Changeset 9325
- Timestamp:
- 03/20/13 21:02:10 (12 years ago)
- Location:
- branches/sluengo/HeuristicLab.Problems.TradeRules
- Files:
-
- 4 edited
-
Evaluator/EvaluatorTradeRules.cs (modified) (1 diff)
-
HeuristicLab.Problems.TradeRules-1.0.csproj (modified) (3 diffs)
-
Interpreter.cs (modified) (12 diffs)
-
Symbols/MACDTreeNode.cs (modified) (4 diffs)
Legend:
- Unmodified
- Added
- Removed
-
branches/sluengo/HeuristicLab.Problems.TradeRules/Evaluator/EvaluatorTradeRules.cs
r9262 r9325 56 56 57 57 double tradingCash = OnlineTradeRulesCalculator.Calculate(estimatedValues, problemData, problemData.TrainingIndices); 58 58 59 59 return tradingCash; 60 60 } -
branches/sluengo/HeuristicLab.Problems.TradeRules/HeuristicLab.Problems.TradeRules-1.0.csproj
r9262 r9325 32 32 </PropertyGroup> 33 33 <ItemGroup> 34 <Reference Include="HeuristicLab.Analysis-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">35 <SpecificVersion>False</SpecificVersion>36 <HintPath>..\bin\HeuristicLab.Analysis-3.3.dll</HintPath>37 </Reference>38 <Reference Include="HeuristicLab.Collections-3.3">39 <HintPath>..\bin\HeuristicLab.Collections-3.3.dll</HintPath>40 </Reference>41 <Reference Include="HeuristicLab.Common-3.3">42 <HintPath>..\bin\HeuristicLab.Common-3.3.dll</HintPath>43 </Reference>44 <Reference Include="HeuristicLab.Common.Resources-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">45 <SpecificVersion>False</SpecificVersion>46 <HintPath>..\bin\HeuristicLab.Common.Resources-3.3.dll</HintPath>47 </Reference>48 <Reference Include="HeuristicLab.Core-3.3">49 <HintPath>..\bin\HeuristicLab.Core-3.3.dll</HintPath>50 </Reference>51 <Reference Include="HeuristicLab.Data-3.3">52 <HintPath>..\bin\HeuristicLab.Data-3.3.dll</HintPath>53 </Reference>54 <Reference Include="HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4">55 <HintPath>..\bin\HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4.dll</HintPath>56 </Reference>57 <Reference Include="HeuristicLab.Operators-3.3">58 <HintPath>..\bin\HeuristicLab.Operators-3.3.dll</HintPath>59 </Reference>60 <Reference Include="HeuristicLab.Operators.Programmable-3.3">61 <HintPath>..\bin\HeuristicLab.Operators.Programmable-3.3.dll</HintPath>62 </Reference>63 <Reference Include="HeuristicLab.Optimization-3.3">64 <HintPath>..\bin\HeuristicLab.Optimization-3.3.dll</HintPath>65 </Reference>66 <Reference Include="HeuristicLab.Optimizer-3.3">67 <HintPath>..\bin\HeuristicLab.Optimizer-3.3.dll</HintPath>68 </Reference>69 <Reference Include="HeuristicLab.Parameters-3.3">70 <HintPath>..\bin\HeuristicLab.Parameters-3.3.dll</HintPath>71 </Reference>72 <Reference Include="HeuristicLab.Persistence-3.3">73 <HintPath>..\bin\HeuristicLab.Persistence-3.3.dll</HintPath>74 </Reference>75 <Reference Include="HeuristicLab.PluginInfrastructure-3.3">76 <HintPath>..\bin\HeuristicLab.PluginInfrastructure-3.3.dll</HintPath>77 </Reference>78 <Reference Include="HeuristicLab.Problems.DataAnalysis-3.4, Version=3.4.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">79 <SpecificVersion>False</SpecificVersion>80 <HintPath>..\bin\HeuristicLab.Problems.DataAnalysis-3.4.dll</HintPath>81 </Reference>82 <Reference Include="HeuristicLab.Problems.Instances-3.3">83 <HintPath>..\bin\HeuristicLab.Problems.Instances-3.3.dll</HintPath>84 </Reference>85 <Reference Include="HeuristicLab.Random-3.3">86 <HintPath>..\bin\HeuristicLab.Random-3.3.dll</HintPath>87 </Reference>88 34 <Reference Include="System" /> 89 35 <Reference Include="System.Core" /> … … 130 76 </ItemGroup> 131 77 <ItemGroup> 78 <ProjectReference Include="..\HeuristicLab.Analysis\3.3\HeuristicLab.Analysis-3.3.csproj"> 79 <Project>{887425B4-4348-49ED-A457-B7D2C26DDBF9}</Project> 80 <Name>HeuristicLab.Analysis-3.3</Name> 81 </ProjectReference> 82 <ProjectReference Include="..\HeuristicLab.Collections\3.3\HeuristicLab.Collections-3.3.csproj"> 83 <Project>{958B43BC-CC5C-4FA2-8628-2B3B01D890B6}</Project> 84 <Name>HeuristicLab.Collections-3.3</Name> 85 </ProjectReference> 86 <ProjectReference Include="..\HeuristicLab.Common.Resources\3.3\HeuristicLab.Common.Resources-3.3.csproj"> 87 <Project>{0E27A536-1C4A-4624-A65E-DC4F4F23E3E1}</Project> 88 <Name>HeuristicLab.Common.Resources-3.3</Name> 89 </ProjectReference> 90 <ProjectReference Include="..\HeuristicLab.Common\3.3\HeuristicLab.Common-3.3.csproj"> 91 <Project>{A9AD58B9-3EF9-4CC1-97E5-8D909039FF5C}</Project> 92 <Name>HeuristicLab.Common-3.3</Name> 93 </ProjectReference> 94 <ProjectReference Include="..\HeuristicLab.Core\3.3\HeuristicLab.Core-3.3.csproj"> 95 <Project>{C36BD924-A541-4A00-AFA8-41701378DDC5}</Project> 96 <Name>HeuristicLab.Core-3.3</Name> 97 </ProjectReference> 98 <ProjectReference Include="..\HeuristicLab.Data\3.3\HeuristicLab.Data-3.3.csproj"> 99 <Project>{BBAB9DF5-5EF3-4BA8-ADE9-B36E82114937}</Project> 100 <Name>HeuristicLab.Data-3.3</Name> 101 </ProjectReference> 102 <ProjectReference Include="..\HeuristicLab.Encodings.SymbolicExpressionTreeEncoding\3.4\HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4.csproj"> 103 <Project>{06D4A186-9319-48A0-BADE-A2058D462EEA}</Project> 104 <Name>HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4</Name> 105 </ProjectReference> 106 <ProjectReference Include="..\HeuristicLab.Operators.Programmable\3.3\HeuristicLab.Operators.Programmable-3.3.csproj"> 107 <Project>{6A5F8C2D-B0C3-4B36-BC20-9B1A91EE6DB6}</Project> 108 <Name>HeuristicLab.Operators.Programmable-3.3</Name> 109 </ProjectReference> 110 <ProjectReference Include="..\HeuristicLab.Operators\3.3\HeuristicLab.Operators-3.3.csproj"> 111 <Project>{23DA7FF4-D5B8-41B6-AA96-F0561D24F3EE}</Project> 112 <Name>HeuristicLab.Operators-3.3</Name> 113 </ProjectReference> 114 <ProjectReference Include="..\HeuristicLab.Optimization\3.3\HeuristicLab.Optimization-3.3.csproj"> 115 <Project>{14AB8D24-25BC-400C-A846-4627AA945192}</Project> 116 <Name>HeuristicLab.Optimization-3.3</Name> 117 </ProjectReference> 118 <ProjectReference Include="..\HeuristicLab.Parameters\3.3\HeuristicLab.Parameters-3.3.csproj"> 119 <Project>{56F9106A-079F-4C61-92F6-86A84C2D84B7}</Project> 120 <Name>HeuristicLab.Parameters-3.3</Name> 121 </ProjectReference> 122 <ProjectReference Include="..\HeuristicLab.Persistence\3.3\HeuristicLab.Persistence-3.3.csproj"> 123 <Project>{102BC7D3-0EF9-439C-8F6D-96FF0FDB8E1B}</Project> 124 <Name>HeuristicLab.Persistence-3.3</Name> 125 </ProjectReference> 126 <ProjectReference Include="..\HeuristicLab.PluginInfrastructure\3.3\HeuristicLab.PluginInfrastructure-3.3.csproj"> 127 <Project>{94186A6A-5176-4402-AE83-886557B53CCA}</Project> 128 <Name>HeuristicLab.PluginInfrastructure-3.3</Name> 129 </ProjectReference> 132 130 <ProjectReference Include="..\HeuristicLab.Problems.DataAnalysis.Symbolic.Regression\3.4\HeuristicLab.Problems.DataAnalysis.Symbolic.Regression-3.4.csproj"> 133 131 <Project>{5AC82412-911B-4FA2-A013-EDC5E3F3FCC2}</Project> … … 138 136 <Name>HeuristicLab.Problems.DataAnalysis.Symbolic-3.4</Name> 139 137 </ProjectReference> 138 <ProjectReference Include="..\HeuristicLab.Problems.DataAnalysis\3.4\HeuristicLab.Problems.DataAnalysis-3.4.csproj"> 139 <Project>{DF87C13E-A889-46FF-8153-66DCAA8C5674}</Project> 140 <Name>HeuristicLab.Problems.DataAnalysis-3.4</Name> 141 </ProjectReference> 142 <ProjectReference Include="..\HeuristicLab.Problems.Instances\3.3\HeuristicLab.Problems.Instances-3.3.csproj"> 143 <Project>{3540E29E-4793-49E7-8EE2-FEA7F61C3994}</Project> 144 <Name>HeuristicLab.Problems.Instances-3.3</Name> 145 </ProjectReference> 146 <ProjectReference Include="..\HeuristicLab.Random\3.3\HeuristicLab.Random-3.3.csproj"> 147 <Project>{F4539FB6-4708-40C9-BE64-0A1390AEA197}</Project> 148 <Name>HeuristicLab.Random-3.3</Name> 149 </ProjectReference> 140 150 </ItemGroup> 141 151 <Import Project="$(MSBuildToolsPath)\Microsoft.CSharp.targets" /> 152 <PropertyGroup> 153 <PreBuildEvent>set Path=%25Path%25;$(ProjectDir);$(SolutionDir) 154 set ProjectDir=$(ProjectDir) 155 set SolutionDir=$(SolutionDir) 156 set Outdir=$(Outdir) 157 </PreBuildEvent> 158 </PropertyGroup> 142 159 <!-- To modify your build process, add your task inside one of the targets below and uncomment it. 143 160 Other similar extension points exist, see Microsoft.Common.targets. -
branches/sluengo/HeuristicLab.Problems.TradeRules/Interpreter.cs
r9262 r9325 24 24 private int initialTraining; 25 25 private int initialTest; 26 26 27 [ThreadStatic] 27 28 private static Dictionary<ISymbolicExpressionTreeNode, double> signalCache; … … 208 209 #endregion 209 210 210 211 212 213 211 private double Evaluate(Dataset dataset, ref int row, InterpreterState state) 214 212 { … … 248 246 { 249 247 double sum = Evaluate(dataset, ref row, state); 250 int integerValue = (int) 248 int integerValue = (int)Math.Floor(sum); 251 249 if (integerValue > 100) integerValue = 100; 252 250 if (row < integerValue) … … 254 252 string variableName = dataset.GetValue(row, 3); 255 253 double inferiorValue = Convert.ToDouble(variableName); 256 return inferiorValue /(row+1);254 return inferiorValue / (row + 1); 257 255 } 258 256 else … … 268 266 { 269 267 double result = Evaluate(dataset, ref row, state); 270 for (int i = 1; i < currentInstr.nArguments; i++) 271 { 272 if (result > 0.0) result = Evaluate(dataset, ref row, state); 273 else 274 { 275 SkipInstructions(state); 276 } 277 } 278 return result > 0.0 ? 1.0 : -1.0; 268 double result2 = Evaluate(dataset, ref row, state); 269 double total = result + result2; 270 return total < 2 ? -1.0 : 1.0; 279 271 } 280 272 case OpCodes.OR: 281 273 { 282 274 double result = Evaluate(dataset, ref row, state); 283 for (int i = 1; i < currentInstr.nArguments; i++) 284 { 285 if (result <= 0.0) result = Evaluate(dataset, ref row, state); 286 else 287 { 288 SkipInstructions(state); 289 } 290 } 291 return result > 0.0 ? 1.0 : -1.0; 275 double result2 = Evaluate(dataset, ref row, state); 276 double total = result + result2; 277 return total > -2 ? 1.0 : -1.0; 292 278 } 293 279 case OpCodes.NOT: … … 339 325 case OpCodes.Max: 340 326 { 341 int n = (int) 327 int n = (int)Evaluate(dataset, ref row, state); 342 328 double max = Double.NegativeInfinity; 343 int i = Math.Min(n, row);344 while (i>=0)329 int i = Math.Min(n, row); 330 while (i >= 0) 345 331 { 346 332 int position = row - i; 347 333 string variableName = dataset.GetValue(position, 2); 348 334 double intValue = Convert.ToDouble(variableName); 349 if (intValue >max) max = intValue;350 i--;335 if (intValue > max) max = intValue; 336 i--; 351 337 } 352 338 return max; … … 365 351 i--; 366 352 } 367 353 return min; 368 354 } 369 355 case OpCodes.Lag: 370 356 { 371 int n = (int) 372 if (n >row) return 0;357 int n = (int)Evaluate(dataset, ref row, state); 358 if (n > row) return 0; 373 359 int position = row - n; 374 360 string variableName = dataset.GetValue(position, 2); … … 378 364 case OpCodes.MACD: 379 365 { 380 var MACDNode = currentInstr.dynamicNode;381 366 //Taking the number of the days for each EMA 382 367 double firstEMA = Evaluate(dataset, ref row, state); 383 368 double secondEMA = Evaluate(dataset, ref row, state); 384 369 double signal = Evaluate(dataset, ref row, state); 385 386 //Calculating the factor for each EMA387 double factor = 2.0 / (firstEMA + 1.0);388 double factor2 = 2.0 / (secondEMA + 1.0);389 double factor3 = 2.0 / (signal + 1.0);390 391 370 //Initiation of the variables 392 371 double firstElementEMA = -1000000; … … 394 373 double signalValue = -100000; 395 374 double macd = 0; 375 double longitud = 0; 396 376 397 377 //Check if this MACD has previous values and retrieve them 398 if (firstEMACache.ContainsKey( MACDNode)) firstElementEMA = firstEMACache[MACDNode];399 if (secondEMACache.ContainsKey( MACDNode)) secondElementEMA = secondEMACache[MACDNode];400 if (signalCache.ContainsKey( MACDNode)) signalValue = signalCache[MACDNode];401 402 //Calculate the first value in the training for the two EMAs and the signal.403 if (row <= initialTraining || row == initialTest)404 {405 double [] meanValues = dataset.GetDoubleValues("\"Close\"", Enumerable.Range(0, (row + 1))).ToArray();406 firstElementEMA = meanValues[0];407 secondElementEMA = meanValues[0]; 408 double max = (Math.Max(firstEMA, secondEMA) - 1);//The first macd happens when the longest EMA has its first value. We need -1 because row begin in 0.409 for (int i = 1; i <= row; i++)378 if (firstEMACache.ContainsKey(currentInstr.dynamicNode)) firstElementEMA = firstEMACache[currentInstr.dynamicNode]; 379 if (secondEMACache.ContainsKey(currentInstr.dynamicNode)) secondElementEMA = secondEMACache[currentInstr.dynamicNode]; 380 if (signalCache.ContainsKey(currentInstr.dynamicNode)) signalValue = signalCache[currentInstr.dynamicNode]; 381 382 383 //Calculating the factor for each EMA 384 double factor = 2.0 / (firstEMA + 1.0); 385 double factor2 = 2.0 / (secondEMA + 1.0); 386 double factor3 = 2.0 / (signal + 1.0); 387 388 //Calculate the first value in the training for the two EMAs and the signal. 389 if (row <= initialTraining || row == initialTest) 410 390 { 411 firstElementEMA = meanValues[i] * factor + (1 - factor) * firstElementEMA; 412 secondElementEMA = meanValues[i] * factor2 + (1 - factor2) * secondElementEMA; 413 if (i == max) signalValue = firstElementEMA - secondElementEMA;//First signal equals to macd. 414 else if (i > max)//Calculation for the next signals 391 double [] meanValues = dataset.GetDoubleValues("\"Close\"", Enumerable.Range(0, row+1)).ToArray(); 392 firstElementEMA = meanValues[0]; 393 secondElementEMA = meanValues[0]; 394 double max = (Math.Max(firstEMA, secondEMA)-1);//The first macd happens when the longest EMA has its first value. We need -1 because row begin in 0. 395 for (int i = 1; i < meanValues.Length; i++) 415 396 { 416 macd = firstElementEMA - secondElementEMA; 417 signalValue = macd * factor3 + (1 - factor3) * signalValue; 397 firstElementEMA = (meanValues[i] * factor) + ((1 - factor) * firstElementEMA); 398 secondElementEMA = (meanValues[i] * factor2) + ((1 - factor2) * secondElementEMA); 399 if (i == max) signalValue = firstElementEMA - secondElementEMA;//First signal equals to macd. 400 else if (i > max)//Calculation for the next signals 401 { 402 macd = firstElementEMA - secondElementEMA; 403 signalValue = macd * factor3 + (1 - factor3) * signalValue; 404 } 418 405 } 419 406 } 420 } 421 else //The rest of the rows are calculating with the standard EMA formula 422 { 423 //Retrieve the dataset values 424 string variableName = dataset.GetValue(row, 2); 425 double meanValue1 = Convert.ToDouble(variableName); 426 427 //Calculating EMA 428 firstElementEMA = meanValue1 * factor + (1 - factor) * firstElementEMA; 429 secondElementEMA = meanValue1 * factor2 + (1 - factor2) * secondElementEMA; 430 431 //Calculating signal 432 macd = firstElementEMA - secondElementEMA; 433 signalValue = macd * factor3 + (1 - factor3) * signalValue; 434 } 435 436 //Save the values for the next iteration 437 firstEMACache[MACDNode] = firstElementEMA; 438 secondEMACache[MACDNode] = secondElementEMA; 439 signalCache[MACDNode] = signalValue; 440 441 macd = firstElementEMA - secondElementEMA; 407 else //The rest of the rows are calculating with the standard EMA formula 408 { 409 //Retrieve the dataset values 410 string variableName = dataset.GetValue(row, 2); 411 double meanValue1 = Convert.ToDouble(variableName); 412 413 //Calculating EMA 414 firstElementEMA = meanValue1 * factor + (1 - factor) * firstElementEMA; 415 secondElementEMA = meanValue1 * factor2 + (1 - factor2) * secondElementEMA; 416 417 //Calculating signal 418 macd = firstElementEMA - secondElementEMA; 419 signalValue = macd * factor3 + (1 - factor3) * signalValue; 420 } 421 422 //Save the values for the next iteration 423 firstEMACache[currentInstr.dynamicNode] = firstElementEMA; 424 secondEMACache[currentInstr.dynamicNode] = secondElementEMA; 425 signalCache[currentInstr.dynamicNode] = signalValue; 426 442 427 return macd > signalValue ? 