public void createGraph()
{
// Input Nodes (Sensor)
var in1 = new Node("in1");
var in2 = new Node("in2");
var in3 = new Node("in3");
var in4 = new Node("in4");
var in5 = new Node("in5");
InputNodes = new List<Node> { in1, in2, in3, in4, in5 };
// Bias Node
BiasNode = new Node("bias");
BiasNode.Output = 1.0;
// Hidden Nodes
var hidden1 = new Node("hidden1");
hidden1.InternalState = 0;
var hidden2 = new Node("hidden2");
hidden2.InternalState = 0;
hidden1.addConnections(in1, in2, in3, in4, in5, hidden1, hidden2, BiasNode);
hidden2.addConnections(in1, in2, in3, in4, in5, hidden1, hidden2, BiasNode);
HiddenNodes = new List<Node> { hidden1, hidden2 };
// Output Nodes (Motor)
var out1 = new Node("out1");
var out2 = new Node("out2");
out1.addConnections(hidden1, hidden2, out1, out2, BiasNode);
out2.addConnections(hidden1, hidden2, out1, out2, BiasNode);
OutputNodes = new List<Node> { out1, out2 };
}
public MinCogPhenotype(List<Node> inputNodes, List<Node> hiddenNodes, List<Node> outputNodes, Node bias)
{
InputNodes = inputNodes;
HiddenNodes = hiddenNodes;
OutputNodes = outputNodes;
BiasNode = bias;
}
/// <summary>
/// Updates Activation level and output on a node
/// </summary>
/// <param name="node">Node to be updated</param>
private void updateNode(Node node)
{
double s = 0.0; // Sum of weighted input
double dy = 0.0; // Change in activation level
foreach (var upstreamConnection in node.UpstreamConnections)
{
s += upstreamConnection.left.Output*upstreamConnection.right;
}
dy = (-node.InternalState + s)/node.TimeConstant; // Bias included in s
node.InternalState = node.InternalState + dy;
node.Output = 1 / (1 + Math.Pow(Math.E, -(node.Gain * node.InternalState)));
}