private void RunModelLogic()
{
if (run && !m_EasyRapidlib.isModelEmpty)
{
// DTW
if (learningType == EasyRapidlib.LearningType.DTW)
{
// We init the running serie if it is null
if (m_RunningSerieDTW.ExampleSerie == null)
{
m_RunningSerieDTW = new RapidlibTrainingSerie();
}
// We start adding examples to the running serie until the user stops
AddExampleToSerie(ref m_RunningSerieDTW);
}
// Classification and Regression
else
{
// Extract all the features from the input object
double[] modelInputs = null;
int sizeModelInput = 0;
// Calculate size of the model input vector [I CAN MOVE THIS TO THE START OF THE PROGRAM]
for (int i = 0; i < m_Features.Count; i++)
{
sizeModelInput += (m_LengthsFeatureVector[i] * inputs.Length);
}
modelInputs = new double[sizeModelInput];
if (modelInputs != null)
{
m_ExtractFeatures(ref modelInputs, inputs);
}
////Debug.Log(input);
////Debug.Log(input.Length);
//for (int i = 0; i < input.Length; i++)
//{
// //Debug.Log(input[i]);
//}
////Debug.Log(outputs);
////Debug.Log(outputs.Length);
//for (int i = 0; i < outputs.Length; i++)
//{
// //Debug.Log(outputs[i]);
//}
// Get the predictions based on the features we input
outputs = m_EasyRapidlib.Run(modelInputs, outputs.Length);
}
}
}
/// <summary>
/// Adds a single example to a serie
/// </summary>
/// <param name="serieToReceiveExample"></param>
public void AddExampleToSerie(ref RapidlibTrainingSerie serieToReceiveExample)
{
// Calculate size of the new example input vector [I CAN MOVE THIS TO THE START OF THE PROGRAM]
int sizeNewExampleInput = 0;
for (int i = 0; i < m_Features.Count; i++)
{
sizeNewExampleInput += (m_LengthsFeatureVector[i] * inputs.Length);
}
double[] singleInputExampleInSerie = new double[sizeNewExampleInput];
m_ExtractFeatures(ref singleInputExampleInSerie, inputs);
// Add a single example to the training serie dtw that we will push later on to easyRapidlib
serieToReceiveExample.AddTrainingExample(singleInputExampleInSerie, OutputDTW.ToString());
}
