new ArrayList<Connection>());
|
||||
}
|
||||
connections.get(child).add(new Connection(child, score, collections));
|
||||
}
|
||||
|
||||
public void removeConnection(String child, Connection connection)
|
||||
{
|
||||
if (connections.containsKey(child))
|
||||
{
|
||||
connections.get(child).remove(connection);
|
||||
}
|
||||
}
|
||||
|
||||
public void removeConnectionByParent(String child, List<String> parents)
|
||||
{
|
||||
List<Connection> listOfConnections = connections.get(child);
|
||||
|
||||
for (int i = listOfConnections.size() - 1; i >= 0; i--)
|
||||
{
|
||||
if (listOfConnections.get(i).equalParents(parents))
|
||||
{
|
||||
listOfConnections.remove(i);
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
public void removeConnection(String child, int index)
|
||||
{
|
||||
if (connections.containsKey(child))
|
||||
{
|
||||
connections.get(child).remove(index);
|
||||
}
|
||||
}
|
||||
|
||||
public List<Connection> getListOfConnections(String s)
|
||||
{
|
||||
return connections.get(s);
|
||||
}
|
||||
|
||||
public Map<String, List<Connection>> getConnections()
|
||||
{
|
||||
return connections;
|
||||
}
|
||||
|
||||
}
|
||||
281
learn/src/main/java/edu/utah/ece/async/learn/genenet/Run.java
Normal file
281
learn/src/main/java/edu/utah/ece/async/learn/genenet/Run.java
Normal file
|
|
@ -0,0 +1,281 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.io.File;
|
||||
import java.io.FileInputStream;
|
||||
import java.io.FileNotFoundException;
|
||||
import java.io.IOException;
|
||||
import java.io.InputStream;
|
||||
import java.util.Scanner;
|
||||
|
||||
import javax.xml.stream.XMLStreamException;
|
||||
|
||||
import org.sbml.jsbml.Model;
|
||||
import org.sbml.jsbml.SBMLDocument;
|
||||
import org.sbml.jsbml.SBMLReader;
|
||||
import org.sbml.jsbml.Species;
|
||||
|
||||
import edu.utah.ece.async.dataModels.util.exceptions.BioSimException;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Run
|
||||
{
|
||||
|
||||
private static int experiment;
|
||||
|
||||
public static boolean run(String filename, String directory) throws BioSimException
|
||||
{
|
||||
experiment = 0;
|
||||
SpeciesCollection S = new SpeciesCollection();
|
||||
Experiments E = new Experiments();
|
||||
Encodings L = new Encodings();
|
||||
Thresholds T = new Thresholds();
|
||||
NetCon C = new NetCon();
|
||||
init(filename, S);
|
||||
loadExperiments(directory, S, E);
|
||||
if(experiment < 1)
|
||||
{
|
||||
return false;
|
||||
}
|
||||
Learn learn = new Learn(3);
|
||||
learn.learnNetwork(S, E, C, T, L);
|
||||
learn.getDotFile("method.gcm", directory, S, C);
|
||||
learn.getDotFile("method.dot", directory, S, C);
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
public static void loadExperiments(String directory, SpeciesCollection S, Experiments E) throws BioSimException
|
||||
{
|
||||
File path = new File(directory);
|
||||
|
||||
for (File file : path.listFiles())
|
||||
{
|
||||
String name = file.getAbsolutePath();
|
||||
if (name.endsWith(".tsd"))
|
||||
{
|
||||
parse(name, S, E);
|
||||
experiment++;
|
||||
}
|
||||
else if (name.endsWith(".csv"))
|
||||
{
|
||||
parseCSV(name, S, E);
|
||||
experiment++;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
public static void init(String filename, SpeciesCollection S)
|
||||
{
|
||||
try
|
||||
{
|
||||
SBMLDocument doc = SBMLReader.read(new File(filename));
|
||||
|
||||
Model model = doc.getModel();
|
||||
|
||||
for (Species species : model.getListOfSpecies())
|
||||
{
|
||||
S.addInterestingSpecies(species.getId());
|
||||
}
|
||||
}
|
||||
|
||||
catch (XMLStreamException e)
|
||||
{
|
||||
e.printStackTrace();
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
e.printStackTrace();
|
||||
}
|
||||
}
|
||||
|
||||
private static void parseCSV(String filename, SpeciesCollection S, Experiments E)
|
||||
{
|
||||
Scanner scan = null;
|
||||
boolean isFirst = true;
|
||||
try
|
||||
{
|
||||
scan = new Scanner(new File(filename));
|
||||
int row = 0;
|
||||
while (scan.hasNextLine())
|
||||
{
|
||||
String line = scan.nextLine();
|
||||
|
||||
String[] values = line.split(",");
|
||||
|
||||
if (isFirst)
|
||||
{
|
||||
for (int i = 0; i < values.length; i++)
|
||||
{
|
||||
S.addSpecies(values[i], i);
|
||||
}
|
||||
isFirst = false;
|
||||
}
|
||||
else
|
||||
{
|
||||
for (int i = 0; i < values.length; i++)
|
||||
{
|
||||
E.addExperiment(experiment, row, i, Double.parseDouble(values[i]));
|
||||
}
|
||||
row++;
|
||||
}
|
||||
}
|
||||
}
|
||||
catch (FileNotFoundException e)
|
||||
{
|
||||
System.out.println("Could not find the file!");
|
||||
}
|
||||
finally
|
||||
{
|
||||
if (scan != null)
|
||||
{
|
||||
scan.close();
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
|
||||
private static void parse(String filename, SpeciesCollection S, Experiments E) throws BioSimException
|
||||
{
|
||||
InputStream input = null;
|
||||
|
||||
try
|
||||
{
|
||||
|
||||
String buffer = "";
|
||||
char state = 1;
|
||||
char read = 0;
|
||||
int row = 0;
|
||||
input = new FileInputStream(filename);
|
||||
|
||||
int data = input.read();
|
||||
while (data != -1)
|
||||
{
|
||||
data = input.read();
|
||||
read = (char) data;
|
||||
|
||||
if (read == ' ' || read == '\n')
|
||||
{
|
||||
continue;
|
||||
}
|
||||
|
||||
switch (state)
|
||||
{
|
||||
case 1:
|
||||
if (read == '(')
|
||||
{
|
||||
state = 2;
|
||||
buffer = "";
|
||||
break;
|
||||
}
|
||||
else
|
||||
{
|
||||
return;
|
||||
}
|
||||
case 2:
|
||||
if (read == ')')
|
||||
{
|
||||
state = 3;
|
||||
|
||||
row = 0;
|
||||
String[] ids = buffer.split(",");
|
||||
|
||||
if (experiment == 0)
|
||||
{
|
||||
for (int i = 0; i < ids.length; i++)
|
||||
{
|
||||
S.addSpecies(ids[i].substring(1, ids[i].length() - 1), i);
|
||||
}
|
||||
}
|
||||
|
||||
break;
|
||||
}
|
||||
else
|
||||
{
|
||||
buffer += read;
|
||||
|
||||
break;
|
||||
}
|
||||
case 3:
|
||||
if (read == '(')
|
||||
{
|
||||
state = 4;
|
||||
buffer = "";
|
||||
break;
|
||||
}
|
||||
if (read == ',')
|
||||
{
|
||||
break;
|
||||
}
|
||||
if (read == ')')
|
||||
{
|
||||
return;
|
||||
}
|
||||
else
|
||||
{
|
||||
return;
|
||||
}
|
||||
default:
|
||||
if (read == ')')
|
||||
{
|
||||
state = 3;
|
||||
String[] values = buffer.replace(" ", "").split(",");
|
||||
|
||||
for (int i = 0; i < values.length; i++)
|
||||
{
|
||||
E.addExperiment(experiment, row, i, Double.parseDouble(values[i]));
|
||||
}
|
||||
|
||||
row++;
|
||||
}
|
||||
else
|
||||
{
|
||||
buffer += read;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
catch (FileNotFoundException e)
|
||||
{
|
||||
throw new BioSimException("Could not find the file!", "Error in Learning");
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
throw new BioSimException("There was a problem when reading the file!", "Error in Learning");
|
||||
}
|
||||
finally
|
||||
{
|
||||
try
|
||||
{
|
||||
if (input != null)
|
||||
{
|
||||
input.close();
|
||||
}
|
||||
}
|
||||