1.0 : -1.0; 443 428 } 444 429 case OpCodes.RSI: 445 { 446 447 //Taking the number of the days for EMA 448 double numberOfDays = Evaluate(dataset, ref row, state); 449 450 //Calculate the factor for the EMA 451 double factor = 1.0 / numberOfDays; 452 453 double positiveEMA = 0; 454 double negativeEMA = 0; 455 double yesterdayRSI = double.NegativeInfinity; 456 double todayRSI = double.NegativeInfinity; 457 double outputRSI = double.NegativeInfinity; 458 //Retrieve EMA values 459 if (RSIPositiveCache.ContainsKey(currentInstr.dynamicNode)) positiveEMA = RSIPositiveCache[currentInstr.dynamicNode]; 460 if (RSINegativeCache.ContainsKey(currentInstr.dynamicNode)) negativeEMA = RSINegativeCache[currentInstr.dynamicNode]; 461 if (RSICache.ContainsKey(currentInstr.dynamicNode)) yesterdayRSI = RSICache[currentInstr.dynamicNode]; 462 if (RSIOutputCache.ContainsKey(currentInstr.dynamicNode)) outputRSI = RSIOutputCache[currentInstr.dynamicNode]; 463 464 if (row == initialTraining || row == initialTest) 465 { 466 double[] closeValues = dataset.GetDoubleValues("\"Close\"", Enumerable.Range(0, (row + 1))).ToArray(); 467 outputRSI = -1.0; 468 for (int i = 1; i <= row; i++) 469 { 470 if (numberOfDays>=i) 471 { 472 if ((closeValues[i] - closeValues[i - 1]) > 0) positiveEMA = ((closeValues[i] - closeValues[i - 1])+positiveEMA); 473 else negativeEMA = Math.Abs(closeValues[i] - closeValues[i - 1]) + negativeEMA; 474 if (numberOfDays == i) 475 { 476 positiveEMA = positiveEMA/numberOfDays; 477 negativeEMA = negativeEMA / numberOfDays; 478 yesterdayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 479 } 480 }else{ 481 if ((closeValues[i] - closeValues[i - 1]) > 0) 430 { 431 //Taking the number of the days for EMA 432 double numberOfDays = Evaluate(dataset, ref row, state); 433 434 double positiveEMA = 0; 435 double negativeEMA = 0; 436 double yesterdayRSI = double.NegativeInfinity; 437 double todayRSI = double.NegativeInfinity; 438 double outputRSI = double.NegativeInfinity; 439 //Retrieve EMA values 440 if (RSIPositiveCache.ContainsKey(currentInstr.dynamicNode)) positiveEMA = RSIPositiveCache[currentInstr.dynamicNode]; 441 if (RSINegativeCache.ContainsKey(currentInstr.dynamicNode)) negativeEMA = RSINegativeCache[currentInstr.dynamicNode]; 442 if (RSICache.ContainsKey(currentInstr.dynamicNode)) yesterdayRSI = RSICache[currentInstr.dynamicNode]; 443 if (RSIOutputCache.ContainsKey(currentInstr.dynamicNode)) outputRSI = RSIOutputCache[currentInstr.dynamicNode]; 444 445 446 //Calculate the factor for the EMA 447 double factor = 1.0 / numberOfDays; 448 449 if (row == initialTraining || row == initialTest) 450 { 451 double[] closeValues = dataset.GetDoubleValues("\"Close\"", Enumerable.Range(0, (row + 1))).ToArray(); 452 outputRSI = -1.0; 453 for (int i = 1; i <= row; i++) 454 { 455 if (numberOfDays >= i) 482 456 { 483 positiveEMA = (closeValues[i]-closeValues[i-1]) * factor + (1 - factor) * positiveEMA; 484 negativeEMA = 0 * factor + (1 - factor) * negativeEMA; 457 if ((closeValues[i] - closeValues[i - 1]) > 0) positiveEMA = ((closeValues[i] - closeValues[i - 1]) + positiveEMA); 458 else negativeEMA = Math.Abs(closeValues[i] - closeValues[i - 1]) + negativeEMA; 459 if (numberOfDays == i) 460 { 461 positiveEMA = positiveEMA / numberOfDays; 462 negativeEMA = negativeEMA / numberOfDays; 463 yesterdayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 464 } 485 465 } 486 466 else 487 467 { 488 positiveEMA = 0 * factor + (1 - factor) * positiveEMA; 489 negativeEMA = Math.Abs(closeValues[i] - closeValues[i - 1]) * factor + (1 - factor) * negativeEMA; 468 if ((closeValues[i] - closeValues[i - 1]) > 0) 469 { 470 positiveEMA = (closeValues[i] - closeValues[i - 1]) * factor + (1 - factor) * positiveEMA; 471 negativeEMA = 0 * factor + (1 - factor) * negativeEMA; 472 } 473 else 474 { 475 positiveEMA = 0 * factor + (1 - factor) * positiveEMA; 476 negativeEMA = Math.Abs(closeValues[i] - closeValues[i - 1]) * factor + (1 - factor) * negativeEMA; 477 } 478 479 todayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 480 481 if ((yesterdayRSI < 30) && (todayRSI > 30)) outputRSI = 1.0; 482 else if ((yesterdayRSI > 70) && (todayRSI < 70)) outputRSI = -1.0; 483 yesterdayRSI = todayRSI; 490 484 } 491 492 todayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 493 494 if ((yesterdayRSI < 30) && (todayRSI > 30)) outputRSI = 1.0; 495 else if ((yesterdayRSI > 70) && (todayRSI < 70)) outputRSI = -1.0; 496 yesterdayRSI = todayRSI; 497 } 498 } 499 } 500 else 501 { 502 string todayCloseString = dataset.GetValue(row, 2); 503 string yesterdayCloseString = dataset.GetValue((row - 1), 2); 504 double todayClose = Convert.ToDouble(todayCloseString); 505 double yesterdayClose = Convert.ToDouble(yesterdayCloseString); 506 507 //Calculating EMA 508 if ((todayClose - yesterdayClose) > 0) 509 { 510 positiveEMA = (todayClose-yesterdayClose) * factor + (1 - factor) * positiveEMA; 511 negativeEMA = 0 * factor + (1 - factor) * negativeEMA; 512 } 513 else 514 { 515 positiveEMA = 0 * factor + (1 - factor) * positiveEMA; 516 negativeEMA = Math.Abs(todayClose - yesterdayClose) * factor + (1 - factor) * negativeEMA; 517 } 518 todayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 519 if ((yesterdayRSI < 30) && (todayRSI > 30)) outputRSI = 1.0; 520 else if ((yesterdayRSI > 70) && (todayRSI < 70)) outputRSI = -1.0; 521 } 522 523 //Save positive and negative EMA for the next iteration 524 RSIPositiveCache[currentInstr.dynamicNode] = positiveEMA; 525 RSINegativeCache[currentInstr.dynamicNode] = negativeEMA; 526 RSICache[currentInstr.dynamicNode] = todayRSI; 527 RSIOutputCache[currentInstr.dynamicNode] = outputRSI; 528 529 return outputRSI; 530 } 531 485 } 486 } 487 else 488 { 489 string todayCloseString = dataset.GetValue(row, 2); 490 string yesterdayCloseString = dataset.GetValue((row - 1), 2); 491 double todayClose = Convert.ToDouble(todayCloseString); 492 double yesterdayClose = Convert.ToDouble(yesterdayCloseString); 493 494 //Calculating EMA 495 if ((todayClose - yesterdayClose) > 0) 496 { 497 positiveEMA = (todayClose - yesterdayClose) * factor + (1 - factor) * positiveEMA; 498 negativeEMA = 0 * factor + (1 - factor) * negativeEMA; 499 } 500 else 501 { 502 positiveEMA = 0 * factor + (1 - factor) * positiveEMA; 503 negativeEMA = Math.Abs(todayClose - yesterdayClose) * factor + (1 - factor) * negativeEMA; 504 } 505 todayRSI = 100 - (100 / (1 + (positiveEMA / negativeEMA))); 506 if ((yesterdayRSI < 30) && (todayRSI > 30)) outputRSI = 1.0; 507 else if ((yesterdayRSI > 70) && (todayRSI < 70)) outputRSI = -1.0; 508 } 509 510 //Save positive and negative EMA for the next iteration 511 RSIPositiveCache[currentInstr.dynamicNode] = positiveEMA; 512 RSINegativeCache[currentInstr.dynamicNode] = negativeEMA; 513 RSICache[currentInstr.dynamicNode] = todayRSI; 514 RSIOutputCache[currentInstr.dynamicNode] = outputRSI; 515 516 517 return outputRSI; 518 } 519 532 520 default: throw new NotSupportedException(); 533 521 } … … 598 586 { 599 587 if (CheckExpressionsWithIntervalArithmetic.Value) 600 throw new NotSupportedException("Interval arithmetic is not yet supported in the symbolic data analysis interpreter.");588 throw new NotSupportedException("Interval arithmetic is not yet supported in the symbolic data analysis interpreter."); 601 589 EvaluatedSolutions.Value++; // increment the evaluated solutions counter 602 590 var compiler = new SymbolicExpressionTreeCompiler(); … … 614 602 } 615 603 616 617 604 if (signalCache == null) signalCache = new Dictionary<ISymbolicExpressionTreeNode, double>(); 618 605 if (firstEMACache == null) firstEMACache = new Dictionary<ISymbolicExpressionTreeNode, double>(); … … 622 609 if (RSICache == null) RSICache = new Dictionary<ISymbolicExpressionTreeNode, double>(); 623 610 if (RSIOutputCache == null) RSIOutputCache = new Dictionary<ISymbolicExpressionTreeNode, double>(); 624 625 626 var state = new InterpreterState(code, necessaryArgStackSize); 627 //Evaluate each row of the datase 628 foreach (var rowEnum in rows) 629 { 630 int row = rowEnum; 631 state.Reset(); 632 if (row < initialTraining) yield return -1; 633 else yield return Evaluate(dataset, ref row, state); 634 } 635 } 636 } 637 611 612 613 var state = new InterpreterState(code, necessaryArgStackSize); 614 //Evaluate each row of the datase 615 foreach (var rowEnum in rows) 616 { 617 int row = rowEnum; 618 state.Reset(); 619 if (row < initialTraining) yield return -1; 620 else yield return Evaluate(dataset, ref row, state); 621 } 622 } 638 623 } 639 624 625 } 626 -
branches/sluengo/HeuristicLab.Problems.TradeRules/Symbols/MACDTreeNode.cs
r9262 r9325 6 6 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable; 7 7 using HeuristicLab.Common; 8 using HeuristicLab.Core; 8 9 9 10 namespace HeuristicLab.Problems.TradeRules.Symbols … … 11 12 public sealed class MACDTreeNode : SymbolicExpressionTreeNode 12 13 { 13 public new MACD Symbol { 14 get { return (MACD)base.Symbol;} 15 } 14 15 public new MACD Symbol 16 { 17 get { return (MACD)base.Symbol; } 18 } 19 20 [Storable] 21 private double idMACD; 22 public double ID 23 { 24 get {return idMACD; } 25 set { idMACD = value; } 26 } 16 27 17 28 [StorableConstructor] … … 20 31 private MACDTreeNode(MACDTreeNode original, Cloner cloner) 21 32 : base(original, cloner) { 22 33 idMACD = original.idMACD; 23 34 } 24 35 … … 31 42 } 32 43 44 public override void ResetLocalParameters(IRandom random) 45 { 46 base.ResetLocalParameters(random); 47 48 idMACD = random.NextDouble(); 49 } 50 33 51 public override IDeepCloneable Clone(Cloner cloner) { 34 52 return new MACDTreeNode(this, cloner);
Note: See TracChangeset
for help on using the changeset viewer.
