/// <summary>
/// Adds a new connection between 2 randomly chosen nodes.
/// The first node can not be an output.
/// The second node cannot be an input node if the first node is an input node, and the layer cannot be the same as the first node.
///
/// If the connection that is going to be created already exists, there will be a maximum of 5 attempts to create a new one.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="innovationFunction">The innovation function.</param>
protected void AddConnectionMutation(TrainingRoomSettings trainingRoomSettings, Func <uint, uint, uint> innovationFunction)
{
// Set initial attempt counter to 0.
int attemptsDone = 0;
// Rebuild structure so we get layer information to make sure we don't get circular dependencies.
RebuildStructure();
while (true)
{
// Prepare start and end node.
Node startNode;
Node endNode;
// Set start node to a random node while start note is of type OutputNode.
do
{
startNode = TempNodes[trainingRoomSettings.Random.Next(TempNodes.Count)];
} while (IsNodeAnOutputNode(startNode));
// Set end node to a random node while the layer of the start and end node are the same or
// start node is of type InputNode and end node is of type InputNode.
do
{
endNode = TempNodes[trainingRoomSettings.Random.Next(TempNodes.Count)];
} while (endNode.Layer == startNode.Layer || (IsNodeAnInputNode(startNode) && IsNodeAnInputNode(endNode)));
// If end node layer is higher than start node layer swap them.
if (endNode.Layer > startNode.Layer)
{
Node temp = endNode;
endNode = startNode;
startNode = temp;
}
// If the connection between the selected start and end node already exists continue the loop, until 5 attempts are done,
// then return early.
if (ConnectionExists(startNode, endNode))
{
if (attemptsDone >= 5)
{
return;
}
attemptsDone += 1;
continue;
}
// Create a new connection gene from the start and end nodes.
ConnectionGene connection = CreateConnectionGene(innovationFunction(startNode.NodeIdentifier, endNode.NodeIdentifier), startNode.NodeIdentifier, endNode.NodeIdentifier, trainingRoomSettings.Random.NextDouble() * 2 - 1);
// Add the connection gene.
ConnectionGenes.Add(connection);
// Update the max innovation.
_maxInnovation = Math.Max(connection.InnovationNumber, _maxInnovation);
// Break the loop.
break;
}
}
/// <summary>
/// Checks if the organism is of the same species.
/// It compares the given organism to a randomly chosen organism from the species.
/// </summary>
/// <param name="organism">The organism to check.</param>
/// <param name="trainingRoomSettings">The training rooms settings.</param>
/// <returns>Returns <c>true</c> if he organism is of the same species; otherwise, <c>false</c>.</returns>
internal bool IsSameSpecies(Organism organism, TrainingRoomSettings trainingRoomSettings)
{
// If the generation is 0, only 1 species should exist.
// So, there is no need check for the same species.
if (organism.Generation == 0)
{
return(true);
}
// Get all organisms from the previous generation.
List <Organism> lastGenOrganisms = Organisms.FindAll(org => org.Generation == organism.Generation - 1);
// If the last generation does not contain any organisms, that indicates that the species is new.
if (!lastGenOrganisms.Any())
{
// The new species will only have organisms from the current generation.
// So, we make the current organisms the representatives of the last generation.
lastGenOrganisms = Organisms;
}
// Gets the randomly selected organism from the last generation organisms.
Organism organismToCompare = lastGenOrganisms[trainingRoomSettings.Random.Next(lastGenOrganisms.Count)];
// Compares if the organism is from the same species.
return(organismToCompare.IsSameSpecies(organism, trainingRoomSettings));
}
/// <summary>
/// Adds a node by selecting a random connection, disabling it and replacing it by 2 genes.
/// The first gene goes from the original input node, to the new node with a weight of 1.
/// The second gene goes from the the new node to the old output node with the same weight as the original connection.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="getAndIncreaseNodeIdFunction">The get and increase node id function.</param>
/// <param name="innovationFunction">The innovation function.</param>
private void AddNodeMutation(TrainingRoomSettings trainingRoomSettings, Func <uint> getAndIncreaseNodeIdFunction, Func <uint, uint, uint> innovationFunction)
{
// If the connection genes count is 0, then there are not connection genes to mutate.
if (ConnectionGenes.Count == 0)
{
return;
}
// Find a random connection.
