Skip to content
Snippets Groups Projects
Commit 9c93a693 authored by Lennart Eichhorn's avatar Lennart Eichhorn
Browse files

Added option to generate images every x generations

parent c7310342
Branches
Tags
No related merge requests found
Hide whitespace changes
Inline Side-by-side
src/main/java/MainClasses/Config.java
+ 6
− 0
View file @ 9c93a693
......@@ -47,6 +47,7 @@ public class Config {
private int imageFps;
private int imagesToFpsIncrease;
private int imageFpsMax;
private int imageInterval;
private boolean zoom;
// For images
......@@ -188,6 +189,7 @@ public class Config {
imageFps = Integer.parseInt(this.properties.getProperty("imgFps"));
imagesToFpsIncrease = Integer.parseInt(this.properties.getProperty("imagesToFpsIncrease"));
imageFpsMax = Integer.parseInt(this.properties.getProperty("imgFpsMax"));
imageInterval = Integer.parseInt(this.properties.getProperty("imageInterval"));
zoom = this.properties.getProperty("zoom").equals("true");
}
......@@ -295,6 +297,10 @@ public class Config {
return imageFpsMax;
}
public int getImageInterval() {
return imageInterval;
}
public boolean isZoom() {
return zoom;
}
......
src/main/java/MainClasses/GeneticAlgorithm.java
+ 1
− 0
View file @ 9c93a693
......@@ -143,6 +143,7 @@ public class GeneticAlgorithm {
// mutateGeneration(Mutator e);
private void evaluateGeneration() {
for (int i = 0; i < population.length; i++) {
this.population[i] = this.evaluator.evaluateFitness(this.population[i]);
}
}
......
src/main/java/Visualizers/BestFoldingToImage.java
+ 97
− 87
View file @ 9c93a693
......@@ -40,123 +40,133 @@ public class BestFoldingToImage implements Visualizer {
}
public void drawProtein(Candidate[] generation, GeneticAlgorithm geneticAlgorithm) {
String filename = config.getImageSequenceDirectory() + "/" + config.getJobName() + "_" + geneticAlgorithm.generation + ".png";
//TODO This should probably be in the new Generation Class
Candidate bestCandidateOfGeneration = generation[0];
for (Candidate evaluatedCandidate : generation) {
if(bestCandidateOfGeneration.getFitness() < evaluatedCandidate.getFitness()){
bestCandidateOfGeneration=evaluatedCandidate;
if(geneticAlgorithm.generation% config.getImageInterval() == 0 || geneticAlgorithm.generation == config.getTotalGenerations()) {
String filename = config.getImageSequenceDirectory() + "/" + config.getJobName() + "_"
+ geneticAlgorithm.generation + ".png";
//TODO This should probably be in the new Generation Class
Candidate bestCandidateOfGeneration = generation[0];
for (Candidate evaluatedCandidate : generation) {
if (bestCandidateOfGeneration.getFitness() < evaluatedCandidate.getFitness()) {
bestCandidateOfGeneration = evaluatedCandidate;
}
}
}
ArrayList<Vertex> vertexList = bestCandidateOfGeneration.getVertices();
ArrayList<Vertex> vertexList = bestCandidateOfGeneration.getVertices();
Cell[][] cellArray = Visualizer.convertProteinTo2DArray(vertexList, this.isHydrophobic);
Cell[][] cellArray = Visualizer.convertProteinTo2DArray(vertexList, this.isHydrophobic);
int height = (cellArray.length * pixelsPerCell) + margin * 2;
int width = (cellArray[0].length * pixelsPerCell) + margin * 2;
int height = (cellArray.length * pixelsPerCell) + margin * 2;
int width = (cellArray[0].length * pixelsPerCell) + margin * 2;
if (height > maxHeight) {
maxHeight = height;
}
if (width > maxWidth) {
maxWidth = width;
}
if (height > maxHeight) {
maxHeight = height;
}
if (width > maxWidth) {
maxWidth = width;
}
BufferedImage image = new BufferedImage(width, height, BufferedImage.TYPE_INT_RGB);
Graphics2D g2 = image.createGraphics();
g2.setRenderingHint(RenderingHints.KEY_TEXT_ANTIALIASING, RenderingHints.VALUE_TEXT_ANTIALIAS_ON);
g2.setFont(config.getFont());
FontMetrics metrics = g2.getFontMetrics();
int ascent = metrics.getAscent();
BufferedImage image = new BufferedImage(width, height, BufferedImage.TYPE_INT_RGB);
Graphics2D g2 = image.createGraphics();