catch (IOException e)
|
||||
{
|
||||
throw new BioSimException("Failed to close input stream", "Error in Learning");
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,78 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.HashMap;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class SpeciesCollection
|
||||
{
|
||||
|
||||
private List<String> interestingSpecies;
|
||||
private Map<Integer, String> columnSpecies;
|
||||
private Map<String, Integer> speciesColumn;
|
||||
|
||||
public SpeciesCollection()
|
||||
{
|
||||
interestingSpecies = new ArrayList<String>();
|
||||
columnSpecies = new HashMap<Integer, String>();
|
||||
speciesColumn = new HashMap<String, Integer>();
|
||||
}
|
||||
|
||||
public void addSpecies(String id, int index)
|
||||
{
|
||||
speciesColumn.put(id, index);
|
||||
columnSpecies.put(index, id);
|
||||
}
|
||||
|
||||
public void addInterestingSpecies(String id)
|
||||
{
|
||||
interestingSpecies.add(id);
|
||||
}
|
||||
|
||||
public int getColumn(String species)
|
||||
{
|
||||
return speciesColumn.get(species);
|
||||
}
|
||||
|
||||
public String getInterestingSpecies(int index)
|
||||
{
|
||||
return interestingSpecies.get(index);
|
||||
}
|
||||
|
||||
public int size()
|
||||
{
|
||||
return interestingSpecies.size();
|
||||
}
|
||||
|
||||
public List<String> getInterestingSpecies()
|
||||
{
|
||||
return interestingSpecies;
|
||||
}
|
||||
|
||||
public boolean containsSpeciesData(String species)
|
||||
{
|
||||
return speciesColumn.containsKey(species);
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,79 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.genenet;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Thresholds
|
||||
{
|
||||
|
||||
private double Ta, Tr, Tv, Tt;
|
||||
|
||||
public Thresholds()
|
||||
{
|
||||
Ta = 1.15;
|
||||
Tr = 0.75;
|
||||
Tv = 0.5;
|
||||
Tt = 0.025;
|
||||
}
|
||||
|
||||
public Thresholds(double Ta, double Tr, double Tv)
|
||||
{
|
||||
this.Ta = Ta;
|
||||
this.Tr = Tr;
|
||||
this.Tv = Tv;
|
||||
this.Tt = 0.025;
|
||||
}
|
||||
|
||||
public double getTa()
|
||||
{
|
||||
return Ta;
|
||||
}
|
||||
|
||||
public double getTr()
|
||||
{
|
||||
return Tr;
|
||||
}
|
||||
|
||||
public double getTv()
|
||||
{
|
||||
return Tv;
|
||||
}
|
||||
|
||||
public void setTa(double Ta)
|
||||
{
|
||||
this.Ta = Ta;
|
||||
}
|
||||
|
||||
public void setTr(double Tr)
|
||||
{
|
||||
this.Tr = Tr;
|
||||
}
|
||||
|
||||
public void setTv(double Tv)
|
||||
{
|
||||
this.Tv = Tv;
|
||||
}
|
||||
|
||||
public double getTt()
|
||||
{
|
||||
return Tt;
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,91 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.util.ArrayList;
|
||||
import java.util.List;
|
||||
|
||||
import javax.xml.stream.XMLStreamException;
|
||||
|
||||
import org.sbml.jsbml.SBMLDocument;
|
||||
|
||||
import edu.utah.ece.async.analysis.simulation.hierarchical.HierarchicalSimulation;
|
||||
import edu.utah.ece.async.analysis.simulation.hierarchical.methods.HierarchicalODERKSimulator;
|
||||
import edu.utah.ece.async.dataModels.util.exceptions.BioSimException;
|
||||
import edu.utah.ece.async.learn.genenet.Experiments;
|
||||
import edu.utah.ece.async.learn.genenet.SpeciesCollection;
|
||||
import edu.utah.ece.async.learn.parameterestimator.methods.sres.EvolutionMethodSetting;
|
||||
import edu.utah.ece.async.learn.parameterestimator.methods.sres.Modelsettings;
|
||||
import edu.utah.ece.async.learn.parameterestimator.methods.sres.ObjectiveSqureError;
|
||||
import edu.utah.ece.async.learn.parameterestimator.methods.sres.SRES;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class ParameterEstimator
|
||||
{
|
||||
static double relativeError = 1e-6;
|
||||
static double absoluteError = 1e-9;
|
||||
static int numSteps;
|
||||
static double maxTimeStep = Double.POSITIVE_INFINITY;
|
||||
static double minTimeStep = 0.0;
|
||||
static long randomSeed = 0;
|
||||
static int runs = 1;
|
||||
static double stoichAmpValue = 1.0;
|
||||
static boolean genStats = false;
|
||||
static String selectedSimulator = "";
|
||||
static ArrayList<String> interestingSpecies = new ArrayList<String>();
|
||||
static String quantityType = "amount";
|
||||
|
||||
public static SBMLDocument estimate(String SBMLFileName, String root, List<String> parameterList, Experiments experiments, SpeciesCollection speciesCollection) throws IOException, XMLStreamException, BioSimException
|
||||
{
|
||||
|
||||
int numberofparameters = parameterList.size();
|
||||
int sp = 0;
|
||||
int n = experiments.getExperiments().get(0).size() - 1;
|
||||
double ep = experiments.getExperiments().get(0).get(n).get(0);
|
||||
double[] lowerbounds = new double[numberofparameters];
|
||||
double[] upperbounds = new double[numberofparameters];
|
||||
HierarchicalSimulation sim = new HierarchicalODERKSimulator(SBMLFileName, root, 0);
|
||||
sim.initialize(randomSeed, 0);
|
||||
|
||||
for (int i = 0; i < numberofparameters; i++)
|
||||
{
|
||||
lowerbounds[i] = sim.getTopLevelValue(parameterList.get(i)) / 100;
|
||||
upperbounds[i] = sim.getTopLevelValue(parameterList.get(i)) * 100;
|
||||
}
|
||||
Modelsettings M1 = new Modelsettings(experiments.getExperiments().get(0).get(0), speciesCollection.size(), sp, (int) ep, lowerbounds, upperbounds, false);
|
||||
// Objective objective1 = new ObjectiveSqureError(M1,0.1);
|
||||
|
||||
EvolutionMethodSetting EMS = new EvolutionMethodSetting();
|
||||
ObjectiveSqureError TP = new ObjectiveSqureError(sim, experiments, parameterList, speciesCollection, M1, 0.1);
|
||||
|
||||
SRES sres = new SRES(TP, EMS);
|
||||
// System.out.println("test");
|
||||
SRES.Solution solution = sres.run(200).getBestSolution();
|
||||
|
||||
// TODO: report results: take average of error
|
||||
// TODO: weight mean square error. Add small value
|
||||
System.out.println(solution.toString());
|
||||
// TODO: copy best parameter to a new SBML document
|
||||
return null;
|
||||
}
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,27 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public interface AbstractEstimator
|
||||
{
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,133 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.pedi;
|
||||
|
||||
public class GeneProduct
|
||||
{
|
||||
private double mRNADegradationConstant;
|
||||
private double translationConstant;
|
||||
private double proteinDegradationConstant;
|
||||
private double[] unkownTranscriptionConstants;
|
||||
private GeneProduct[] transcFactors;
|
||||
private double[] getLowerBounds;
|
||||
private double[] getUpperBounds;
|
||||
private double mRNALevel;
|
||||
private double proteinLevel;
|
||||
|
||||
public GeneProduct()
|
||||
{
|
||||
|
||||
}
|
||||
|
||||
public double evaluateTranscriptionRate(double[] proteins)
|
||||
{
|
||||
return 0;
|
||||
}
|
||||
|
||||
/**
|
||||
* RNA degradation constants for each gene. NaN if the constant is unknown,
|
||||
* value if the constant is known.