ConnectionGene theChosenOne = ConnectionGenes.ElementAt(trainingRoomSettings.Random.Next(ConnectionGenes.Count));
// Disable the connection gene.
theChosenOne.Enabled = false;
// Generate the new node's id.
uint newNodeId = getAndIncreaseNodeIdFunction();
// Get the old in and out node identifiers.
uint oldInId = theChosenOne.InNodeIdentifier;
uint oldOutId = theChosenOne.OutNodeIdentifier;
// Create a connectionGene from oldIn (a) to new (c) and from new (c) to oldOut (b)
ConnectionGene aToC = CreateConnectionGene(innovationFunction(oldInId, newNodeId), oldInId, newNodeId, 1);
ConnectionGene cToB = CreateConnectionGene(innovationFunction(newNodeId, oldOutId), newNodeId, oldOutId, theChosenOne.Weight);
// Add new connection genes.
ConnectionGenes.Add(aToC);
ConnectionGenes.Add(cToB);
// Update the max innovation.
_maxInnovation = Math.Max(Math.Max(cToB.InnovationNumber, aToC.InnovationNumber), _maxInnovation);
}
/// <summary>
/// Mutates the the organism based on the training room settings.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="getAndIncreaseNodeIdFunction">The get and increase node id function.</param>
/// <param name="innovationFunction">The innovation function.</param>
public void Mutate(TrainingRoomSettings trainingRoomSettings, Func <uint> getAndIncreaseNodeIdFunction, Func <uint, uint, uint> innovationFunction)
{
// If the random value is lower than the add connection chance add a connection mutation.
if (trainingRoomSettings.Random.NextDouble() < trainingRoomSettings.AddConnectionChance)
{
AddConnectionMutation(trainingRoomSettings, innovationFunction);
}
// If the random value is lower than the add node chance add a node mutation.
if (trainingRoomSettings.Random.NextDouble() < trainingRoomSettings.AddNodeChance)
{
AddNodeMutation(trainingRoomSettings, getAndIncreaseNodeIdFunction, innovationFunction);
}
// If the random value is higher than the mutate weight chance or the connection gene count is lower than 0 return early.
if (trainingRoomSettings.Random.NextDouble() > trainingRoomSettings.MutateWeightChance || ConnectionGenes.Count <= 0)
{
return;
}
// Find a random connection gene.
ConnectionGene connectionGene = ConnectionGenes[trainingRoomSettings.Random.Next(ConnectionGenes.Count)];
// If the random value is lower than the weight reassign chance set a new weight else add weight.
if (trainingRoomSettings.Random.NextDouble() < trainingRoomSettings.WeightReassignChance)
{
connectionGene.Weight = trainingRoomSettings.Random.NextDouble() * 2 - 1;
}
else
{
connectionGene.Weight += (trainingRoomSettings.Random.NextDouble() * 2 - 1) * 0.1;
}
//TODO: Inputs mutate?
}
/// <summary>
/// Checks if the other organism is of the same species.
/// </summary>
/// <param name="organism">The other organism.</param>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <returns>Returns <c>true</c> if the other organism is of the same species; otherwise, <c>false</c>.</returns>
public bool IsSameSpecies(Organism organism, TrainingRoomSettings trainingRoomSettings)
{
//TODO: Check inputs and outputs when they mutate
// Sets the initial excess.
int excess = 0;
// If the organism has connection genes then add the amount of connection genes
// where the innovation number is higher than the max innovation to the excess.
if (organism.ConnectionGenes.Any())
{
excess += organism.ConnectionGenes.Count(gene => gene.InnovationNumber > _maxInnovation);
}
// Sets initial variables.
double weightDiff = 0;
int disjoint = 0;
int weightDiffCount = 0;
// For each connection gene where the given organism contains a single connection gene with the same innovation number
// add the weight difference to the weight difference variable and add 1 to the weight difference counter; otherwise, add one
// to the disjoint counter.
foreach (ConnectionGene gene in ConnectionGenes)
{
ConnectionGene otherGene = organism.ConnectionGenes.SingleOrDefault(gen => gen.InnovationNumber == gene.InnovationNumber);
if (otherGene != default)
{
weightDiff += Math.Abs(gene.Weight - otherGene.Weight);
weightDiffCount++;
}
else
{
disjoint++;
}
}
// If the weight difference count is not 0 than compound weight difference to weight difference count.
if (weightDiffCount != 0)
{
weightDiff /= weightDiffCount;
}
// Set genome count to the max between connection genes count of the current organism and the other organism connection genes count.
int genomeCount = Math.Max(ConnectionGenes.Count, organism.ConnectionGenes.Count);
// The coefficients SpeciesExcessGeneWeight, SpeciesDisjointGeneWeight, and SpeciesAverageWeightDiffWeight
// allow us to adjust the importance of the three factors, and the factor N,
// the number of genes in the larger genome, normalizes for genome size
// (N can be set to 1 if both genomes are small, i.e., consist of fewer than 20 genes).
if (genomeCount < 20)
{
genomeCount = 1;
}
// Depending on the training room settings return whether the organism is of the same species.
return(((trainingRoomSettings.SpeciesExcessGeneWeight * excess) / genomeCount +
(trainingRoomSettings.SpeciesDisjointGeneWeight * disjoint) / genomeCount +
trainingRoomSettings.SpeciesAverageWeightDiffWeight * weightDiff) <
trainingRoomSettings.Threshold);
}
/// <summary>
/// Clones the organism.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <returns>Returns a clone of the organism.</returns>
public virtual Organism Clone(TrainingRoomSettings trainingRoomSettings)
{
// Prepares a new id for the organisms to clone with.
Guid newGuid = Guid.NewGuid();
// Create a new organism with the given node id and training room settings.
return(new Organism(newGuid, trainingRoomSettings, Generation, ConnectionGenes.Select(gene => gene.Clone(newGuid)).ToList()));
//TODO: When inputs can mutate they should also be cloned
}
/// <summary>
/// Initializes an instance of the <see cref="Organism"/> class.
/// </summary>
/// <param name="id">The id.</param>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="generation">The current generation.</param>
/// <param name="connectionGenes">The connection genes.</param>
public Organism(Guid id, TrainingRoomSettings trainingRoomSettings, uint generation, List <ConnectionGene> connectionGenes)
{
Id = id;
Generation = generation;
ConnectionGenes = new List <ConnectionGene>(connectionGenes);
if (connectionGenes.Any())
{
_maxInnovation = ConnectionGenes.Max(p => p.InnovationNumber);
}
GenerateInputAndOutputNodes(trainingRoomSettings);
Name = GenerateName(trainingRoomSettings.Random.Next);
}
/// <summary>
/// Generates the input and output nodes.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
protected virtual void GenerateInputAndOutputNodes(TrainingRoomSettings trainingRoomSettings)
{
Inputs = new List <OrganismInputNode>((int)trainingRoomSettings.InputCount);
for (uint nodeIdentifier = 0; nodeIdentifier < trainingRoomSettings.InputCount; nodeIdentifier++)
{
Inputs.Add(new OrganismInputNode(this, new InputNode(nodeIdentifier)));
}
Outputs = new List <OrganismOutputNode>((int)trainingRoomSettings.OutputCount);
for (uint nodeIdentifier = 0; nodeIdentifier < trainingRoomSettings.OutputCount; nodeIdentifier++)
{
Outputs.Add(new OrganismOutputNode(this, new OutputNode(nodeIdentifier + trainingRoomSettings.InputCount)));
}
}
/// <summary>
/// Initializes an instance of the <see cref="Organism"/> class.
/// </summary>
/// <param name="generation">The current generation.</param>
/// <param name="trainingRoomSettings">The training room settings.</param>
public Organism(uint generation, TrainingRoomSettings trainingRoomSettings)
{
// Generates a new guid for the organism, this is needed so EF can set the
// foreign shadow key: OrganismId, on OrganismInputNode/OrganismOutputNode.
Id = Guid.NewGuid();
// Sets the current generation.
Generation = generation;
// Initializes the list of connection genes.
ConnectionGenes = new List <ConnectionGene>();
// Initializes the Input & Output lists for the many to many relations.
GenerateInputAndOutputNodes(trainingRoomSettings);
// Generates a random name for the organism.
