private async Task <StringBuilder> Predict(List <List <string> > trainData,
List <List <string> > rowNames, List <List <string> > testData)
{
StringBuilder sb = null;
try
{
//ml algo rule: iterations can also set rowcount and -1 for mlinstructs removed
int iRowCount = (Shared.GetRowCount(_iterations, trainData.Count) - 1);
//columns of data used and returned in DataResults
_actualColNames = Shared.GetActualColNames(_colNames, _depColNames).ToArray();
//ml instructions associated with actual colNames
List <string> normTypes = Shared.GetNormTypes(trainData[0], _colNames, _depColNames);
//instructions in both row names and datasets
List <string> actualMLInstructs = Shared.GetAlgoInstructs(rowNames);
actualMLInstructs.AddRange(normTypes);
// error allowance
double dbPlusorMinus
= CalculatorHelpers.ConvertStringToDouble(actualMLInstructs[0]);
//converts rows to columns with normalized data
List <List <double> > trainDB = Shared.GetNormalizedDData(trainData,
this.IndicatorQT, _colNames, _depColNames, normTypes, "F2");
List <List <double> > testDB = Shared.GetNormalizedDData(testData,
this.IndicatorQT, _colNames, _depColNames, normTypes, "F2");
//make a new list with same matrix, to be replaced with results
int iColCount = testDB.Count;
if (_subalgorithm == MATHML_SUBTYPES.subalgorithm_03.ToString().ToString())
{
//subalgo02 needs qtm and percent probability of accuracy, qtm, low ci, high ci
iColCount = testDB.Count + 5;
//normtypes need full columns before insertion
normTypes = Shared.FixNormTypes(normTypes, iColCount);
}
//row count comes from original testdata to account for the instructions row
DataResults = CalculatorHelpers.GetList(testData.Count, iColCount);
DataResults[0] = normTypes;
//dep var output count
int numOutput = 1;
//less col[0]
int numInput = trainDB.Count - 1;
int numHidden = 12;
//can truncate the data to iRowCount
double[][] trainInputs = Shared.MakeInputDData(trainDB, iRowCount, this.IndicatorQT,
numInput);
//build a neural network
NeuralNetwork2 nn2 = new NeuralNetwork2(numInput, numHidden, numOutput);
int maxEpochs = iRowCount;
double learnRate = 0.001;
//train nn2
double[] wts = nn2.Train(trainInputs, maxEpochs, learnRate, sb);
//mean squared error
double trainErr = nn2.Error(trainInputs);
//final model accuracy
double trainAcc = nn2.Accuracy(trainInputs, dbPlusorMinus);
//add classified test data to DataResults
bool bHasNewClassifs = await AddNewClassifications(nn2, testDB,
trainAcc, trainErr, iRowCount, dbPlusorMinus, _ciLevel);
}
catch (Exception ex)
{
IndicatorQT.ErrorMessage = ex.Message;
}
return(sb);
}
private async Task <StringBuilder> Classify(List <List <string> > trainData,
List <List <string> > rowNames, List <List <string> > testData)
{
StringBuilder sb = null;
try
{
//columns of data used and returned in DataResults
int iRowCount = (Shared.GetRowCount(_iterations, trainData.Count) - 1);
_actualColNames = Shared.GetActualColNames(_colNames, _depColNames).ToArray();
//ml instructions associated with actual colNames
List <string> normTypes = Shared.GetNormTypes(trainData[0], _colNames, _depColNames);
//instructions in both row names and datasets
List <string> actualMLInstructs = Shared.GetAlgoInstructs(rowNames);
actualMLInstructs.AddRange(normTypes);
// prevent joint counts with 0
bool withLaplacian
= actualMLInstructs[0].ToLower().Contains("true") ? true : false;
//converts rows to columns with normalized data
List <List <string> > trainDB = Shared.GetNormalizedSData(trainData,
this.IndicatorQT, _colNames, _depColNames, normTypes, "F0");
List <List <string> > testDB = Shared.GetNormalizedSData(testData,
this.IndicatorQT, _colNames, _depColNames, normTypes, "F0");
int iColCount = testDB.Count;
if (_subalgorithm == MATHML_SUBTYPES.subalgorithm_01.ToString().ToString())
{
//subalgo02 needs qtm and percent probability of accuracy, mse, qtm, low ci, high ci
iColCount = testDB.Count + 2;
}
//row count comes from original testdata to account for the instructions row
DataResults = CalculatorHelpers.GetList(testData.Count, iColCount);
DataResults[0] = normTypes;
// trainData columns define number of rows (depcolumns.Length + 1)
string[][] attributeValues = new string[trainDB.Count][];
//for each column of trainDB, fill in the unique attribute names (i.e. gender = 2 unique atts)
for (int i = 0; i < trainDB.Count; i++)
{
attributeValues[i] = Shared.GetAttributeGroups(i, trainDB, this.IndicatorQT);
}
int[][][] jointCounts = MakeJointCounts(trainDB, attributeValues);
int[] dependentCounts = MakeDependentCounts(jointCounts, attributeValues[0].Length);
//classify everything in test dataset and add result to new columns in test dataset
List <string> predictors = new List <string>();
int d = 0;
int iRowLength = DataResults[1].Count;
string sAttribute = string.Empty;
for (int r = 0; r < DataResults.Count - 1; r++)
{
predictors = new List <string>();
//cols have separate set of predictors
for (int j = 0; j < testDB.Count; j++)
{
//prepare mathresults
DataResults[r + 1][j] = testDB[j][r];
if (j > 0)
{
//going down the rows (j) in the column (r)
predictors.Add(testDB[j][r]);
}
}
d = await Classify(r + 1, attributeValues, predictors.ToArray(),
jointCounts, dependentCounts, withLaplacian, attributeValues.Length - 1);
for (int l = 0; l < attributeValues[0].Length; l++)
{
if (d == l)
{
sAttribute = Shared.ConvertAttributeToLabel(attributeValues[0][l],
this.IndicatorQT);
DataResults[r + 1][iRowLength - 2] = sAttribute;
}
}
}
}
catch (Exception ex)
{
IndicatorQT.ErrorMessage = ex.Message;
}
return(sb);
}