|
||||
*
|
||||
* @return
|
||||
*/
|
||||
public double getmRNADegradationConstant()
|
||||
{
|
||||
return mRNADegradationConstant;
|
||||
}
|
||||
|
||||
public void setmRNADegradationConstant(double mRNADegradationConstant)
|
||||
{
|
||||
this.mRNADegradationConstant = mRNADegradationConstant;
|
||||
}
|
||||
|
||||
public double getTranslationConstant()
|
||||
{
|
||||
return translationConstant;
|
||||
}
|
||||
|
||||
public void setTranslationConstant(double translationConstant)
|
||||
{
|
||||
this.translationConstant = translationConstant;
|
||||
}
|
||||
|
||||
public double getProteinDegradationConstant()
|
||||
{
|
||||
return proteinDegradationConstant;
|
||||
}
|
||||
|
||||
public void setProteinDegradationConstant(double proteinDegradationConstant)
|
||||
{
|
||||
this.proteinDegradationConstant = proteinDegradationConstant;
|
||||
}
|
||||
|
||||
public double[] getUnkownTranscriptionConstants()
|
||||
{
|
||||
return unkownTranscriptionConstants;
|
||||
}
|
||||
|
||||
public void setUnkownTranscriptionConstants(double[] unkownTranscriptionConstants)
|
||||
{
|
||||
this.unkownTranscriptionConstants = unkownTranscriptionConstants;
|
||||
}
|
||||
|
||||
public GeneProduct[] getTranscFactors()
|
||||
{
|
||||
return transcFactors;
|
||||
}
|
||||
|
||||
public void setTranscFactors(GeneProduct[] transcFactors)
|
||||
{
|
||||
this.transcFactors = transcFactors;
|
||||
}
|
||||
|
||||
public double[] getGetLowerBounds()
|
||||
{
|
||||
return getLowerBounds;
|
||||
}
|
||||
|
||||
public void setGetLowerBounds(double[] getLowerBounds)
|
||||
{
|
||||
this.getLowerBounds = getLowerBounds;
|
||||
}
|
||||
|
||||
public double[] getGetUpperBounds()
|
||||
{
|
||||
return getUpperBounds;
|
||||
}
|
||||
|
||||
public void setGetUpperBounds(double[] getUpperBounds)
|
||||
{
|
||||
this.getUpperBounds = getUpperBounds;
|
||||
}
|
||||
|
||||
public double getmRNALevel()
|
||||
{
|
||||
return mRNALevel;
|
||||
}
|
||||
|
||||
public void setmRNALevel(double mRNALevel)
|
||||
{
|
||||
this.mRNALevel = mRNALevel;
|
||||
}
|
||||
|
||||
public double getProteinLevel()
|
||||
{
|
||||
return proteinLevel;
|
||||
}
|
||||
|
||||
public void setProteinLevel(double proteinLevel)
|
||||
{
|
||||
this.proteinLevel = proteinLevel;
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,21 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.pedi;
|
||||
|
||||
public interface PEDIBridge
|
||||
{
|
||||
|
||||
public double[][] simulate(GeneProduct[] genes, double startTime, double endTime, double printInterval);
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,139 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.sres;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class EvolutionMethodSetting
|
||||
{
|
||||
|
||||
/*
|
||||
* * @param verbose if set to true, will print the number of generations
|
||||
* passed and other statistics to {@code stderr}
|
||||
*
|
||||
* @param lambda solution set size
|
||||
*
|
||||
* @param mu number of top-ranking solutions selected to produce new
|
||||
* solution set at each generation
|
||||
*
|
||||
* @param expectedConvergenceRate expected convergence rate
|
||||
*
|
||||
* @param numberOfSweeps number of times stochastic ranking bubble-sort is
|
||||
* applied to solution set
|
||||
*
|
||||
* @param rankingPenalizationFactor constraint breaking penalization factor,
|
||||
* should be in { [0, 1]};
|
||||
*/
|
||||
boolean verbose;
|
||||
int lambda;
|
||||
int mu;
|
||||
// double tau;
|
||||
double expectedConvergenceRate;
|
||||
int numberOfSweeps;
|
||||
double rankingPenalizationFactor;
|
||||
double tauDash;
|
||||
int numberOfgenerations;
|
||||
|
||||
public EvolutionMethodSetting()
|
||||
{
|
||||
this(true, 200, 30, 1.0, 200, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose)
|
||||
{
|
||||
this(verbose, 200, 30, 1.0, 200, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda)
|
||||
{
|
||||
this(verbose, lambda, 30, 1.0, 200, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda, int mu)
|
||||
{
|
||||
this(verbose, lambda, mu, 1.0, 200, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda, int mu, double expectedConvergenceRate)
|
||||
{
|
||||
this(verbose, lambda, mu, expectedConvergenceRate, 200, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda, int mu, double expectedConvergenceRate, int numberOfSweeps)
|
||||
{
|
||||
this(verbose, lambda, mu, expectedConvergenceRate, numberOfSweeps, 1000, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda, int mu, double expectedConvergenceRate, int numberOfSweeps, int numberofgenerations)
|
||||
{
|
||||
this(verbose, lambda, mu, expectedConvergenceRate, numberOfSweeps, numberofgenerations, 0.45);
|
||||
}
|
||||
|
||||
public EvolutionMethodSetting(boolean verbose, int lambda, int mu, double expectedConvergenceRate, int numberOfSweeps, int numberofgenerations, double rankingPenalizationFactor)
|
||||
{
|
||||
this.verbose = verbose;
|
||||
this.lambda = lambda;
|
||||
this.mu = mu;
|
||||
// this.tau = expectedConvergenceRate / Math.sqrt(2 *
|
||||
// Math.sqrt(numberOfFeatures));
|
||||
// this.tauDash = expectedConvergenceRate / Math.sqrt(2 *
|
||||
// numberOfFeatures);
|
||||
this.numberOfSweeps = numberOfSweeps;
|
||||
this.rankingPenalizationFactor = rankingPenalizationFactor;
|
||||
this.numberOfgenerations = numberofgenerations;
|
||||
this.expectedConvergenceRate = expectedConvergenceRate;
|
||||
}
|
||||
|
||||
public boolean getverbose()
|
||||
{
|
||||
return verbose;
|
||||
}
|
||||
|
||||
public int getlamda()
|
||||
{
|
||||
return lambda;
|
||||
}
|
||||
|
||||
public int getmu()
|
||||
{
|
||||
return mu;
|
||||
}
|
||||
|
||||
public double getexpectedConvergenceRate()
|
||||
{
|
||||
return expectedConvergenceRate;
|
||||
}
|
||||
|
||||
public int getnumberOfSweeps()
|
||||
{
|
||||
return numberOfSweeps;
|
||||
}
|
||||
|
||||
public double getrankingPenalizationFactor()
|
||||
{
|
||||
return rankingPenalizationFactor;
|
||||
}
|
||||
|
||||
public double getnumberOfgenerations()
|
||||
{
|
||||
return numberOfgenerations;
|
||||
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,109 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.sres;
|
||||
|
||||
import java.io.BufferedReader;
|
||||
import java.io.File;
|
||||
import java.io.FileInputStream;
|
||||
import java.io.FileNotFoundException;
|
||||
import java.io.IOException;
|
||||
import java.io.InputStreamReader;
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class Modelsettings
|
||||
{
|
||||
|
||||
public List<Double> IC;
|
||||
public int ngenes;
|
||||
public int startp;
|
||||
public int endp;
|
||||
double[] lowerBounds;
|
||||
double[] upperBounds;
|
||||
boolean verbose;
|
||||
int nums;
|
||||
|
||||
public Modelsettings(List<Double> ic, int ngenes, int sp, int ep, double[] lowerbounds, double[] upperbounds, boolean verbose)
|
||||
{
|
||||
|
||||
this.IC = ic;
|
||||
this.ngenes = ngenes;
|
||||
this.startp = sp;
|
||||
this.endp = ep;
|
||||
this.lowerBounds = lowerbounds;
|
||||
this.upperBounds = upperbounds;
|
||||
this.verbose = verbose;
|
||||
}
|
||||
|
||||
public double[][] loaddata(String filename)
|
||||
{
|
||||
// StringBuffer sb=new StringBuffer();
|
||||
String tempstr = null;
|
||||
int lines = 0;
|
||||
int rows = 0;
|
||||
double[][] tmps = new double[100][100];
|
||||
try
|
||||
{
|
||||
// String path="/Users/mfan/Documents/program/data/model1.txt";
|
||||
String path = new String(filename);
|
||||
File file = new File(path);
|
||||
if (!file.exists())
|
||||
{
|
||||
throw new FileNotFoundException();
|
||||
}
|
||||
// BufferedReader br=new BufferedReader(new FileReader(file));
|
||||
// while((tempstr=br.readLine())!=null)
|
||||
// sb.append(tempstr);
|
||||
FileInputStream fis = new FileInputStream(file);
|
||||
BufferedReader br = new BufferedReader(new InputStreamReader(fis));
|
||||
|
||||
while ((tempstr = br.readLine()) != null)
|
||||
{
|
||||
String s[] = tempstr.split(" ");
|
||||
// System.out.println(s.length);
|
||||
for (int i = 0; i < s.length; ++i)
|
||||
{
|
||||
tmps[lines][i] = Double.parseDouble(s[i]);
|
||||
}
|
||||
lines = lines + 1;
|
||||
rows = s.length;
|
||||
}
|
||||
br.close();
|
||||
// double rows=tmps[0].length;
|
||||
|
||||
}
|
||||
catch (IOException ex)
|
||||
{
|
||||
System.out.println(ex.getStackTrace());
|
||||
}
|
||||
double[][] data = new double[lines][rows];
|
||||
for (int i = 0; i < lines; i++)
|
||||
{
|
||||
for (int j = 0; j < rows; j++)
|
||||
{
|
||||
data[i][j] = tmps[i][j];
|
||||
}
|
||||
}
|
||||
return data;
|
||||
// return sb.toString();
|
||||
}
|
||||
|
||||
}
|
||||
|
|
@ -0,0 +1,257 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.sres;
|
||||
|
||||
import java.util.Arrays;
|
||||
|
||||
/**
|
||||
* A class representing a real multivariate objective function and constraints
|
||||
* that are subject to optimization. The objective functionmethod is left for
|
||||
* the user to specify. Arbitrary value range is allowed for objective function.
|
||||
* Constraints are provided as { constraint_value <= 0} inequalities, the
|
||||
* computation of array should be implemented by user.
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public abstract class Objective
|
||||
{
|
||||
private final int numberOfFeatures;
|
||||
private final double[] featureLowerBounds, featureUpperBounds;
|
||||
private final double[] mutationRates;
|
||||
private final boolean maximizationProblem;
|
||||
|
||||
/**
|
||||
* Generates an objective function instance and specifies the feature
|
||||
* (argument) space.
|
||||
*
|
||||
* @param featureLowerBounds
|
||||
* lower bounds of feature space
|
||||
* @param featureUpperBounds
|
||||
* upper bounds of feature space
|
||||
* @param maximizationProblem
|
||||
* if set to true, the objective function is subject to
|
||||
* maximization task, minimization will be performed otherwise
|
||||
*/
|
||||
public Objective(Modelsettings Ms)
|
||||
{
|
||||
|
||||
// this.numberOfFeatures = featureLowerBounds.length;
|
||||
this.numberOfFeatures = Ms.lowerBounds.length;
|
||||
|
||||
if (Ms.upperBounds.length != numberOfFeatures)
|
||||
{
|
||||
throw new IllegalArgumentException("Lengths of upper and lower bounds should match.");
|
||||
}
|
||||
this.featureLowerBounds = Ms.lowerBounds;
|
||||
this.featureUpperBounds = Ms.upperBounds;
|
||||
for (int i = 0; i < numberOfFeatures; i++)
|
||||
{
|
||||
if (featureUpperBounds[i] < featureLowerBounds[i])
|
||||
{
|
||||
throw new IllegalArgumentException("Feature upper bound is smaller than lower bound for index " + i + ".");
|
||||
}
|
||||
}
|
||||
this.maximizationProblem = Ms.verbose;
|
||||
|
||||
this.mutationRates = new double[numberOfFeatures];
|
||||
double factor = Math.sqrt(numberOfFeatures);
|
||||
|
||||
for (int i = 0; i < numberOfFeatures; i++)
|
||||
{
|
||||
mutationRates[i] = (featureUpperBounds[i] - featureLowerBounds[i]) / factor;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Generate a random feature array based on feature space constraints.
|
||||
*
|
||||
* @return a random array from feature space
|
||||
*/
|
||||
double[] generateFeatureVector()
|
||||
{
|
||||
double[] features = new double[numberOfFeatures];
|
||||
|
||||
for (int i = 0; i < numberOfFeatures; i++)
|
||||
{
|
||||
double x = SRES.random.nextDouble();
|
||||
features[i] = featureLowerBounds[i] + x * (featureUpperBounds[i] - featureLowerBounds[i]);
|
||||
}
|
||||
|
||||
return features;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the mutation rates for { SRES} algorithm.
|
||||
*
|
||||
* @return an array of feature mutation rates
|
||||
*/
|
||||
final double[] getMutationRates()
|
||||
{
|
||||
return mutationRates;
|
||||
}
|
||||
|
||||
/**
|
||||
* Check whether a given feature value belongs to feature space.
|
||||
*
|
||||
* @param feature
|
||||
* feature value
|
||||
* @param index
|
||||
* feature index
|
||||
* @return true if feature value is in specified bounds, false otherwise
|
||||
*/
|
||||
final boolean inBounds(double feature, int index)
|
||||
{
|
||||
return featureLowerBounds[index] <= feature && feature <= featureUpperBounds[index];
|
||||
}
|
||||
|
||||
/**
|
||||
* A user-implemented function that computes the value of objective function
|
||||
* and constraint values using provided features (parameters).
|
||||
*
|
||||
* @param features
|
||||
* an array of objective function features
|
||||
* @return objective function evaluation result
|
||||
* @see com.antigenomics.jsres.Objective.Result
|
||||
*/
|
||||
public abstract Result evaluate(double[] features);
|
||||
|
||||
/**
|
||||
* Gets the dimensionality of feature (parameter) space.
|
||||
*
|
||||
* @return length of feature arrays
|
||||
*/
|
||||
public final int getNumberOfFeatures()
|
||||
{
|
||||
return numberOfFeatures;
|
||||
}
|
||||
|
||||
public double[] getfeatureLowerBounds()
|
||||
{
|
||||
return featureLowerBounds;
|
||||
}
|
||||
|
||||
public double[] getfeatureUpperBounds()
|
||||
{
|
||||
return featureUpperBounds;
|
||||
}
|
||||
|
||||
/**
|
||||
* Tells if the problem is maximization or minimization one.
|
||||
*
|
||||
* @return true for maximization task, false for minimization
|
||||
*/
|
||||
public boolean isMaximizationProblem()
|
||||
{
|
||||
return maximizationProblem;
|
||||
}
|
||||
|
||||
/**
|
||||
* An object containing results of objective function and constraint
|
||||
* evaluation.
|
||||
*/
|
||||
public class Result
|
||||
{
|
||||
private final double value, penalty;
|
||||
private final double[] constraintValues;
|
||||
|
||||
/**
|
||||
* Creates a new instance of objective function evaluation result.
|
||||
*
|
||||
* @param value
|
||||
* objective function value
|
||||
*/
|
||||
public Result(double value)
|
||||
{
|
||||
this.value = value;
|
||||
this.constraintValues = new double[0];
|
||||
this.penalty = 0;
|
||||
}
|
||||
|
||||
/**
|
||||
* Creates a new instance of objective function evaluation result.
|
||||
* Constraints should be re-written as {constraint_value <= 0}
|
||||
* inequalities and {@code constraint_value} array should be provided.
|
||||
*
|
||||
* @param value
|
||||
* objective function value
|
||||
* @param constraintValues
|
||||
* constraint values
|
||||
*/
|
||||
public Result(double value, double[] constraintValues)
|
||||
{
|
||||
this.value = value;
|
||||
this.constraintValues = constraintValues;
|
||||
double penalty = 0;
|
||||
for (double p : constraintValues)
|
||||
{
|
||||
double pp = Math.max(0, p);
|
||||
penalty += pp * pp;
|
||||
}
|
||||
this.penalty = penalty;
|
||||
}
|
||||
|
||||
/**
|
||||
* Get the array of constraint values.
|
||||
*
|
||||
* @return constraint value array
|
||||
*/
|
||||
public double[] getConstraintValues()
|
||||
{
|
||||
return constraintValues;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the objective function value.
|
||||
*
|
||||
* @return objective function value
|
||||
*/
|
||||
public double getValue()
|
||||
{
|
||||
return value;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the { SRES} penalty computed from objective values. Penalty
|
||||
* value is computed as {@code sum(max(0, constraint_value)^2}.
|
||||
*
|
||||
* @return penalty value
|
||||
*/
|
||||
double getPenalty()
|
||||
{
|
||||
return penalty;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the parent objective function instance.
|
||||
*
|
||||
* @return parent objective function instance
|
||||
*/
|
||||
public Objective getObjective()
|
||||
{
|
||||
return Objective.this;
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
@Override
|
||||
public String toString()
|
||||
{
|
||||
return "RESULT objective function value is " + value + ", constraint values are " + Arrays.toString(constraintValues);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,148 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.sres;
|
||||
|
||||
import static java.lang.Math.abs;
|
||||
|
||||
import java.util.List;
|
||||
|
||||
import edu.utah.ece.async.analysis.simulation.hierarchical.HierarchicalSimulation;
|
||||
import edu.utah.ece.async.analysis.simulation.hierarchical.methods.HierarchicalODERKSimulator;
|
||||
import edu.utah.ece.async.learn.genenet.Experiments;
|
||||
import edu.utah.ece.async.learn.genenet.SpeciesCollection;
|
||||
|
||||
/**
|
||||
*
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public class ObjectiveSqureError extends Objective
|
||||
{
|
||||
private final double valueAtSolution;
|
||||
private final double absolutePrecision, relativePrecision;
|
||||
private final int allowedViolatedConstraintsCount;
|
||||
private final Modelsettings Models;
|
||||
|
||||
private Experiments experiment;
|
||||
private List<String> listOfParams;
|
||||
private String[] speciesCollection;
|
||||
private HierarchicalSimulation sim;
|
||||
|
||||
/*
|
||||
* public TestProblem(double[] featureUpperBounds, double[]
|
||||
* featureLowerBounds, boolean maximizationProblem, double valueAtSolution)
|
||||
* { this(featureUpperBounds, featureLowerBounds, maximizationProblem,
|
||||
* valueAtSolution, 0); }
|
||||
*
|
||||
* public TestProblem(double[] featureUpperBounds, double[]
|
||||
* featureLowerBounds, boolean maximizationProblem, double valueAtSolution,
|
||||
* int allowedViolatedConstraintsCount) { this(featureUpperBounds,
|
||||
* featureLowerBounds, maximizationProblem, valueAtSolution,
|
||||
* allowedViolatedConstraintsCount, 1e-6, 1e-3); }
|
||||
*/
|
||||
|
||||