g2.setRenderingHint(RenderingHints.KEY_TEXT_ANTIALIASING,
RenderingHints.VALUE_TEXT_ANTIALIAS_ON);
g2.setFont(config.getFont());
FontMetrics metrics = g2.getFontMetrics();
int ascent = metrics.getAscent();
// Background
g2.setColor(config.getImageBackground());
g2.fillRect(0, 0, width, height);
// Background
g2.setColor(config.getImageBackground());
g2.fillRect(0, 0, width, height);
for (int yIndex = cellArray.length-1; yIndex >= 0; yIndex--) {
int yIndexPosition = cellArray.length-1 - yIndex;
for (int yIndex = cellArray.length - 1; yIndex >= 0; yIndex--) {
int yIndexPosition = cellArray.length - 1 - yIndex;
for (int xIndex = 0; xIndex < cellArray[0].length; xIndex++) {
Color aminoColor = null;
for (int xIndex = 0; xIndex < cellArray[0].length; xIndex++) {
Color aminoColor = null;
State cellState = cellArray[yIndex][xIndex].getRelevantDrawState();
State cellState = cellArray[yIndex][xIndex].getRelevantDrawState();
switch (cellState) {
case Empty:
break;
switch (cellState) {
case Empty:
break;
case Hydrophobic:
case HydrophobicMulti:
aminoColor = config.getImageHydrophobic();
break;
case Hydrophobic:
case HydrophobicMulti:
aminoColor = config.getImageHydrophobic();
break;
case Hydrophilic:
case HydrophilicMulti:
aminoColor = config.getImageHydrophilic();
break;
case Hydrophilic:
case HydrophilicMulti:
aminoColor = config.getImageHydrophilic();
break;
case Mixed:
aminoColor = config.getImageMixed();
break;
case Mixed:
aminoColor = config.getImageMixed();
break;
case ConnectionVertical:
g2.setColor(config.getImageConnection());
g2.fillRect((xIndex * pixelsPerCell) + margin + (pixelsPerCell/2)-outline,
case ConnectionVertical:
g2.setColor(config.getImageConnection());
g2.fillRect(
(xIndex * pixelsPerCell) + margin + (pixelsPerCell / 2) - outline,
(yIndexPosition * pixelsPerCell) + margin,
outline * 2, pixelsPerCell);
break;
break;
case ConnectionHorizontal:
g2.setColor(config.getImageConnection());
g2.fillRect((xIndex * pixelsPerCell) + margin,
(yIndexPosition * pixelsPerCell) + margin + (pixelsPerCell/2)-outline,
case ConnectionHorizontal:
g2.setColor(config.getImageConnection());
g2.fillRect((xIndex * pixelsPerCell) + margin,
(yIndexPosition * pixelsPerCell) + margin + (pixelsPerCell / 2)
- outline,
pixelsPerCell, outline * 2);
break;
}
break;
}
if (aminoColor != null) {
g2.setColor(config.getImageOutline());
g2.fillRect((xIndex * pixelsPerCell) + margin, (yIndexPosition * pixelsPerCell) + margin,
if (aminoColor != null) {
g2.setColor(config.getImageOutline());
g2.fillRect((xIndex * pixelsPerCell) + margin,
(yIndexPosition * pixelsPerCell) + margin,
pixelsPerCell, pixelsPerCell);
g2.setColor(aminoColor);
g2.fillRect((xIndex * pixelsPerCell) + outline + margin, (yIndexPosition * pixelsPerCell) + outline + margin,
g2.setColor(aminoColor);
g2.fillRect((xIndex * pixelsPerCell) + outline + margin,
(yIndexPosition * pixelsPerCell) + outline + margin,
pixelsPerCell - (outline * 2), pixelsPerCell - (outline * 2));
g2.setColor(config.getImageAminoText());
String label = "";
for (int aminoIndex : cellArray[yIndex][xIndex].aminoIndexes) {
label += aminoIndex + " ";
g2.setColor(config.getImageAminoText());
String label = "";
for (int aminoIndex : cellArray[yIndex][xIndex].aminoIndexes) {
label += aminoIndex + " ";
}
int labelWidth = metrics.stringWidth(label);
int x =
(xIndex * pixelsPerCell) + margin + (pixelsPerCell / 2) - (labelWidth
/ 2) + outline;
int y = (yIndexPosition * pixelsPerCell) + margin + (pixelsPerCell / 2) + (
ascent / 2) - outline;
g2.drawString(label, x, y);
}
int labelWidth = metrics.stringWidth(label);
int x = (xIndex * pixelsPerCell) + margin + (pixelsPerCell/2) - (labelWidth/2) + outline;
int y = (yIndexPosition * pixelsPerCell) + margin + (pixelsPerCell/2) + (ascent/2) - outline;
g2.drawString(label, x, y);
}
}
}
g2.setColor(config.getImageText());
//TODO Get the labels from the new Generation class?
