FightMicroGA/Simulation/Main.java at master · TheSimpleSoldier/FightMicroGA · GitHub

History

286 lines (237 loc) · 8.65 KB

Raw

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

package Simulation;

public class Main

{

/**

* This is the main function which triggers simulations

* @param args

public static void main(String[] args)

{

double mutationRate = 0.3;

double crossOverRate = 0.1;

double mutationAmount = 0.1;

boolean verbose = false;

double globalScale = 0.1;

double localScale = 0.2;

double randomScale = 0.3;

//runFightSimulation(inputs, inputs);

// System.out.println("Basic vs. Advanced");

// runFightSimulation(null, null, 1, 2, true, 2);

// System.out.println("Advanced vs. Basic");

// runFightSimulation(null, null, 2, 1, true, 2);

double[][] idealWeights = getIdealWeights(10, 1, mutationRate, crossOverRate, mutationAmount);

PSO pso = new PSO(globalScale, localScale, randomScale);

double[][] idealWeights2 = pso.getBestWeights(200, 10);

long startTime = System.currentTimeMillis();

System.out.println();

for (int i = 0; i < idealWeights2.length; i++)

{

for (int j = 0; j < idealWeights2[i].length; j++)

{

System.out.print(idealWeights2[i][j] + ", ");

}

System.out.println();

}

verbose = true;

for (int i = 0; i < 2; i++)

{

System.out.println("Net is Red:");

runFightSimulation(idealWeights, idealWeights2, 0, 0, verbose, 1);

System.out.println("Net is blue");

runFightSimulation(idealWeights2, idealWeights, 0, 0, verbose, 1);

System.out.println("Net vs. basic");

runFightSimulation(idealWeights, idealWeights, 0, 1, verbose, 1);

System.out.println("basic vs. Net");

runFightSimulation(idealWeights, idealWeights, 1, 0, verbose, 1);

System.out.println("PSO vs. basic");

runFightSimulation(idealWeights2, idealWeights2, 0, 1, verbose, 1);

System.out.println("basic vs. PSO");

runFightSimulation(idealWeights2, idealWeights2, 1, 0, verbose, 1);

}

System.out.println("Run Time: " + (System.currentTimeMillis() - startTime));

}

public static double[][] getIdealWeights(int popSize, int rounds, double mutationRate, double crossOverRate, double mutationAmount)

{

FeedForwardNeuralNetwork net = new FeedForwardNeuralNetwork(1, new int[]{6, 10, 4}, ActivationFunction.LOGISTIC, ActivationFunction.LOGISTIC);

double[][][] population = new double[popSize][][];

for (int i = 0; i < popSize; i++)

{

net.generateRandomWeights();

population[i] = new double[1][];

population[i][0] = net.getWeights();

// System.out.println("net length:" + net.getWeights().length);

}

for (int i = 0; i < rounds; i++)

{

population = runTheGA(population, mutationRate, crossOverRate, mutationAmount);

double[] totalFitness = new double[popSize];

for (int j = 0; j < popSize; j++)

{

for (int k = 0; k < popSize; k++)

{

if (j != k)

{

double[][] results = runFightSimulation(population[j], population[k], 0, 0, false, i);

totalFitness[j] += results[0][0];

totalFitness[k] += results[1][0];

// System.out.println("Score for Red: " + results[0][0]);

// System.out.println("Score for Blue: " + results[1][0]);

}

population = sortPopulation(population, totalFitness);

for (int j = 0; j < totalFitness.length; j++)

{

totalFitness[j] = 0;

}

System.out.println("Finished round: " + i + " of the GA");

}

return population[0];

}

/**

* This method sorts the population based on their fitness values

* @param initialPop

* @param fitness

* @return

public static double[][][] sortPopulation(double[][][] initialPop, double[] fitness)

{

double[][][] sortedPop = new double[initialPop.length][][];

for (int i = 0; i < sortedPop.length; i++)

{

sortedPop[i] = initialPop[i];

}

// bubble sort FTW!!!!

for (int i = 0; i < fitness.length; i++)

{

for (int j = i+1; j < fitness.length; j++)

{

if (fitness[j] > fitness[i])

{

double temp = fitness[j];

fitness[j] = fitness[i];

fitness[i] = temp;

double[][] temp2 = sortedPop[j];

sortedPop[j] = sortedPop[i];

sortedPop[i] = temp2;

}

return sortedPop;

}

/**

* This method takes an initial population and runs the GA to evolve a superior poputation

* Note: this method assumes that the initial population has been sorted from best to worst

* @param initialPop

* @param mutationRate

* @param crossOverRate

* @param mutationAmount

* @return

public static double[][][] runTheGA(double[][][] initialPop, double mutationRate, double crossOverRate, double mutationAmount)

{

double[][][] evolvedPop = new double[initialPop.length][][];

int currentIndex = 0;

int len = initialPop.length;

for (int i = 0; i < len; i++)

{

double[][] selected = null;

while (selected == null)

{

if (Math.random() < (((double) ((len + 1) - currentIndex) / (len + 2)) / 2))

{

selected = initialPop[currentIndex];

}

currentIndex = (currentIndex + 1) % len;

}

// Cross Over

if (Math.random() < crossOverRate)

{

double[][] selected2 = null;

while (selected2 == null)

{

if (Math.random() < (((double) ((len + 1) - currentIndex) / (len + 2)) / 2))

{

selected2 = initialPop[currentIndex];

}

currentIndex = (currentIndex + 1) % len;

}

for (int j = 0; j < selected.length; j++)

{

int crossOverPoint = (int) (Math.random() * selected[j].length);

for (int k = crossOverPoint; k < selected[j].length; k++)

{

selected[j] = selected2[j];

}

// Mutation

for (int j = 0; j < selected.length; j++)

{

for (int k = 0; k < selected[j].length; k++)

{

if (Math.random() < mutationRate)

{

selected[j][k] += (Math.random() * 2 * mutationAmount) - mutationAmount;

}

evolvedPop[i] = selected;

}

return evolvedPop;

}

/**

* This method returns an array with the total fitness values of all units

* for both teams in the form

* [

* Team 1: [

* Soldiers: [

* ]

* Tanks: [

* ]

* etc...

* ]

* Team 2: [

* ]

* @param team1Inputs

* @param team2Inputs

* @return

public static double[][] runFightSimulation(double[][] team1Inputs, double[][] team2Inputs, int teamA, int teamB, boolean verbose, int round)

{

if (verbose)

{

System.out.println("Simulating a match");

}

Game game = new Game(team1Inputs, team2Inputs, verbose);

String map = "FightMicroGA/Simulation/simulationMaps/onetower.xml";

if (round % 4 == 1)

{

map = "FightMicroGA/Simulation/simulationMaps/barren.xml";

}

else if (round % 4 == 2)

{

map = "FightMicroGA/Simulation/simulationMaps/frontlines.xml";

}

// else if (round % 4 == 3)

// {

// map = "FightMicroGA/Simulation/simulationMaps/noeffort.xml";

// }

game.runMatch(map, teamA, teamB);

double[][] results = new double[2][];

results[0] = game.getTeamResults(0);

results[1] = game.getTeamResults(1);

return results;

}

FightMicroGA/Simulation/Main.java at master · TheSimpleSoldier/FightMicroGA · GitHub

Expand file tree

Breadcrumbs

Main.java

Latest commit

History

Breadcrumbs

Main.java

File metadata and controls