public ObjectiveSqureError(HierarchicalSimulation sim, Experiments experiments, List<String> parameterList, SpeciesCollection speciesCollection, Modelsettings Ms, double valueAtSolution)
|
||||
{
|
||||
super(Ms);
|
||||
this.valueAtSolution = valueAtSolution;
|
||||
this.relativePrecision = 0;
|
||||
this.allowedViolatedConstraintsCount = 0;
|
||||
this.absolutePrecision = 0;
|
||||
this.Models = Ms;
|
||||
this.experiment = experiments;
|
||||
this.listOfParams = parameterList;
|
||||
this.speciesCollection = new String[speciesCollection.size() + 1];
|
||||
for (String species : speciesCollection.getInterestingSpecies())
|
||||
{
|
||||
this.speciesCollection[speciesCollection.getColumn(species)] = species;
|
||||
}
|
||||
this.sim = sim;
|
||||
}
|
||||
|
||||
public ObjectiveSqureError(Modelsettings Ms, double valueAtSolution, int allowedViolatedConstraintsCount, double absolutePrecision, double relativePrecision)
|
||||
{
|
||||
super(Ms);
|
||||
this.valueAtSolution = valueAtSolution;
|
||||
this.absolutePrecision = absolutePrecision;
|
||||
this.relativePrecision = relativePrecision;
|
||||
this.allowedViolatedConstraintsCount = allowedViolatedConstraintsCount;
|
||||
this.Models = Ms;
|
||||
}
|
||||
|
||||
@Override
|
||||
public Result evaluate(double[] features)
|
||||
{
|
||||
double sum = 0;
|
||||
List<List<Double>> experiment = this.experiment.getExperiments().get(0);
|
||||
|
||||
HierarchicalODERKSimulator odeSim = (HierarchicalODERKSimulator) sim;
|
||||
|
||||
for (int i = 1; i < speciesCollection.length; i++)
|
||||
{
|
||||
odeSim.setTopLevelValue(speciesCollection[i], experiment.get(0).get(i));
|
||||
}
|
||||
|
||||
for (int i = 0; i < listOfParams.size(); i++)
|
||||
{
|
||||
odeSim.setTopLevelValue(listOfParams.get(i), features[i]);
|
||||
}
|
||||
for (int i = 0; i < experiment.size() - 1; i++)
|
||||
{
|
||||
|
||||
odeSim.setTimeLimit(experiment.get(i + 1).get(0));
|
||||
odeSim.simulate();
|
||||
for (int j = 1; j < speciesCollection.length; j++)
|
||||
{
|
||||
double tmp = odeSim.getTopLevelValue(speciesCollection[j]) - experiment.get(i + 1).get(j);
|
||||
tmp = tmp * tmp;
|
||||
sum = sum + tmp;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
odeSim.setCurrentTime(0);
|
||||
|
||||
odeSim.setTimeLimit(0);
|
||||
|
||||
odeSim.setupForNewRun(0);
|
||||
|
||||
return new Result(sum);
|
||||
}
|
||||
|
||||
public boolean isSolved(SRES.Solution solution)
|
||||
{
|
||||
Objective.Result result = solution.getObjectiveResult();
|
||||
|
||||
int violatedConstraints = 0;
|
||||
|
||||
for (int i = 0; i < result.getConstraintValues().length; i++)
|
||||
{
|
||||
if (result.getConstraintValues()[i] > 0)
|
||||
{
|
||||
violatedConstraints++;
|
||||
}
|
||||
}
|
||||
|
||||
double absoluteError = abs(valueAtSolution - result.getValue()), relativeError = absoluteError / abs(valueAtSolution);
|
||||
System.out.print(allowedViolatedConstraintsCount);
|
||||
System.out.print(absolutePrecision);
|
||||
System.out.print(valueAtSolution);
|
||||
System.out.print(violatedConstraints <= allowedViolatedConstraintsCount && absoluteError <= absolutePrecision && (abs(valueAtSolution) < relativePrecision || relativeError <= relativePrecision));
|
||||
return violatedConstraints <= allowedViolatedConstraintsCount && absoluteError <= absolutePrecision && (abs(valueAtSolution) < relativePrecision || relativeError <= relativePrecision);
|
||||
|
||||
}
|
||||
}
|
||||
|
|
@ -0,0 +1,484 @@
|
|||
/*******************************************************************************
|
||||
*
|
||||
* This file is part of iBioSim. Please visit <http://www.async.ece.utah.edu/ibiosim>
|
||||
* for the latest version of iBioSim.
|
||||
*
|
||||
* Copyright (C) 2017 University of Utah
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or modify it
|
||||
* under the terms of the Apache License. A copy of the license agreement is provided
|
||||
* in the file named "LICENSE.txt" included with this software distribution
|
||||
* and also available online at <http://www.async.ece.utah.edu/ibiosim/License>.
|
||||
*
|
||||
*******************************************************************************/
|
||||
package edu.utah.ece.async.learn.parameterestimator.methods.sres;
|
||||
|
||||
import java.util.Arrays;
|
||||
import java.util.Random;
|
||||
|
||||
import edu.utah.ece.async.learn.parameterestimator.methods.AbstractEstimator;
|
||||
|
||||
/**
|
||||
* A class that implements Stochastic Ranking Evolutionary Strategy (SRES), an
|
||||
* evolutionary algorithm for constrained optimization of real multivariate
|
||||
* objective functions. User should provide an objective function instance
|
||||
* inherited from abstract {@link com.antigenomics.jsres.Objective} class. The
|
||||
* algorithm is executed via { #run} method. Objective function evaluation is
|
||||
* optimized by implementing .
|
||||
*
|
||||
* @author
|
||||
* @author Chris Myers
|
||||
* @author <a href="http://www.async.ece.utah.edu/ibiosim#Credits"> iBioSim Contributors </a>
|
||||
* @version %I%
|
||||
*/
|
||||
public final class SRES implements AbstractEstimator
|
||||
{
|
||||
static final Random random = new Random(51102);
|
||||
private static final int defaultNumberOfGenerations = 1000, boundingTries = 10;
|
||||
private final Objective objective;
|
||||
private final double tau, tauDash, rankingPenalizationFactor;
|
||||
private final int lambda, mu, numberOfFeatures, numberOfSweeps;
|
||||
private final double[] featureLowerBounds;
|
||||
private final double[] featureUpperBounds;
|
||||
private final boolean verbose;
|
||||
|
||||
/**
|
||||
* Creates an instance of SRES algorithm for a given objective function.
|
||||
* Default algorithm parameters from Runarsson TP and Yao X work are used.
|
||||
*
|
||||
* @param objective
|
||||
* objective function instance
|
||||
*/
|
||||
/*
|
||||
* public SRES(Objective objective) { this(objective, true); }
|
||||
*/
|
||||
/**
|
||||
* Creates an instance of SRES algorithm for a given objective function.
|
||||
* Default algorithm parameters from Runarsson TP and Yao X work are used.
|
||||
*
|
||||
* @param objective
|
||||
* objective function instance
|
||||
* @param verbose
|
||||
* if set to true, will print the number of generations passed
|
||||
* and other statistics to {@code stderr}
|
||||
*/
|
||||
/*
|
||||
* public SRES(Objective objective, boolean verbose) { this(objective,
|
||||
* verbose, 200, 30, 1.0, 200, 0.45); }
|
||||
*/
|
||||
|
||||
/**
|
||||
* Creates an instance of SRES algorithm for a given objective function.