EvaluatorNESW evaluator = new EvaluatorNESW(1,isHydrophobic);
int bonds = evaluator.evaluateBonds(bestCandidateOfGeneration);
int overlaps = evaluator.evaluateOverlaps(bestCandidateOfGeneration);
String label = "Gen: " + geneticAlgorithm.generation
g2.setColor(config.getImageText());
//TODO Get the labels from the new Generation class?
EvaluatorNESW evaluator = new EvaluatorNESW(1, isHydrophobic);
int bonds = evaluator.evaluateBonds(bestCandidateOfGeneration);
int overlaps = evaluator.evaluateOverlaps(bestCandidateOfGeneration);
String label = "Gen: " + geneticAlgorithm.generation
+ " Fitness: " + String.format("%.4f", bestCandidateOfGeneration.getFitness())
+ " H/H Bonds: " + bonds
+ " Overlaps: " + overlaps;
int labelWidth = metrics.stringWidth(label);
int x = margin / 4;
int y = margin / 4;
g2.drawString(label, x, y);
try {
//ImageIO.write(image, "png", new File(folder + File.separator + filename));
ImageIO.write(image, "png", new File(filename));
} catch (IOException e) {
e.printStackTrace();
System.exit(0);
int labelWidth = metrics.stringWidth(label);
int x = margin / 4;
int y = margin / 4;
g2.drawString(label, x, y);
try {
//ImageIO.write(image, "png", new File(folder + File.separator + filename));
ImageIO.write(image, "png", new File(filename));
} catch (IOException e) {
e.printStackTrace();
System.exit(0);
}
}
}
......
src/main/resources/genetic.properties
+ 13
− 11
View file @ 9c93a693
......@@ -7,13 +7,13 @@
### Algorithm settings
# Size of the population in one generation
populationSize = 100
populationSize = 1000
# Number of generations to run the algorithm for
noGenerations = 20
noGenerations = 10000
# Type of population initialization [curl | straight | random]
initializationMethod = random
# Type of selection that should be used [proportional | tournament | onlybest]
selectionMethod = proportional
selectionMethod = tournament
# Number of tournament participants, only relevant when selection is set to tournament
k = 5
# Type(s) of mutators to use in the algorithm, separated by comma [singlePoint / crossover / singlePointGlobal]
......@@ -24,21 +24,21 @@
### Mutation settings
# How often a single point mutation should be attempted per candidate
mutationAttemptsPerCandidate = 1
mutationAttemptsPerCandidate = 100
# Chance for a successful single point mutation [1.0 = 100%]
mutationChance = 1.0
mutationChance = 0.011
# The mutation chance cannot drop below this value [1.0 = 100%]
mutationMinimalChance = 0.001
mutationMinimalChance = 0.01
# Multiplicand for mutation probability with each generation -> ex with 0.05: 1st 1.0, 2nd 0.95, 3rd 0.9025, 4th 0.857
mutationMultiplier = 0.001
mutationMultiplier = 0.0
# How often a crossover should be attempted per candidate
crossoverAttemptsPerCandidate = 1
# Chance for a successful crossover [1.0 = 100%]
crossoverChance = 0.6
crossoverChance = 0.051
# The crossover chance cannot drop below this value [1.0 = 100%]
crossoverMinimalChance = 0.001
crossoverMinimalChance = 0.05
# Multiplicand for mutation probability with each generation -> ex with -0.05: 1st 1.0, 2nd 1.05, 3rd 1,1025, 4th 1,157
crossoverMultiplier = 0.005
crossoverMultiplier = 0.0
### Output settings
......@@ -48,7 +48,7 @@
# video = Generate a video from the created images. Requires image
# log = Generate tab seperated file with a bit of information about each generation
# generation = print a short overview about each generation to stdout
visualizerType = log,generation,image,video
visualizerType = generation,image
# Name of the job, filenames will be based on this name. If left empty a timestamp is used
jobName =
# Directory for the image sequence
......@@ -57,6 +57,8 @@
videoDirectory = visualization
# Directory for the log file, always active
logfileDirectory = log
# Interval of generations between rendered images
imageInterval = 100
# Frames per second in the beginning of the video
imgFps = 10
# Number of images until fps increases by one
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment