public AxisVector(AxisVector other) : this(CNTKLibPINVOKE.new_AxisVector__SWIG_1(AxisVector.getCPtr(other)), true)
{
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
}
static void Main(string[] args)
{
int[] shape = new int[] { 6, 3 };
float[] data = new float[] { 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9 };
NDArrayView array = new NDArrayView(shape, data, DeviceDescriptor.CPUDevice);
Variable variable = new Variable(shape, VariableKind.Parameter, CNTK.DataType.Float, array, false, new AxisVector(), false, "", "");
var slicedData = CNTKLib.Slice(variable, AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat(1, 1), IntVector.Repeat(3, 1));
var resultArray = GetArray(slicedData);
Global.UseEngine(SiaNet.Backend.CNTKLib.SiaNetBackend.Instance, DeviceType.CPU);
var K = Global.CurrentBackend;
var a = K.CreateVariable(new float[] { 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, new long[] { 6, 3 });
a.Print();
var sliced = K.SliceRows(a, 1, 2);
//var d = K.CreateVariable(new float[] { 1 }, new long[] { 1, 1 });
//c = c + d;
Console.ReadLine();
}
public AxisVectorEnumerator(AxisVector collection)
{
collectionRef = collection;
currentIndex = -1;
currentObject = null;
currentSize = collectionRef.Count;
}
public QNetworkSimple(int stateSize, int actionSize, int numLayers, int hiddenSize, DeviceDescriptor device, float initialWeightScale = 0.01f)
{
Device = device;
StateSize = stateSize;
ActionSize = actionSize;
//create actor network part
var inputA = new InputLayerDense(stateSize);
var outputA = new OutputLayerDense(hiddenSize, null, OutputLayerDense.LossFunction.None);
outputA.HasBias = false;
outputA.InitialWeightScale = initialWeightScale;
SequentialNetworkDense qNetwork = new SequentialNetworkDense(inputA, LayerDefineHelper.DenseLayers(numLayers, hiddenSize, false, NormalizationMethod.None, 0, initialWeightScale, new ReluDef()), outputA, device);
//seperate the advantage and value part. It is said to be better
var midStream = outputA.GetOutputVariable();
var advantageStream = CNTKLib.Slice(midStream, AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat(0, 1), IntVector.Repeat(hiddenSize / 2, 1));
var valueStream = CNTKLib.Slice(midStream, AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat(hiddenSize / 2, 1), IntVector.Repeat(hiddenSize, 1));
var adv = Layers.Dense(advantageStream, actionSize, device, false, "QNetworkAdvantage", initialWeightScale);
var value = Layers.Dense(valueStream, 1, device, false, "QNetworkValue", initialWeightScale);
InputState = inputA.InputVariable;
//OutputQs = outputA.GetOutputVariable();
OutputQs = value.Output + CNTKLib.Minus(adv, CNTKLib.ReduceMean(adv, Axis.AllStaticAxes())).Output;
}
protected override IEnumerable <DxfPoint> GetExtentsPoints()
{
var radiusX = new DxfVector(Radius, 0.0, 0.0);
var radiusY = new DxfVector(0.0, Radius, 0.0);
// base of the helix
yield return(StartPoint + radiusX);
yield return(StartPoint - radiusX);
yield return(StartPoint + radiusY);
yield return(StartPoint - radiusY);
// other side of the helix
var endPoint = StartPoint + AxisVector.Normalize() * NumberOfTurns * TurnHeight;
yield return(endPoint + radiusX);
yield return(endPoint - radiusX);
yield return(endPoint + radiusY);
yield return(endPoint - radiusY);
}
public void SetRange(int index, AxisVector values)
{
CNTKLibPINVOKE.AxisVector_SetRange(swigCPtr, index, AxisVector.getCPtr(values));
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
}
public static AxisVector UnknownDynamicAxes()
{
AxisVector ret = new AxisVector(CNTKLibPINVOKE.Axis_UnknownDynamicAxes(), false);
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
public static AxisVector DefaultInputVariableDynamicAxes()
{
AxisVector ret = new AxisVector(CNTKLibPINVOKE.Axis_DefaultInputVariableDynamicAxes(), false);
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
private AxisVector _DynamicAxes()
{
AxisVector ret = new AxisVector(CNTKLibPINVOKE.Variable__DynamicAxes(swigCPtr), false);
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
public AxisVector GetRange(int index, int count)
{
global::System.IntPtr cPtr = CNTKLibPINVOKE.AxisVector_GetRange(swigCPtr, index, count);
AxisVector ret = (cPtr == global::System.IntPtr.Zero) ? null : new AxisVector(cPtr, true);
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
public static AxisVector Repeat(Axis value, int count)
{
global::System.IntPtr cPtr = CNTKLibPINVOKE.AxisVector_Repeat(Axis.getCPtr(value), count);
AxisVector ret = (cPtr == global::System.IntPtr.Zero) ? null : new AxisVector(cPtr, true);
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
public static Function Upsample2D2(Variable input, int xOffset = 0, int yOffset = 0, string name = "")
{
var xr = CNTKLib.Reshape(input, new int[] { 1, input.Shape[0], 1, input.Shape[1], input.Shape[2] });
var xx = CNTKLib.Splice(VariableVector.Repeat(xr, 2), new Axis(0));
var xy = CNTKLib.Splice(VariableVector.Repeat(xx, 2), new Axis(2));
var r = CNTKLib.Reshape(xy, new int[] { input.Shape[0] * 2, input.Shape[1] * 2, input.Shape[2] });
var sliceAxis = new AxisVector();
sliceAxis.Add(new Axis(0)); sliceAxis.Add(new Axis(1));
r = CNTKLib.Slice(r, sliceAxis, new IntVector(new int[] { xOffset, yOffset }), new IntVector(new int[] { input.Shape[0] * 2, input.Shape[1] * 2 }));
//name
if (!string.IsNullOrEmpty(name))
{
r.SetName(name);
}
return(r);
}
protected override Function BuildNetwork(Variable input, DeviceDescriptor device, string name)
{
var c1 = UnityCNTK.Layers.Dense(input, HiddenSize, device, true, name + ".Dense", InitialWeightScale);
resultOutput = CNTKLib.Slice(c1, AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat(0, 1), IntVector.Repeat(HiddenSize - 1, 1));
varianceOutput = CNTKLib.Square(CNTKLib.Slice(c1, AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat(HiddenSize - 1, 1), IntVector.Repeat(HiddenSize, 1)));
targetInput = CNTKLib.InputVariable(resultOutput.Shape, resultOutput.DataType, name + ".TargetInput");
var squareErrorRoot = CNTKLib.Sqrt(CNTKLib.SquaredError(resultOutput, targetInput));
var l = CNTKLib.ElementDivide(squareErrorRoot, CNTKLib.ElementTimes(Constant.Scalar(DataType.Float, 2), varianceOutput));
lossOutput = l.Output + CNTKLib.ElementTimes(CNTKLib.Log(varianceOutput), Constant.Scalar(DataType.Float, 1));
//lossOutput = squareError;//test
//add parameters to list
ParameterNames.Add(ParamTypeToName(DenseParamType.Weight));
ParameterNames.Add(ParamTypeToName(DenseParamType.Bias));
return(lossOutput);
}
private static double[][] sum(params int[] axes)
{
var arr = new[]
{
/* total:
* /* */1.0, 2.0, 3.0, /* 6.0 */
/* */ 4.0, 5.0, 6.0, /* 15.0 */
/* total: 5.0, 7.0, 9.0 21.0 */
};
var shape = NDShape.CreateNDShape(new[] { 2, 3 });
Value vx = Value.CreateBatch(shape, arr, DeviceDescriptor.CPUDevice, readOnly: true);
Variable x = Variable.InputVariable(shape, CNTK.DataType.Double, name: "input");
CNTK.Function f;
if (axes == null)
{
f = CNTKLib.Alias(x);
}
else
{
var axisVector = new AxisVector(axes.Select(ax => new Axis(ax)).ToArray());
f = CNTKLib.ReduceSum(x, axis: axisVector);
}
var inputs = new Dictionary <Variable, Value>()
{
{ x, vx }
};
var outputs = new Dictionary <Variable, Value>()
{
{ f, null }
};
f.Evaluate(inputs, outputs, DeviceDescriptor.CPUDevice);
var r = outputs[f].GetDenseData <double>((Variable)f);
return(r.Select(ri => ri.ToArray()).ToArray());
}
public Tensor SliceRows(Tensor x, long start, long end)
{
return(Out(C.Slice(In(x), AxisVector.Repeat(new Axis(0), 1), IntVector.Repeat((int)start, 1), IntVector.Repeat((int)end, 1))));
}
public Variable(NDShape shape, VariableKind varType, DataType dataType, NDArrayView value, bool needsGradient, AxisVector dynamicAxes, bool isSparse, string name, string uid) : this(CNTKLibPINVOKE.new_Variable__SWIG_3(NDShape.getCPtr(shape), (int)varType, (int)dataType, NDArrayView.getCPtr(value), needsGradient, AxisVector.getCPtr(dynamicAxes), isSparse, name, uid), true)
{
if (CNTKLibPINVOKE.SWIGPendingException.Pending)
{
throw CNTKLibPINVOKE.SWIGPendingException.Retrieve();
}
}
internal static global::System.Runtime.InteropServices.HandleRef getCPtr(AxisVector obj)
{
return((obj == null) ? new global::System.Runtime.InteropServices.HandleRef(null, global::System.IntPtr.Zero) : obj.swigCPtr);
}
public static void ValueCopyToSparseCSCTest <T>(DeviceDescriptor device)
{
try
{
// uint numAxes = 2;
int dimension = 3;
IEnumerable <int> dims = new List <int>()
{
dimension
};
var sampleShape = NDShape.CreateNDShape(dims);
int seqLen = 1;
int[] colStarts = { 0, 1 };
int[] rowIndices = { 2 };
T[] nonZeroValues = { (typeof(T) == typeof(float)) ? (T)(object)(0.5F) : (T)(object)(0.5) };
bool sequenceStartFlag = true;
bool readOnly = false;
var sparseValue = Value.CreateSequence <T>(sampleShape, seqLen, colStarts, rowIndices, nonZeroValues, sequenceStartFlag, device, readOnly);
NDArrayView value = null;
bool needsGradient = false;
var dynamicAxes = new AxisVector(0);
bool isSparse = true;
var sampleVariable = new Variable(sampleShape, VariableKind.Output,
(typeof(T) == typeof(float)) ? DataType.Float : DataType.Double, value, needsGradient,
dynamicAxes, isSparse, "sampleVariable", "sampleVariableUid");
int output_seqLen;
IList <int> output_colsStarts;
IList <int> output_rowIndices;
IList <T> output_nonZeroValues;
int output_numNonZeroValues;
sparseValue.GetSparseData <T>(sampleVariable, out output_seqLen,
out output_colsStarts, out output_rowIndices, out output_nonZeroValues, out output_numNonZeroValues);
if (output_seqLen != seqLen)
{
throw new ApplicationException("CopyVariableValueTo returns incorrect sequenceLength");
}
if (!output_colsStarts.SequenceEqual(colStarts))
{
throw new ApplicationException("CopyVariableValueTo returns incorrect colStarts");
}
if (!output_rowIndices.SequenceEqual(rowIndices))
{
throw new ApplicationException("CopyVariableValueTo returns incorrect rowIndices");
}
if (!output_nonZeroValues.SequenceEqual(nonZeroValues))
{
throw new ApplicationException("CopyVariableValueTo returns incorrect nonZeroValues");
}
if (output_numNonZeroValues != nonZeroValues.Count())
{
throw new ApplicationException("CopyVariableValueTo returns incorrect numNonZeroValues");
}
}
catch (Exception ex)
{
Console.WriteLine("Error: {0}\nCallStack: {1}\n Inner Exception: {2}", ex.Message, ex.StackTrace, ex.InnerException != null ? ex.InnerException.Message : "No Inner Exception");
throw ex;
}
}
void createNDArrayViewObjects()
{
NDShape shape = new int[] { 4 };
var v1 = Variable.InputVariable(shape, DataType.Float);
//create NDArray by setting the values manually
var dna1 = new NDArrayView(DataType.Float, shape, device);
dna1.SetValue(1);
dna1.SetValue(2);
dna1.SetValue(3);
dna1.SetValue(4);
dna1.SetValue(5);
//create NDArray
var value = new Value(dna1);
var dd = value.GetDenseData <float>(v1);
Assert.Equal(5f, dd[0][0], 1);
Assert.Equal(5f, dd[0][1], 1);
Assert.Equal(5f, dd[0][2], 1);
Assert.Equal(5f, dd[0][3], 1);
Console.WriteLine("NDArrayView SetValue");
print(dd, v1.Shape);
//2
var array = new float[] { 10f, 11f, 12f, 13f };
var dna2 = new NDArrayView(shape, array, device);
value = new Value(dna2);
dd = value.GetDenseData <float>(v1);
Assert.Equal(10f, dd[0][0], 1);
Assert.Equal(11f, dd[0][1], 1);
Assert.Equal(12f, dd[0][2], 1);
Assert.Equal(13f, dd[0][3], 1);
Console.WriteLine("NDArrayView add array of value");
print(dd, v1.Shape);
//3
var dna3 = new NDArrayView(shape, array, 5 /* must be greater or equal then array length*/, device);
value = new Value(dna3);
dd = value.GetDenseData <float>(v1);
Assert.Equal(10f, dd[0][0], 1);
Assert.Equal(11f, dd[0][1], 1);
Assert.Equal(12f, dd[0][2], 1);
Assert.Equal(13f, dd[0][3], 1);
Console.WriteLine("NDArrayView add array and number of elements");
print(dd, v1.Shape);
//4
var axes = new AxisVector()
{
{ Axis.DefaultBatchAxis() }, { Axis.DefaultDynamicAxis() }
};
NDShape shape2 = new int[] { 2, 2 };
var v2 = Variable.InputVariable(shape2, DataType.Float, "", axes);
var dna4 = new NDArrayView(shape, array, device);
var dna5 = dna4.AsShape(shape2);
value = new Value(dna5);
Assert.Equal(10f, dd[0][0], 1);
Assert.Equal(11f, dd[0][1], 1);
Assert.Equal(12f, dd[0][2], 1);
Assert.Equal(13f, dd[0][3], 1);
dd = value.GetDenseData <float>(v2);
print(dd, v2.Shape);
}