public void CanLearnAndInfer1OnRecognizedSubInputs_5By4Layer0_5Coverage()
{
uint maxoutputsize = 5;
var coverage = 0.5;
var clone = false;
uint layerheight = 5;
uint layerwidth = 4;
int inputtolayerratio = 2;
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, layerheight, layerwidth, coverage, clone, maxoutputsize);
var input1 = new SparseMatrix((int)layerheight * inputtolayerratio, (int)layerwidth * inputtolayerratio, 1.0);
var input2 = 2 * input1;
//
layer.Learn(input1);
layer.Learn(input2);
var infer2 = layer.Infer(input2);
var infer1 = layer.Infer(input1);
var input3 = (SparseMatrix)new SparseMatrix((int)layerwidth * inputtolayerratio, (int)layerwidth * inputtolayerratio, 1.0)
.Stack(new SparseMatrix((int)(layerheight - layerwidth) * inputtolayerratio, (int)layerwidth * inputtolayerratio, 3.0));
var infer3 = layer.Infer(input3);
//
foreach (var infer in new SparseMatrix[] { infer1, infer2, infer3 })
{
for (int row = 0; row < layerheight; row++)
{
for (int col = 0; col < layerwidth * layer.MaxNodeOutputSize; col += layer.MaxNodeOutputSize)
{
var subInput = infer.SubMatrix(row, 1, col, layer.MaxNodeOutputSize);
if (infer.Equals(infer3) && row >= layerheight - 2 && row < layerheight)
for (int i = 0; i < subInput.ColumnCount; i++)
Assert.IsTrue(subInput[0, i] < 0.1);
else
Assert.IsTrue(subInput[0, 0] == 1.0 || subInput[0, 1] == 1.0);
}
}
}
}
public void SpatialLayer2DGaussianCanLearnAndInferFromBitmapPicture2DSensor()
{
var sensor = new BitmapPicture2DSensor(presentationsPerInput: 1, pathSpeed: 2, useRandomOrigin: false);
sensor.SetTrainingFolder(TrainingSetPath);
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, 16, 16, 0.1, true);
foreach (var input in sensor.GetTrainingInputs(true))
{
foreach (var iteration in input)
{
layer.Learn(iteration);
}
}
Assert.IsTrue(((SpatialNode2DGaussian)layer.ClonedNode).CoincidencesFrequencies.Keys.Count > 0);
}
public void SeesCorrectSubmatrixWhenTrainingWith0_5Coverage()
{
var inputsize = 10;
var size = (uint)inputsize / 2;
var coverage = 0.5;
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, size, size, coverage, true, 1000);
var list = from i in Enumerable.Range(0, inputsize)
select (double)i;
var matrix = new SparseMatrix(inputsize, inputsize);
for (int i = 0; i < matrix.RowCount; ++i)
{
matrix.SetRow(i, list.ToArray());
}
// for each node in the 2D layer
for (int nodeRow = 0; nodeRow < layer.Height; nodeRow++)
{
for (int nodeCol = 0; nodeCol < layer.Width; nodeCol++)
{
// Act
var subMatrix = layer.GetSubMatrixForNodeAt(nodeRow, nodeCol, matrix);
// check if its 2D input is ok
var width = inputsize / size + coverage * (inputsize - inputsize / size);
var delta = (inputsize - width) / (size - 1);
for (int i = 0; i < (int)width; i++)
{
for (int j = 0; j < (int)width; j++)
{
Assert.AreEqual(nodeCol * delta + j, subMatrix[i, j]);
}
}
}
}
}
public void SeesCorrectSubmatrixWhenCloneTrainingWith0Coverage()
{
var inputsize = 10;
var size = (uint)inputsize / 2;
var coverage = 0.0;
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, size, size, coverage, true, 1000);
var list = from i in Enumerable.Range(0, inputsize)
select (double)i;
var matrix = new SparseMatrix(inputsize, inputsize);
for (int i = 0; i < matrix.RowCount; ++i)
{
matrix.SetRow(i, list.ToArray());
}
// Act
var subMatrix = layer.GetSubMatrixForNodeAt(layer.ClonedNodeRow, layer.ClonedNodeCol, matrix);
//
var width = inputsize / size + coverage * (inputsize - inputsize / size);
var delta = (inputsize - width) / (size - 1);
var centerRow = layer.Height / 2;
var centerCol = layer.Width / 2;
for (int i = 0; i < (int)width; i++)
{
for (int j = 0; j < (int)width; j++)
{
Assert.AreEqual(centerCol * delta + j, subMatrix[i, j]);
}
}
}
public void ErrorWhenLearningAfterLayerIsTrained()
{
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, 1, 1, 0.0, true, 1000);
layer.Learn(new SparseMatrix(5));
layer.Infer(new SparseMatrix(5));
try
{
layer.Learn(new SparseMatrix(5));
Assert.Inconclusive("Should have fired an exception");
}
catch (Exception e)
{
Debug.WriteLine(e.Message);
Assert.IsInstanceOfType(e, typeof(HtmRuleException));
}
}
public void CanLearnNewInputs_5By4Layer0Coverage()
{
uint maxoutputsize = 5;
var coverage = 0.0;
var clone = false;
uint height = 5;
uint width = 4;
var layer = new Spatial2DLayer(SpatialLayerType.Gaussian, height, width, coverage, clone, maxoutputsize);
var input1 = (SparseMatrix) SparseMatrix.Identity(8).Stack(new SparseMatrix(2, 8));
var input2 = 2*input1;
//
layer.Learn(input1);
layer.Learn(input2);
//
for (int i = 0; i < height; i++)
{
for (int j = 0; j < width; j++)
{
var node = (SpatialNode2DGaussian)layer.NodeArray[i, j];
foreach (var coinc in node.CoincidencesFrequencies)
Assert.AreEqual(1, coinc.Value);
}
}
}