public void ComplexCyclic()
{
var connList = new LightweightList <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, -2.0),
new WeightedDirectedConnection <double>(0, 2, 1.0),
new WeightedDirectedConnection <double>(1, 2, 1.0),
new WeightedDirectedConnection <double>(2, 1, 1.0)
};
var connSpan = connList.AsSpan();
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connSpan, 1, 1);
// Create neural net and run tests.
var actFn = new Logistic();
var net = new NeuralNetCyclic(digraph, actFn.Fn, 1);
ComplexCyclic_Inner(net, actFn);
// Create vectorized neural net and run tests.
var vnet = new NeuralNets.Double.Vectorized.NeuralNetCyclic(digraph, actFn.Fn, 1);
ComplexCyclic_Inner(vnet, actFn);
}
public void SimpleAcyclic_DefinedNodes_NodeIdGap()
{
// Simple acyclic graph.
var connList = new LightweightList <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(100, 103, 0.0),
new WeightedDirectedConnection <double>(101, 103, 1.0),
new WeightedDirectedConnection <double>(102, 103, 2.0),
new WeightedDirectedConnection <double>(102, 104, 3.0)
};
var connSpan = connList.AsSpan();
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connSpan, 0, 10);
// The gaps in the node IDs should be removed such that node IDs form a contiguous span starting from zero.
var connListExpected = new LightweightList <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(10, 13, 0.0),
new WeightedDirectedConnection <double>(11, 13, 1.0),
new WeightedDirectedConnection <double>(12, 13, 2.0),
new WeightedDirectedConnection <double>(12, 14, 3.0)
};
var connSpanExpected = connListExpected.AsSpan();
// The graph should be unchanged from the input connections.
CompareConnectionLists(connSpanExpected, digraph.ConnectionIdArrays, digraph.WeightArray);
// Check the node count.
Assert.Equal(15, digraph.TotalNodeCount);
}
public void SingleInput_WeightZero()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, 0.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 1, 1);
// Create neural net
var actFn = new LogisticFunction();
var net = new CyclicNeuralNet(digraph, actFn.Fn, 2, false);
// Note. The single connection weight is zero, so the input value has no affect.
// Activate and test.
net.InputVector[0] = 100.0;
net.Activate();
Assert.AreEqual(0.5, net.OutputVector[0]);
// Activate and test.
net.InputVector[0] = 0;
net.Activate();
Assert.AreEqual(0.5, net.OutputVector[0]);
// Activate and test.
net.InputVector[0] = -100;
net.Activate();
Assert.AreEqual(0.5, net.OutputVector[0]);
}
public void CyclicOutput()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, 1.0),
new WeightedDirectedConnection <double>(1, 1, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 1, 1);
// Create neural net
var actFn = new LogisticFunction();
var net = new CyclicNeuralNet(digraph, actFn.Fn, 1, false);
// Activate and test.
const double input = 0.1;
double inputVal = input;
net.InputVector[0] = inputVal;
for (int i = 0; i < 10; i++)
{
net.Activate();
double outputExpected = actFn.Fn(inputVal);
Assert.AreEqual(outputExpected, net.OutputVector[0]);
inputVal = input + outputExpected;
}
}
public void MultipleInputsOutputs()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 5, 1.0),
new WeightedDirectedConnection <double>(1, 3, 1.0),
new WeightedDirectedConnection <double>(2, 4, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 3, 3);
// Create neural net
var actFn = new LogisticFunction();
var net = new CyclicNeuralNet(digraph, actFn.Fn, 1, false);
// Activate and test.
net.InputVector[0] = 1.0;
net.InputVector[1] = 2.0;
net.InputVector[2] = 3.0;
net.Activate();
Assert.AreEqual(actFn.Fn(2.0), net.OutputVector[0]);
Assert.AreEqual(actFn.Fn(3.0), net.OutputVector[1]);
Assert.AreEqual(actFn.Fn(1.0), net.OutputVector[2]);
}
private static DirectedGraphViewModel CreateGraphViewModel()
{
// Simple acyclic graph.
var connList = new LightweightList <WeightedDirectedConnection <float> >
{
new WeightedDirectedConnection <float>(0, 4, 1f),
new WeightedDirectedConnection <float>(4, 5, 2f),
new WeightedDirectedConnection <float>(5, 2, 3f),
new WeightedDirectedConnection <float>(1, 2, 4f),
new WeightedDirectedConnection <float>(2, 2, 5f),
new WeightedDirectedConnection <float>(2, 3, 5f),
new WeightedDirectedConnection <float>(2, 4, 5f),
new WeightedDirectedConnection <float>(2, 5, 5f)
};
var connSpan = connList.AsSpan();
connSpan.Sort(WeightedDirectedConnectionComparer <float> .Default);
// Create graph.
var digraph = WeightedDirectedGraphBuilder <float> .Create(connSpan, 2, 2);
// Create graph view model, and return.
INodeIdMap nodeIdByIdx = CreateNodeIdByIdx(digraph.TotalNodeCount);
DirectedGraphViewModel graphViewModel = new(digraph, digraph.WeightArray, nodeIdByIdx);
return(graphViewModel);
}
/// <summary>
/// Create with the provided list of connections, and input/output node counts.
/// </summary>
/// <param name="connectionList">A list of weighted connections that describe the graph.</param>
/// <param name="inputCount">Input node count.</param>
/// <param name="outputCount">Output node count.</param>
/// <returns>A new instance of <see cref="WeightedDirectedGraphAcyclic{T}"/>.</returns>
public static WeightedDirectedGraphAcyclic <T> Create(
IList <WeightedDirectedConnection <T> > connectionList,
int inputCount, int outputCount)
{
// Convert the set of connections to a standardised graph representation.
WeightedDirectedGraph <T> digraph = WeightedDirectedGraphBuilder <T> .Create(connectionList, inputCount, outputCount);
// Invoke factory logic specific to acyclic graphs.
return(Create(digraph));
}
public void SimpleAcyclic_DefinedNodes()
{
// Simple acyclic graph.
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(10, 13, 0.0),
new WeightedDirectedConnection <double>(11, 13, 1.0),
new WeightedDirectedConnection <double>(12, 13, 2.0),
new WeightedDirectedConnection <double>(12, 14, 3.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 0, 10);
// The graph should be unchanged from the input connections.
CompareConnectionLists(connList, digraph.ConnectionIdArrays, digraph.WeightArray);
// Check the node count.
Assert.AreEqual(15, digraph.TotalNodeCount);
}
public void SimpleAcyclic()
{
// Simple acyclic graph.
var connList = new LightweightList <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 3, 0.0),
new WeightedDirectedConnection <double>(1, 3, 1.0),
new WeightedDirectedConnection <double>(2, 3, 2.0),
new WeightedDirectedConnection <double>(2, 4, 3.0)
};
var connSpan = connList.AsSpan();
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connSpan, 0, 0);
// The graph should be unchanged from the input connections.
CompareConnectionLists(connSpan, digraph.ConnectionIdArrays, digraph.WeightArray);
// Check the node count.
Assert.Equal(5, digraph.TotalNodeCount);
}
public void SingleInput_WeightOne()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 1, 1);
// Create neural net and run tests.
var actFn = new Logistic();
var net = new NeuralNetCyclic(digraph, actFn.Fn, 1);
SingleInput_WeightOne_Inner(net, actFn);
// Create vectorized neural net and run tests.
var vnet = new NeuralNets.Double.Vectorized.NeuralNetCyclic(digraph, actFn.Fn, 1);
SingleInput_WeightOne_Inner(vnet, actFn);
}
public void SingleInput_WeightOne()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 1, 1);
// Create neural net
var actFn = new LogisticFunction();
var net = new CyclicNeuralNet(digraph, actFn.Fn, 1, false);
// Activate and test.
net.InputVector[0] = 0.0;
for (int i = 0; i < 10; i++)
{
net.Activate();
Assert.AreEqual(0.5, net.OutputVector[0]);
}
// Activate and test.
net.InputVector[0] = 1.0;
for (int i = 0; i < 10; i++)
{
net.Activate();
Assert.AreEqual(actFn.Fn(1), net.OutputVector[0]);
}
// Activate and test.
net.InputVector[0] = 10.0;
for (int i = 0; i < 10; i++)
{
net.Activate();
Assert.AreEqual(actFn.Fn(10), net.OutputVector[0]);
}
}
public void MultipleInputsOutputs()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 5, 1.0),
new WeightedDirectedConnection <double>(1, 3, 1.0),
new WeightedDirectedConnection <double>(2, 4, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 3, 3);
// Create neural net and run tests.
var actFn = new Logistic();
var net = new NeuralNetCyclic(digraph, actFn.Fn, 1);
MultipleInputsOutputs_Inner(net, actFn);
// Create neural net and run tests.
var vnet = new NeuralNets.Double.Vectorized.NeuralNetCyclic(digraph, actFn.Fn, 1);
MultipleInputsOutputs_Inner(vnet, actFn);
}
public void ComplexCyclic()
{
var connList = new List <WeightedDirectedConnection <double> >
{
new WeightedDirectedConnection <double>(0, 1, -2.0),
new WeightedDirectedConnection <double>(0, 2, 1.0),
new WeightedDirectedConnection <double>(1, 2, 1.0),
new WeightedDirectedConnection <double>(2, 1, 1.0)
};
// Create graph.
var digraph = WeightedDirectedGraphBuilder <double> .Create(connList, 1, 1);
// Create neural net
var actFn = new LogisticFunction();
var net = new CyclicNeuralNet(digraph, actFn.Fn, 1, false);
// Simulate network in C# and compare calculated outputs with actual network outputs.
double[] preArr = new double[3];
double[] postArr = new double[3];
postArr[0] = 3.0;
net.InputVector[0] = 3.0;
for (int i = 0; i < 10; i++)
{
preArr[1] = postArr[0] * -2.0 + postArr[2];
preArr[2] = postArr[0] + postArr[1];
postArr[1] = actFn.Fn(preArr[1]);
postArr[2] = actFn.Fn(preArr[2]);
net.Activate();
Assert.AreEqual(postArr[1], net.OutputVector[0]);
}
}