|
||||
*
|
||||
* @param objective
|
||||
* objective function instance
|
||||
* @param verbose
|
||||
* if set to true, will print the number of generations passed
|
||||
* and other statistics to {@code stderr}
|
||||
* @param lambda
|
||||
* solution set size
|
||||
* @param mu
|
||||
* number of top-ranking solutions selected to produce new
|
||||
* solution set at each generation
|
||||
* @param expectedConvergenceRate
|
||||
* expected convergence rate
|
||||
* @param numberOfSweeps
|
||||
* number of times stochastic ranking bubble-sort is applied to
|
||||
* solution set
|
||||
* @param rankingPenalizationFactor
|
||||
* constraint breaking penalization factor, should be in { [0,
|
||||
* 1]}; no penalization is performed if set to 1, all solutions
|
||||
* will be penalized if set to 0
|
||||
*/
|
||||
/*
|
||||
* public SRES(Objective objective, boolean verbose, int lambda, int mu,
|
||||
* double expectedConvergenceRate, int numberOfSweeps, double
|
||||
* rankingPenalizationFactor) { this.objective = objective; this.verbose =
|
||||
* verbose; this.numberOfFeatures = objective.getNumberOfFeatures();
|
||||
* this.featureLowerBounds = objective.getfeatureLowerBounds();
|
||||
* this.featureUpperBounds = objective.getfeatureUpperBounds(); this.lambda
|
||||
* = lambda; this.mu = mu; this.tau = expectedConvergenceRate / Math.sqrt(2
|
||||
* * Math.sqrt(numberOfFeatures)); this.tauDash = expectedConvergenceRate /
|
||||
* Math.sqrt(2 * numberOfFeatures); this.numberOfSweeps = numberOfSweeps;
|
||||
* this.rankingPenalizationFactor = rankingPenalizationFactor; }
|
||||
*/
|
||||
|
||||
public SRES(Objective objective, EvolutionMethodSetting EMS)
|
||||
{
|
||||
this.objective = objective;
|
||||
this.verbose = EMS.getverbose();
|
||||
this.numberOfFeatures = objective.getNumberOfFeatures();
|
||||
this.featureLowerBounds = objective.getfeatureLowerBounds();
|
||||
this.featureUpperBounds = objective.getfeatureUpperBounds();
|
||||
this.lambda = EMS.getlamda();
|
||||
this.mu = EMS.getmu();
|
||||
this.tau = EMS.getexpectedConvergenceRate() / Math.sqrt(2 * Math.sqrt(numberOfFeatures));
|
||||
this.tauDash = EMS.getexpectedConvergenceRate() / Math.sqrt(2 * numberOfFeatures);
|
||||
this.numberOfSweeps = EMS.getnumberOfSweeps();
|
||||
this.rankingPenalizationFactor = EMS.getrankingPenalizationFactor();
|
||||
}
|
||||
|
||||
/**
|
||||
* Runs the SRES algorithm for {@value #defaultNumberOfGenerations}
|
||||
* generations.
|
||||
*
|
||||
* @return a sorted (from best to worst) set of solutions obtained after
|
||||
* running the algorithm
|
||||
*/
|
||||
public SolutionSet run()
|
||||
{
|
||||
return run(defaultNumberOfGenerations);
|
||||
}
|
||||
|
||||
/**
|
||||
* Generate a random feature array based on feature space constraints.
|
||||
*
|
||||
* @return a random array from feature space
|
||||
*/
|
||||
double[] generateFeatureVector()
|
||||
{
|
||||
double[] features = new double[numberOfFeatures];
|
||||
|
||||
for (int i = 0; i < numberOfFeatures; i++)
|
||||
{
|
||||
double x = random.nextDouble();
|
||||
features[i] = featureLowerBounds[i] + x * (featureUpperBounds[i] - featureLowerBounds[i]);
|
||||
}
|
||||
|
||||
return features;
|
||||
}
|
||||
|
||||
/**
|
||||
* Runs the SRES algorithm for {@code numberOfGenerations} generations.
|
||||
*
|
||||
* @param numberOfGenerations
|
||||
* number of generations to run
|
||||
* @return a sorted (from best to worst) set of solutions obtained after
|
||||
* running the algorithm
|
||||
*/
|
||||
public SolutionSet run(int numberOfGenerations)
|
||||
{
|
||||
SolutionSet solutionSet = new SolutionSet();
|
||||
|
||||
for (int i = 0; i < numberOfGenerations; i++)
|
||||
{
|
||||
solutionSet.evaluate();
|
||||
solutionSet.sort();
|
||||
|
||||
if (verbose && i % 100 == 0)
|
||||
{
|
||||
Solution bestSolution = solutionSet.getBestSolution();
|
||||
System.err.println("SRES ran for " + i + " generations, best solution fitness is " + bestSolution.getFitness() + ", penalty is " + bestSolution.getPenalty());
|
||||
}
|
||||
|
||||
solutionSet = solutionSet.evolve();
|
||||
}
|
||||
|
||||
solutionSet.evaluate();
|
||||
solutionSet.sort();
|
||||
|
||||
if (verbose)
|
||||
{
|
||||
Solution bestSolution = solutionSet.getBestSolution();
|
||||
System.err.println("SRES finished, best solution fitness is " + bestSolution.getFitness() + ", penalty is " + bestSolution.getPenalty());
|
||||
}
|
||||
|
||||
return solutionSet;
|
||||
}
|
||||
|
||||
/**
|
||||
* A set of solutions for the objective function and constraints.
|
||||
*/
|
||||
public class SolutionSet
|
||||
{
|
||||
private final Solution[] solutions;
|
||||
|
||||
/**
|
||||
* Generates initial solution set at random.
|
||||
*/
|
||||
SolutionSet()
|
||||
{
|
||||
this.solutions = new Solution[lambda];
|
||||
for (int i = 0; i < lambda; i++)
|
||||
{
|
||||
solutions[i] = new Solution(generateFeatureVector(), objective.getMutationRates());
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Constructs a new instance of solution set.
|
||||
*
|
||||
* @param solutions
|
||||
* solutions array
|
||||
*/
|
||||
SolutionSet(Solution[] solutions)
|
||||
{
|
||||
this.solutions = solutions;
|
||||
}
|
||||
|
||||
/**
|
||||
* Sorts solutions using stochastic ranking bubble sort.
|
||||
*/
|
||||
void sort()
|
||||
{
|
||||
for (int i = 0; i < numberOfSweeps; i++)
|
||||
{
|
||||
boolean swapped = false;
|
||||
for (int j = 0; j < lambda - 1; j++)
|
||||
{
|
||||
double p = random.nextDouble(), p1 = solutions[j].getPenalty(), p2 = solutions[j + 1].getPenalty();
|
||||
if (p < rankingPenalizationFactor || (p1 == 0 && p2 == 0))
|
||||
{
|
||||
// no solution break constraints or penalization is not
|
||||
// applied
|
||||
if (solutions[j].getFitness() < solutions[j + 1].getFitness())
|
||||
{
|
||||
swap(j, j + 1);
|
||||
swapped = true;
|
||||
}
|
||||
}
|
||||
else if (p1 > p2)
|
||||
{
|
||||
// constraint penalization
|
||||
swap(j, j + 1);
|
||||
swapped = true;
|
||||
}
|
||||
}
|
||||
if (!swapped)
|
||||
{
|
||||
// sorted
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Evolve the population by selecting top { #mu} solutions and
|
||||
* generating new population of { #lambda} solutions using recombination
|
||||
* of mutation rates and random weighted mutation of features.
|
||||
*
|
||||
* @return evolved solution set, no evaluation of objective function is
|
||||
* constraints is performed on this step
|
||||
*/
|
||||
SolutionSet evolve()
|
||||
{
|
||||
Solution[] newSolutions = new Solution[lambda];
|
||||
|
||||
int j = 0;
|
||||
for (int i = 0; i < lambda; i++)
|
||||
{
|
||||
// mutation rates are sampled from top solutions
|
||||
double[] sampledMutationRates = new double[numberOfFeatures];
|
||||
|
||||
for (int k = 0; k < numberOfFeatures; k++)
|
||||
{
|
||||
sampledMutationRates[k] = solutions[random.nextInt(mu)].mutationRates[k];
|
||||
}
|
||||
|
||||
// top individual generates offspring
|
||||
newSolutions[i] = solutions[j].generateOffspring(sampledMutationRates);
|
||||
|
||||
if (j++ > mu)
|
||||
{
|
||||
// cycle top mu solutions
|
||||
j = 0;
|
||||
}
|
||||
}
|
||||
|
||||
return new SolutionSet(newSolutions);
|
||||
}
|
||||
|
||||
/**
|
||||
* Evaluates objective function and constraints for the solution set in
|
||||
* parallel.
|
||||
*/
|
||||
void evaluate()
|
||||
{
|
||||
// Arrays.stream(solutions).parallel().forEach(Solution::evaluate);
|
||||
for (Solution solution : solutions)
|
||||
{
|
||||
solution.evaluate();
|
||||
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Swaps two solutions.
|
||||
*
|
||||
* @param i
|
||||
* index of fist solution
|
||||
* @param j
|
||||
* index of second solution
|
||||
*/
|
||||
private void swap(int i, int j)
|
||||
{
|
||||
Solution tmp = solutions[i];
|
||||
solutions[i] = solutions[j];
|
||||
solutions[j] = tmp;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the solutions array.
|
||||
*
|
||||
* @return array of solutions in this set
|
||||
*/
|
||||
public Solution[] getSolutions()
|
||||
{
|
||||
return solutions;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the best solution from solution set.
|
||||
*
|
||||
* @return best solution
|
||||
*/
|
||||
public Solution getBestSolution()
|
||||
{
|
||||
return solutions[0];
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* A class representing solution to a problem that consists of objective
|
||||
* function and constraints.
|
||||
*/
|
||||
public class Solution
|
||||
{
|
||||
private final double[] features, mutationRates;
|
||||
private Objective.Result objectiveResult = null;
|
||||
|
||||
/**
|
||||
* Creates a solution, features and mutation rates are not set.
|
||||
*/
|
||||
Solution()
|
||||
{
|
||||
this.features = new double[numberOfFeatures];
|
||||
this.mutationRates = new double[numberOfFeatures];
|
||||
}
|
||||
|
||||
/**
|
||||
* Creates a solution with specified features and mutation rates. Used
|
||||
* to instantiate a descendants of solutions selected by stochastic
|
||||
* ranking.
|
||||
*
|
||||
* @param features
|
||||
* feature array
|
||||
* @param mutationRates
|
||||
* array of mutation rates
|
||||
*/
|
||||
Solution(double[] features, double[] mutationRates)
|
||||
{
|
||||
this.features = features;
|
||||
this.mutationRates = mutationRates;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the features (objective function parameters) for this solution.
|
||||
*
|
||||
* @return objective function parameters
|
||||
*/
|
||||
public double[] getFeatures()
|
||||
{
|
||||
return features;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the result of evaluation of objective function and constraints
|
||||
* for this solution.
|
||||
*
|
||||
* @return objective function and constraints evaluation results
|
||||
*/
|
||||
public Objective.Result getObjectiveResult()
|
||||
{
|
||||
return objectiveResult;
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the fitness of this solution.
|
||||
*
|
||||
* @return {@code value} of objective function for maximization problem,
|
||||
* {@code -value} for minimization problem
|
||||
* @see com.antigenomics.jsres.Objective.Result#getValue
|
||||
*/
|
||||
double getFitness()
|
||||
{
|
||||
return objectiveResult.getObjective().isMaximizationProblem() ? objectiveResult.getValue() : -objectiveResult.getValue();
|
||||
}
|
||||
|
||||
/**
|
||||
* Gets the penalty value of this solution.
|
||||
*
|
||||
* @return penalty value, sum of squares of constraint values for
|
||||
* constraints that are violated
|
||||
* @see com.antigenomics.jsres.Objective.Result#getPenalty
|
||||
*/
|
||||
double getPenalty()
|
||||
{
|
||||
return objectiveResult.getPenalty();
|
||||
}
|
||||
|
||||
/**
|
||||
* Generates an offspring for this solution. Mutation rates are
|
||||
* recombined and applied to randomly mutate the feature vector.
|
||||
*
|
||||
* @param sampledMutationRates
|
||||
* random sample of mutation rates from top {@link #mu}
|
||||
* solutions for recombination
|
||||
* @return a recombined and mutated offspring
|
||||
*/
|
||||
Solution generateOffspring(double[] sampledMutationRates)
|
||||
{
|
||||
Solution offspring = new Solution();
|
||||
|
||||
double globalLearningRate = tauDash * random.nextGaussian();
|
||||
|
||||
double[] mutationRateConstraints = objective.getMutationRates();
|
||||
|
||||
for (int i = 0; i < numberOfFeatures; i++)
|
||||
{
|
||||
// global intermediate recombination for mutation rates
|
||||
double newMutationRate = (mutationRates[i] + sampledMutationRates[i]) / 2 *
|
||||
// lognormal update for mutation rates
|
||||
Math.exp(globalLearningRate + tau * random.nextGaussian());
|
||||
// important: prevent mutation rates from growing to infinity
|
||||
offspring.mutationRates[i] = Math.min(newMutationRate, mutationRateConstraints[i]);
|
||||
// generate offspring features
|
||||
offspring.features[i] = features[i]; // in case fail to bound
|
||||
|
||||
double newFeatureValue;
|
||||
for (int j = 0; j < boundingTries; j++)
|
||||
{
|
||||
// try generating new feature value and check whether it is
|
||||
// in bounds
|
||||
newFeatureValue = features[i] + newMutationRate * random.nextGaussian();
|
||||
if (objective.inBounds(newFeatureValue, i))
|
||||
{
|
||||
offspring.features[i] = newFeatureValue;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
return offspring;
|
||||
}
|
||||
|
||||
/**
|
||||
* Evaluates given solution using objective function and constraints
|
||||
* specified in parent {@code SRES} algorithm.
|
||||
*/
|
||||
void evaluate()
|
||||
{
|
||||
objectiveResult = objective.evaluate(features);
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
@Override
|
||||
public String toString()
|
||||
{
|
||||
return "SOLUTION " + Arrays.toString(features) + ",fitness is " + getFitness() + ", penalty is " + getPenalty() + "\n" + (objectiveResult == null ? "OBJECTIVE NOT EVALUATED" : objectiveResult.toString());
|
||||
}
|
||||
}
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue