[6256] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[14185] | 3 | * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[6256] | 4 | *
|
---|
| 5 | * This file is part of HeuristicLab.
|
---|
| 6 | *
|
---|
| 7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
| 8 | * it under the terms of the GNU General Public License as published by
|
---|
| 9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
| 10 | * (at your option) any later version.
|
---|
| 11 | *
|
---|
| 12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 15 | * GNU General Public License for more details.
|
---|
| 16 | *
|
---|
| 17 | * You should have received a copy of the GNU General Public License
|
---|
| 18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
| 19 | */
|
---|
| 20 | #endregion
|
---|
| 21 |
|
---|
[8704] | 22 | using System;
|
---|
[6256] | 23 | using System.Collections.Generic;
|
---|
| 24 | using System.Linq;
|
---|
[8704] | 25 | using AutoDiff;
|
---|
[6256] | 26 | using HeuristicLab.Common;
|
---|
| 27 | using HeuristicLab.Core;
|
---|
| 28 | using HeuristicLab.Data;
|
---|
| 29 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
| 30 | using HeuristicLab.Parameters;
|
---|
| 31 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 32 |
|
---|
| 33 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
|
---|
[6555] | 34 | [Item("Constant Optimization Evaluator", "Calculates Pearson R² of a symbolic regression solution and optimizes the constant used.")]
|
---|
[6256] | 35 | [StorableClass]
|
---|
| 36 | public class SymbolicRegressionConstantOptimizationEvaluator : SymbolicRegressionSingleObjectiveEvaluator {
|
---|
| 37 | private const string ConstantOptimizationIterationsParameterName = "ConstantOptimizationIterations";
|
---|
| 38 | private const string ConstantOptimizationImprovementParameterName = "ConstantOptimizationImprovement";
|
---|
| 39 | private const string ConstantOptimizationProbabilityParameterName = "ConstantOptimizationProbability";
|
---|
| 40 | private const string ConstantOptimizationRowsPercentageParameterName = "ConstantOptimizationRowsPercentage";
|
---|
[8823] | 41 | private const string UpdateConstantsInTreeParameterName = "UpdateConstantsInSymbolicExpressionTree";
|
---|
[13670] | 42 | private const string UpdateVariableWeightsParameterName = "Update Variable Weights";
|
---|
[6256] | 43 |
|
---|
| 44 | public IFixedValueParameter<IntValue> ConstantOptimizationIterationsParameter {
|
---|
| 45 | get { return (IFixedValueParameter<IntValue>)Parameters[ConstantOptimizationIterationsParameterName]; }
|
---|
| 46 | }
|
---|
| 47 | public IFixedValueParameter<DoubleValue> ConstantOptimizationImprovementParameter {
|
---|
| 48 | get { return (IFixedValueParameter<DoubleValue>)Parameters[ConstantOptimizationImprovementParameterName]; }
|
---|
| 49 | }
|
---|
| 50 | public IFixedValueParameter<PercentValue> ConstantOptimizationProbabilityParameter {
|
---|
| 51 | get { return (IFixedValueParameter<PercentValue>)Parameters[ConstantOptimizationProbabilityParameterName]; }
|
---|
| 52 | }
|
---|
| 53 | public IFixedValueParameter<PercentValue> ConstantOptimizationRowsPercentageParameter {
|
---|
| 54 | get { return (IFixedValueParameter<PercentValue>)Parameters[ConstantOptimizationRowsPercentageParameterName]; }
|
---|
| 55 | }
|
---|
[8823] | 56 | public IFixedValueParameter<BoolValue> UpdateConstantsInTreeParameter {
|
---|
| 57 | get { return (IFixedValueParameter<BoolValue>)Parameters[UpdateConstantsInTreeParameterName]; }
|
---|
| 58 | }
|
---|
[13670] | 59 | public IFixedValueParameter<BoolValue> UpdateVariableWeightsParameter {
|
---|
| 60 | get { return (IFixedValueParameter<BoolValue>)Parameters[UpdateVariableWeightsParameterName]; }
|
---|
| 61 | }
|
---|
[6256] | 62 |
|
---|
[13670] | 63 |
|
---|
[6256] | 64 | public IntValue ConstantOptimizationIterations {
|
---|
| 65 | get { return ConstantOptimizationIterationsParameter.Value; }
|
---|
| 66 | }
|
---|
| 67 | public DoubleValue ConstantOptimizationImprovement {
|
---|
| 68 | get { return ConstantOptimizationImprovementParameter.Value; }
|
---|
| 69 | }
|
---|
| 70 | public PercentValue ConstantOptimizationProbability {
|
---|
| 71 | get { return ConstantOptimizationProbabilityParameter.Value; }
|
---|
| 72 | }
|
---|
| 73 | public PercentValue ConstantOptimizationRowsPercentage {
|
---|
| 74 | get { return ConstantOptimizationRowsPercentageParameter.Value; }
|
---|
| 75 | }
|
---|
[8823] | 76 | public bool UpdateConstantsInTree {
|
---|
| 77 | get { return UpdateConstantsInTreeParameter.Value.Value; }
|
---|
| 78 | set { UpdateConstantsInTreeParameter.Value.Value = value; }
|
---|
| 79 | }
|
---|
[6256] | 80 |
|
---|
[13670] | 81 | public bool UpdateVariableWeights {
|
---|
| 82 | get { return UpdateVariableWeightsParameter.Value.Value; }
|
---|
| 83 | set { UpdateVariableWeightsParameter.Value.Value = value; }
|
---|
| 84 | }
|
---|
| 85 |
|
---|
[6256] | 86 | public override bool Maximization {
|
---|
| 87 | get { return true; }
|
---|
| 88 | }
|
---|
| 89 |
|
---|
| 90 | [StorableConstructor]
|
---|
| 91 | protected SymbolicRegressionConstantOptimizationEvaluator(bool deserializing) : base(deserializing) { }
|
---|
| 92 | protected SymbolicRegressionConstantOptimizationEvaluator(SymbolicRegressionConstantOptimizationEvaluator original, Cloner cloner)
|
---|
| 93 | : base(original, cloner) {
|
---|
| 94 | }
|
---|
| 95 | public SymbolicRegressionConstantOptimizationEvaluator()
|
---|
| 96 | : base() {
|
---|
[8938] | 97 | Parameters.Add(new FixedValueParameter<IntValue>(ConstantOptimizationIterationsParameterName, "Determines how many iterations should be calculated while optimizing the constant of a symbolic expression tree (0 indicates other or default stopping criterion).", new IntValue(10), true));
|
---|
[13916] | 98 | Parameters.Add(new FixedValueParameter<DoubleValue>(ConstantOptimizationImprovementParameterName, "Determines the relative improvement which must be achieved in the constant optimization to continue with it (0 indicates other or default stopping criterion).", new DoubleValue(0), true) { Hidden = true });
|
---|
[6256] | 99 | Parameters.Add(new FixedValueParameter<PercentValue>(ConstantOptimizationProbabilityParameterName, "Determines the probability that the constants are optimized", new PercentValue(1), true));
|
---|
| 100 | Parameters.Add(new FixedValueParameter<PercentValue>(ConstantOptimizationRowsPercentageParameterName, "Determines the percentage of the rows which should be used for constant optimization", new PercentValue(1), true));
|
---|
[13916] | 101 | Parameters.Add(new FixedValueParameter<BoolValue>(UpdateConstantsInTreeParameterName, "Determines if the constants in the tree should be overwritten by the optimized constants.", new BoolValue(true)) { Hidden = true });
|
---|
| 102 | Parameters.Add(new FixedValueParameter<BoolValue>(UpdateVariableWeightsParameterName, "Determines if the variable weights in the tree should be optimized.", new BoolValue(true)) { Hidden = true });
|
---|
[6256] | 103 | }
|
---|
| 104 |
|
---|
| 105 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 106 | return new SymbolicRegressionConstantOptimizationEvaluator(this, cloner);
|
---|
| 107 | }
|
---|
| 108 |
|
---|
[8823] | 109 | [StorableHook(HookType.AfterDeserialization)]
|
---|
| 110 | private void AfterDeserialization() {
|
---|
| 111 | if (!Parameters.ContainsKey(UpdateConstantsInTreeParameterName))
|
---|
| 112 | Parameters.Add(new FixedValueParameter<BoolValue>(UpdateConstantsInTreeParameterName, "Determines if the constants in the tree should be overwritten by the optimized constants.", new BoolValue(true)));
|
---|
[13670] | 113 | if (!Parameters.ContainsKey(UpdateVariableWeightsParameterName))
|
---|
| 114 | Parameters.Add(new FixedValueParameter<BoolValue>(UpdateVariableWeightsParameterName, "Determines if the variable weights in the tree should be optimized.", new BoolValue(true)));
|
---|
[8823] | 115 | }
|
---|
| 116 |
|
---|
[10291] | 117 | public override IOperation InstrumentedApply() {
|
---|
[6256] | 118 | var solution = SymbolicExpressionTreeParameter.ActualValue;
|
---|
| 119 | double quality;
|
---|
| 120 | if (RandomParameter.ActualValue.NextDouble() < ConstantOptimizationProbability.Value) {
|
---|
| 121 | IEnumerable<int> constantOptimizationRows = GenerateRowsToEvaluate(ConstantOptimizationRowsPercentage.Value);
|
---|
| 122 | quality = OptimizeConstants(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, solution, ProblemDataParameter.ActualValue,
|
---|
[13670] | 123 | constantOptimizationRows, ApplyLinearScalingParameter.ActualValue.Value, ConstantOptimizationIterations.Value, updateVariableWeights: UpdateVariableWeights, lowerEstimationLimit: EstimationLimitsParameter.ActualValue.Lower, upperEstimationLimit: EstimationLimitsParameter.ActualValue.Upper, updateConstantsInTree: UpdateConstantsInTree);
|
---|
[8938] | 124 |
|
---|
[6256] | 125 | if (ConstantOptimizationRowsPercentage.Value != RelativeNumberOfEvaluatedSamplesParameter.ActualValue.Value) {
|
---|
| 126 | var evaluationRows = GenerateRowsToEvaluate();
|
---|
[8664] | 127 | quality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, solution, EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper, ProblemDataParameter.ActualValue, evaluationRows, ApplyLinearScalingParameter.ActualValue.Value);
|
---|
[6256] | 128 | }
|
---|
| 129 | } else {
|
---|
| 130 | var evaluationRows = GenerateRowsToEvaluate();
|
---|
[8664] | 131 | quality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, solution, EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper, ProblemDataParameter.ActualValue, evaluationRows, ApplyLinearScalingParameter.ActualValue.Value);
|
---|
[6256] | 132 | }
|
---|
| 133 | QualityParameter.ActualValue = new DoubleValue(quality);
|
---|
| 134 |
|
---|
[10291] | 135 | return base.InstrumentedApply();
|
---|
[6256] | 136 | }
|
---|
| 137 |
|
---|
| 138 | public override double Evaluate(IExecutionContext context, ISymbolicExpressionTree tree, IRegressionProblemData problemData, IEnumerable<int> rows) {
|
---|
| 139 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = context;
|
---|
| 140 | EstimationLimitsParameter.ExecutionContext = context;
|
---|
[8664] | 141 | ApplyLinearScalingParameter.ExecutionContext = context;
|
---|
[6256] | 142 |
|
---|
[9209] | 143 | // Pearson R² evaluator is used on purpose instead of the const-opt evaluator,
|
---|
| 144 | // because Evaluate() is used to get the quality of evolved models on
|
---|
| 145 | // different partitions of the dataset (e.g., best validation model)
|
---|
[8664] | 146 | double r2 = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, tree, EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper, problemData, rows, ApplyLinearScalingParameter.ActualValue.Value);
|
---|
[6256] | 147 |
|
---|
| 148 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = null;
|
---|
| 149 | EstimationLimitsParameter.ExecutionContext = null;
|
---|
[9209] | 150 | ApplyLinearScalingParameter.ExecutionContext = null;
|
---|
[6256] | 151 |
|
---|
| 152 | return r2;
|
---|
| 153 | }
|
---|
| 154 |
|
---|
[8823] | 155 | #region derivations of functions
|
---|
[8730] | 156 | // create function factory for arctangent
|
---|
| 157 | private readonly Func<Term, UnaryFunc> arctan = UnaryFunc.Factory(
|
---|
[8823] | 158 | eval: Math.Atan,
|
---|
| 159 | diff: x => 1 / (1 + x * x));
|
---|
[8730] | 160 | private static readonly Func<Term, UnaryFunc> sin = UnaryFunc.Factory(
|
---|
[8823] | 161 | eval: Math.Sin,
|
---|
| 162 | diff: Math.Cos);
|
---|
[8730] | 163 | private static readonly Func<Term, UnaryFunc> cos = UnaryFunc.Factory(
|
---|
[8823] | 164 | eval: Math.Cos,
|
---|
| 165 | diff: x => -Math.Sin(x));
|
---|
[8730] | 166 | private static readonly Func<Term, UnaryFunc> tan = UnaryFunc.Factory(
|
---|
[8823] | 167 | eval: Math.Tan,
|
---|
| 168 | diff: x => 1 + Math.Tan(x) * Math.Tan(x));
|
---|
[8730] | 169 | private static readonly Func<Term, UnaryFunc> erf = UnaryFunc.Factory(
|
---|
[8823] | 170 | eval: alglib.errorfunction,
|
---|
| 171 | diff: x => 2.0 * Math.Exp(-(x * x)) / Math.Sqrt(Math.PI));
|
---|
[8730] | 172 | private static readonly Func<Term, UnaryFunc> norm = UnaryFunc.Factory(
|
---|
[8823] | 173 | eval: alglib.normaldistribution,
|
---|
| 174 | diff: x => -(Math.Exp(-(x * x)) * Math.Sqrt(Math.Exp(x * x)) * x) / Math.Sqrt(2 * Math.PI));
|
---|
| 175 | #endregion
|
---|
[8730] | 176 |
|
---|
| 177 |
|
---|
[13670] | 178 | public static double OptimizeConstants(ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, ISymbolicExpressionTree tree, IRegressionProblemData problemData, IEnumerable<int> rows, bool applyLinearScaling, int maxIterations, bool updateVariableWeights = true, double lowerEstimationLimit = double.MinValue, double upperEstimationLimit = double.MaxValue, bool updateConstantsInTree = true) {
|
---|
[8704] | 179 |
|
---|
| 180 | List<AutoDiff.Variable> variables = new List<AutoDiff.Variable>();
|
---|
| 181 | List<AutoDiff.Variable> parameters = new List<AutoDiff.Variable>();
|
---|
| 182 | List<string> variableNames = new List<string>();
|
---|
[14349] | 183 | List<int> lags = new List<int>();
|
---|
[8704] | 184 |
|
---|
| 185 | AutoDiff.Term func;
|
---|
[14349] | 186 | if (!TryTransformToAutoDiff(tree.Root.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out func))
|
---|
[8828] | 187 | throw new NotSupportedException("Could not optimize constants of symbolic expression tree due to not supported symbols used in the tree.");
|
---|
[8704] | 188 | if (variableNames.Count == 0) return 0.0;
|
---|
| 189 |
|
---|
[13670] | 190 | AutoDiff.IParametricCompiledTerm compiledFunc = func.Compile(variables.ToArray(), parameters.ToArray());
|
---|
[8704] | 191 |
|
---|
[13670] | 192 | List<SymbolicExpressionTreeTerminalNode> terminalNodes = null;
|
---|
| 193 | if (updateVariableWeights)
|
---|
| 194 | terminalNodes = tree.Root.IterateNodesPrefix().OfType<SymbolicExpressionTreeTerminalNode>().ToList();
|
---|
| 195 | else
|
---|
| 196 | terminalNodes = new List<SymbolicExpressionTreeTerminalNode>(tree.Root.IterateNodesPrefix().OfType<ConstantTreeNode>());
|
---|
| 197 |
|
---|
| 198 | //extract inital constants
|
---|
[8704] | 199 | double[] c = new double[variables.Count];
|
---|
| 200 | {
|
---|
| 201 | c[0] = 0.0;
|
---|
| 202 | c[1] = 1.0;
|
---|
| 203 | int i = 2;
|
---|
| 204 | foreach (var node in terminalNodes) {
|
---|
| 205 | ConstantTreeNode constantTreeNode = node as ConstantTreeNode;
|
---|
| 206 | VariableTreeNode variableTreeNode = node as VariableTreeNode;
|
---|
| 207 | if (constantTreeNode != null)
|
---|
| 208 | c[i++] = constantTreeNode.Value;
|
---|
[13670] | 209 | else if (updateVariableWeights && variableTreeNode != null)
|
---|
[8704] | 210 | c[i++] = variableTreeNode.Weight;
|
---|
| 211 | }
|
---|
[6256] | 212 | }
|
---|
[8938] | 213 | double[] originalConstants = (double[])c.Clone();
|
---|
| 214 | double originalQuality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(interpreter, tree, lowerEstimationLimit, upperEstimationLimit, problemData, rows, applyLinearScaling);
|
---|
[6256] | 215 |
|
---|
[8704] | 216 | alglib.lsfitstate state;
|
---|
| 217 | alglib.lsfitreport rep;
|
---|
| 218 | int info;
|
---|
[6256] | 219 |
|
---|
[12509] | 220 | IDataset ds = problemData.Dataset;
|
---|
[8704] | 221 | double[,] x = new double[rows.Count(), variableNames.Count];
|
---|
| 222 | int row = 0;
|
---|
| 223 | foreach (var r in rows) {
|
---|
| 224 | for (int col = 0; col < variableNames.Count; col++) {
|
---|
[14349] | 225 | int lag = lags[col];
|
---|
| 226 | x[row, col] = ds.GetDoubleValue(variableNames[col], r + lag);
|
---|
[8704] | 227 | }
|
---|
| 228 | row++;
|
---|
| 229 | }
|
---|
| 230 | double[] y = ds.GetDoubleValues(problemData.TargetVariable, rows).ToArray();
|
---|
| 231 | int n = x.GetLength(0);
|
---|
| 232 | int m = x.GetLength(1);
|
---|
| 233 | int k = c.Length;
|
---|
[6256] | 234 |
|
---|
[8704] | 235 | alglib.ndimensional_pfunc function_cx_1_func = CreatePFunc(compiledFunc);
|
---|
| 236 | alglib.ndimensional_pgrad function_cx_1_grad = CreatePGrad(compiledFunc);
|
---|
[6256] | 237 |
|
---|
[8704] | 238 | try {
|
---|
| 239 | alglib.lsfitcreatefg(x, y, c, n, m, k, false, out state);
|
---|
[8938] | 240 | alglib.lsfitsetcond(state, 0.0, 0.0, maxIterations);
|
---|
| 241 | //alglib.lsfitsetgradientcheck(state, 0.001);
|
---|
[8704] | 242 | alglib.lsfitfit(state, function_cx_1_func, function_cx_1_grad, null, null);
|
---|
| 243 | alglib.lsfitresults(state, out info, out c, out rep);
|
---|
[14349] | 244 | } catch (ArithmeticException) {
|
---|
[8984] | 245 | return originalQuality;
|
---|
[14349] | 246 | } catch (alglib.alglibexception) {
|
---|
[8984] | 247 | return originalQuality;
|
---|
[8704] | 248 | }
|
---|
[8823] | 249 |
|
---|
[8938] | 250 | //info == -7 => constant optimization failed due to wrong gradient
|
---|
[13670] | 251 | if (info != -7) UpdateConstants(tree, c.Skip(2).ToArray(), updateVariableWeights);
|
---|
[8938] | 252 | var quality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(interpreter, tree, lowerEstimationLimit, upperEstimationLimit, problemData, rows, applyLinearScaling);
|
---|
| 253 |
|
---|
[13670] | 254 | if (!updateConstantsInTree) UpdateConstants(tree, originalConstants.Skip(2).ToArray(), updateVariableWeights);
|
---|
[8938] | 255 | if (originalQuality - quality > 0.001 || double.IsNaN(quality)) {
|
---|
[13670] | 256 | UpdateConstants(tree, originalConstants.Skip(2).ToArray(), updateVariableWeights);
|
---|
[8938] | 257 | return originalQuality;
|
---|
[8704] | 258 | }
|
---|
[8938] | 259 | return quality;
|
---|
[6256] | 260 | }
|
---|
| 261 |
|
---|
[13670] | 262 | private static void UpdateConstants(ISymbolicExpressionTree tree, double[] constants, bool updateVariableWeights) {
|
---|
[8938] | 263 | int i = 0;
|
---|
| 264 | foreach (var node in tree.Root.IterateNodesPrefix().OfType<SymbolicExpressionTreeTerminalNode>()) {
|
---|
| 265 | ConstantTreeNode constantTreeNode = node as ConstantTreeNode;
|
---|
| 266 | VariableTreeNode variableTreeNode = node as VariableTreeNode;
|
---|
| 267 | if (constantTreeNode != null)
|
---|
| 268 | constantTreeNode.Value = constants[i++];
|
---|
[13670] | 269 | else if (updateVariableWeights && variableTreeNode != null)
|
---|
[8938] | 270 | variableTreeNode.Weight = constants[i++];
|
---|
| 271 | }
|
---|
| 272 | }
|
---|
| 273 |
|
---|
[8704] | 274 | private static alglib.ndimensional_pfunc CreatePFunc(AutoDiff.IParametricCompiledTerm compiledFunc) {
|
---|
| 275 | return (double[] c, double[] x, ref double func, object o) => {
|
---|
| 276 | func = compiledFunc.Evaluate(c, x);
|
---|
| 277 | };
|
---|
| 278 | }
|
---|
[6256] | 279 |
|
---|
[8704] | 280 | private static alglib.ndimensional_pgrad CreatePGrad(AutoDiff.IParametricCompiledTerm compiledFunc) {
|
---|
| 281 | return (double[] c, double[] x, ref double func, double[] grad, object o) => {
|
---|
| 282 | var tupel = compiledFunc.Differentiate(c, x);
|
---|
| 283 | func = tupel.Item2;
|
---|
| 284 | Array.Copy(tupel.Item1, grad, grad.Length);
|
---|
[6256] | 285 | };
|
---|
| 286 | }
|
---|
| 287 |
|
---|
[14349] | 288 | private static bool TryTransformToAutoDiff(ISymbolicExpressionTreeNode node, List<AutoDiff.Variable> variables, List<AutoDiff.Variable> parameters, List<string> variableNames, List<int> lags, bool updateVariableWeights, out AutoDiff.Term term) {
|
---|
[8704] | 289 | if (node.Symbol is Constant) {
|
---|
| 290 | var var = new AutoDiff.Variable();
|
---|
| 291 | variables.Add(var);
|
---|
| 292 | term = var;
|
---|
| 293 | return true;
|
---|
| 294 | }
|
---|
| 295 | if (node.Symbol is Variable) {
|
---|
| 296 | var varNode = node as VariableTreeNode;
|
---|
| 297 | var par = new AutoDiff.Variable();
|
---|
| 298 | parameters.Add(par);
|
---|
| 299 | variableNames.Add(varNode.VariableName);
|
---|
[14349] | 300 | lags.Add(0);
|
---|
[13670] | 301 |
|
---|
| 302 | if (updateVariableWeights) {
|
---|
| 303 | var w = new AutoDiff.Variable();
|
---|
| 304 | variables.Add(w);
|
---|
| 305 | term = AutoDiff.TermBuilder.Product(w, par);
|
---|
| 306 | } else {
|
---|
| 307 | term = par;
|
---|
| 308 | }
|
---|
[8704] | 309 | return true;
|
---|
| 310 | }
|
---|
[14349] | 311 | if (node.Symbol is LaggedVariable) {
|
---|
| 312 | var varNode = node as LaggedVariableTreeNode;
|
---|
| 313 | var par = new AutoDiff.Variable();
|
---|
| 314 | parameters.Add(par);
|
---|
| 315 | variableNames.Add(varNode.VariableName);
|
---|
| 316 | lags.Add(varNode.Lag);
|
---|
| 317 |
|
---|
| 318 | if (updateVariableWeights) {
|
---|
| 319 | var w = new AutoDiff.Variable();
|
---|
| 320 | variables.Add(w);
|
---|
| 321 | term = AutoDiff.TermBuilder.Product(w, par);
|
---|
| 322 | } else {
|
---|
| 323 | term = par;
|
---|
| 324 | }
|
---|
| 325 | return true;
|
---|
| 326 | }
|
---|
[8704] | 327 | if (node.Symbol is Addition) {
|
---|
| 328 | List<AutoDiff.Term> terms = new List<Term>();
|
---|
| 329 | foreach (var subTree in node.Subtrees) {
|
---|
| 330 | AutoDiff.Term t;
|
---|
[14349] | 331 | if (!TryTransformToAutoDiff(subTree, variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8704] | 332 | term = null;
|
---|
| 333 | return false;
|
---|
| 334 | }
|
---|
| 335 | terms.Add(t);
|
---|
| 336 | }
|
---|
| 337 | term = AutoDiff.TermBuilder.Sum(terms);
|
---|
| 338 | return true;
|
---|
| 339 | }
|
---|
[8823] | 340 | if (node.Symbol is Subtraction) {
|
---|
| 341 | List<AutoDiff.Term> terms = new List<Term>();
|
---|
| 342 | for (int i = 0; i < node.SubtreeCount; i++) {
|
---|
| 343 | AutoDiff.Term t;
|
---|
[14349] | 344 | if (!TryTransformToAutoDiff(node.GetSubtree(i), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8823] | 345 | term = null;
|
---|
| 346 | return false;
|
---|
| 347 | }
|
---|
| 348 | if (i > 0) t = -t;
|
---|
| 349 | terms.Add(t);
|
---|
| 350 | }
|
---|
[14036] | 351 | if (terms.Count == 1) term = -terms[0];
|
---|
| 352 | else term = AutoDiff.TermBuilder.Sum(terms);
|
---|
[8823] | 353 | return true;
|
---|
| 354 | }
|
---|
[8704] | 355 | if (node.Symbol is Multiplication) {
|
---|
[13900] | 356 | List<AutoDiff.Term> terms = new List<Term>();
|
---|
| 357 | foreach (var subTree in node.Subtrees) {
|
---|
| 358 | AutoDiff.Term t;
|
---|
[14349] | 359 | if (!TryTransformToAutoDiff(subTree, variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[13869] | 360 | term = null;
|
---|
| 361 | return false;
|
---|
| 362 | }
|
---|
[13900] | 363 | terms.Add(t);
|
---|
[8704] | 364 | }
|
---|
[13900] | 365 | if (terms.Count == 1) term = terms[0];
|
---|
| 366 | else term = terms.Aggregate((a, b) => new AutoDiff.Product(a, b));
|
---|
[13869] | 367 | return true;
|
---|
[13900] | 368 |
|
---|
[8704] | 369 | }
|
---|
| 370 | if (node.Symbol is Division) {
|
---|
[13900] | 371 | List<AutoDiff.Term> terms = new List<Term>();
|
---|
| 372 | foreach (var subTree in node.Subtrees) {
|
---|
| 373 | AutoDiff.Term t;
|
---|
[14349] | 374 | if (!TryTransformToAutoDiff(subTree, variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[13869] | 375 | term = null;
|
---|
| 376 | return false;
|
---|
| 377 | }
|
---|
[13900] | 378 | terms.Add(t);
|
---|
[8704] | 379 | }
|
---|
[13900] | 380 | if (terms.Count == 1) term = 1.0 / terms[0];
|
---|
| 381 | else term = terms.Aggregate((a, b) => new AutoDiff.Product(a, 1.0 / b));
|
---|
[13869] | 382 | return true;
|
---|
[8704] | 383 | }
|
---|
| 384 | if (node.Symbol is Logarithm) {
|
---|
| 385 | AutoDiff.Term t;
|
---|
[14349] | 386 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8704] | 387 | term = null;
|
---|
| 388 | return false;
|
---|
| 389 | } else {
|
---|
| 390 | term = AutoDiff.TermBuilder.Log(t);
|
---|
| 391 | return true;
|
---|
| 392 | }
|
---|
| 393 | }
|
---|
| 394 | if (node.Symbol is Exponential) {
|
---|
| 395 | AutoDiff.Term t;
|
---|
[14349] | 396 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8704] | 397 | term = null;
|
---|
| 398 | return false;
|
---|
| 399 | } else {
|
---|
| 400 | term = AutoDiff.TermBuilder.Exp(t);
|
---|
| 401 | return true;
|
---|
| 402 | }
|
---|
[11680] | 403 | }
|
---|
| 404 | if (node.Symbol is Square) {
|
---|
[8730] | 405 | AutoDiff.Term t;
|
---|
[14349] | 406 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 407 | term = null;
|
---|
| 408 | return false;
|
---|
| 409 | } else {
|
---|
[11680] | 410 | term = AutoDiff.TermBuilder.Power(t, 2.0);
|
---|
[8730] | 411 | return true;
|
---|
| 412 | }
|
---|
[13869] | 413 | }
|
---|
| 414 | if (node.Symbol is SquareRoot) {
|
---|
[8730] | 415 | AutoDiff.Term t;
|
---|
[14349] | 416 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 417 | term = null;
|
---|
| 418 | return false;
|
---|
| 419 | } else {
|
---|
[11680] | 420 | term = AutoDiff.TermBuilder.Power(t, 0.5);
|
---|
[8730] | 421 | return true;
|
---|
| 422 | }
|
---|
[13869] | 423 | }
|
---|
| 424 | if (node.Symbol is Sine) {
|
---|
[8730] | 425 | AutoDiff.Term t;
|
---|
[14349] | 426 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 427 | term = null;
|
---|
| 428 | return false;
|
---|
| 429 | } else {
|
---|
[11680] | 430 | term = sin(t);
|
---|
[8730] | 431 | return true;
|
---|
| 432 | }
|
---|
[13869] | 433 | }
|
---|
| 434 | if (node.Symbol is Cosine) {
|
---|
[8730] | 435 | AutoDiff.Term t;
|
---|
[14349] | 436 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 437 | term = null;
|
---|
| 438 | return false;
|
---|
| 439 | } else {
|
---|
[11680] | 440 | term = cos(t);
|
---|
[8730] | 441 | return true;
|
---|
| 442 | }
|
---|
[13869] | 443 | }
|
---|
| 444 | if (node.Symbol is Tangent) {
|
---|
[11680] | 445 | AutoDiff.Term t;
|
---|
[14349] | 446 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[11680] | 447 | term = null;
|
---|
| 448 | return false;
|
---|
| 449 | } else {
|
---|
| 450 | term = tan(t);
|
---|
| 451 | return true;
|
---|
| 452 | }
|
---|
[13869] | 453 | }
|
---|
| 454 | if (node.Symbol is Erf) {
|
---|
[8730] | 455 | AutoDiff.Term t;
|
---|
[14349] | 456 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 457 | term = null;
|
---|
| 458 | return false;
|
---|
| 459 | } else {
|
---|
| 460 | term = erf(t);
|
---|
| 461 | return true;
|
---|
| 462 | }
|
---|
[13869] | 463 | }
|
---|
| 464 | if (node.Symbol is Norm) {
|
---|
[8730] | 465 | AutoDiff.Term t;
|
---|
[14349] | 466 | if (!TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out t)) {
|
---|
[8730] | 467 | term = null;
|
---|
| 468 | return false;
|
---|
| 469 | } else {
|
---|
| 470 | term = norm(t);
|
---|
| 471 | return true;
|
---|
| 472 | }
|
---|
| 473 | }
|
---|
[8704] | 474 | if (node.Symbol is StartSymbol) {
|
---|
| 475 | var alpha = new AutoDiff.Variable();
|
---|
| 476 | var beta = new AutoDiff.Variable();
|
---|
| 477 | variables.Add(beta);
|
---|
| 478 | variables.Add(alpha);
|
---|
| 479 | AutoDiff.Term branchTerm;
|
---|
[14349] | 480 | if (TryTransformToAutoDiff(node.GetSubtree(0), variables, parameters, variableNames, lags, updateVariableWeights, out branchTerm)) {
|
---|
[8704] | 481 | term = branchTerm * alpha + beta;
|
---|
| 482 | return true;
|
---|
| 483 | } else {
|
---|
| 484 | term = null;
|
---|
| 485 | return false;
|
---|
| 486 | }
|
---|
| 487 | }
|
---|
| 488 | term = null;
|
---|
| 489 | return false;
|
---|
| 490 | }
|
---|
[8730] | 491 |
|
---|
| 492 | public static bool CanOptimizeConstants(ISymbolicExpressionTree tree) {
|
---|
| 493 | var containsUnknownSymbol = (
|
---|
| 494 | from n in tree.Root.GetSubtree(0).IterateNodesPrefix()
|
---|
| 495 | where
|
---|
| 496 | !(n.Symbol is Variable) &&
|
---|
[14349] | 497 | !(n.Symbol is LaggedVariable) &&
|
---|
[8730] | 498 | !(n.Symbol is Constant) &&
|
---|
| 499 | !(n.Symbol is Addition) &&
|
---|
| 500 | !(n.Symbol is Subtraction) &&
|
---|
| 501 | !(n.Symbol is Multiplication) &&
|
---|
| 502 | !(n.Symbol is Division) &&
|
---|
| 503 | !(n.Symbol is Logarithm) &&
|
---|
| 504 | !(n.Symbol is Exponential) &&
|
---|
[11680] | 505 | !(n.Symbol is SquareRoot) &&
|
---|
| 506 | !(n.Symbol is Square) &&
|
---|
[8730] | 507 | !(n.Symbol is Sine) &&
|
---|
| 508 | !(n.Symbol is Cosine) &&
|
---|
| 509 | !(n.Symbol is Tangent) &&
|
---|
| 510 | !(n.Symbol is Erf) &&
|
---|
| 511 | !(n.Symbol is Norm) &&
|
---|
| 512 | !(n.Symbol is StartSymbol)
|
---|
| 513 | select n).
|
---|
| 514 | Any();
|
---|
| 515 | return !containsUnknownSymbol;
|
---|
| 516 | }
|
---|
[6256] | 517 | }
|
---|
| 518 | }
|
---|