Name = GenerateName(trainingRoomSettings.Random.Next);
}
/// <summary>
/// Initializes an instance of the <see cref="Organism"/> class.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="innovationFunction">The innovation function.</param>
/// <param name="generation">The generation.</param>
public Organism(TrainingRoomSettings trainingRoomSettings, Func <uint, uint, uint> innovationFunction, uint generation) : this(generation, trainingRoomSettings)
{
foreach (OrganismInputNode organismInputNode in Inputs)
{
InputNode inputNode = organismInputNode.InputNode;
foreach (OrganismOutputNode organismOutputNode in Outputs)
{
OutputNode outputNode = organismOutputNode.OutputNode;
// Create a new connection gene from the start and end nodes.
ConnectionGene connection = CreateConnectionGene(innovationFunction(inputNode.NodeIdentifier, outputNode.NodeIdentifier), inputNode.NodeIdentifier, outputNode.NodeIdentifier, trainingRoomSettings.Random.NextDouble() * 2 - 1);
// Add the connection gene.
ConnectionGenes.Add(connection);
}
}
}
/// <summary>
/// Gets a random organism from this species.
/// </summary>
/// <param name="generation">The current generation.</param>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <returns>Returns a randomly chosen <see cref="Organism"/>.</returns>
internal Organism GetRandomOrganism(uint generation, TrainingRoomSettings trainingRoomSettings)
{
// Gets the current generation organisms.
List <Organism> currentGen = Organisms.FindAll(o => o.Generation == generation);
// Return the only organism or a random organism.
if (currentGen.Count == 1)
{
return(currentGen.First());
}
if (currentGen.Count > 1)
{
return(currentGen[trainingRoomSettings.Random.Next(currentGen.Count)]);
}
// Should never happen!
throw new Exception($"The organisms does not contain a organism at the generation: {generation}");
}
/// <summary>
/// Creates an organism based on the genes of 2 parent organisms.
/// For each gene in parent1 we check if parent2 has a gene with the same innovation number, if this is the case we randomly choose a parent to get the gene from.
/// If a gene only exists in one parent we add it no matter what.
/// </summary>
/// <param name="parent2Organism">The other parent.</param>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="organismFactory">The organism factory.</param>
/// <returns>Returns a child organism based on the genes of this organism and the passed organism.</returns>
public Organism Crossover(Organism parent2Organism, TrainingRoomSettings trainingRoomSettings, IFactory <Organism, OrganismFactoryArgument> organismFactory)
{
// Pre-generates a child id for creating connection genes.
Guid childId = Guid.NewGuid();
// Copies the organisms from parent2.
List <ConnectionGene> parent2 = parent2Organism.ConnectionGenes.Select(gene => gene.Clone(childId)).ToList();
foreach (ConnectionGene gene in ConnectionGenes)
{
// Prepares a gene to remove and add.
ConnectionGene geneToRemove = default;
// Clone the gene to add with the new id.
ConnectionGene geneToAdd = gene.Clone(childId);
// Tries to find a connection gene based on the current connection gene in the parent, and if not found, gives a default.
ConnectionGene gene2 = parent2.SingleOrDefault(gen => gen.InnovationNumber == gene.InnovationNumber);
// If the gene is not a default connection gene.
if (gene2 != default)
{
// Set it for removal.
geneToRemove = gene2;
// Depending on a random set the current gene or a random gene to add.
if (trainingRoomSettings.Random.NextDouble() > 0.5)
{
geneToAdd = gene2.Clone(childId);
}
}
// If the gene to remove is not default, remove it from its parent.
if (geneToRemove != default)
{
parent2.Remove(geneToRemove);
}
// If the gene to add is not enabled and a random is lower than the enable connection chance, enable the gene.
if (!geneToAdd.Enabled && trainingRoomSettings.Random.NextDouble() < trainingRoomSettings.EnableConnectionChance)
{
geneToAdd.Enabled = true;
}
// Add gene to the temporary child genes list.
_childGenes.Add(geneToAdd);
}
// Add all remaining genes in the parent to the child genes.
foreach (ConnectionGene geneToAdd in parent2)
{
//Make sure this gene doesn't cause a loop
if (!ConnectionLoops(geneToAdd.InNodeIdentifier, geneToAdd.OutNodeIdentifier, _childGenes))
{
_childGenes.Add(geneToAdd);
}
}
// Create a new organism with all new genes.
Organism organism = organismFactory.Create(new OrganismFactoryArgument
{
Id = childId,
TrainingRoomSettings = trainingRoomSettings,
ConnectionGenes = _childGenes,
Generation = Generation + 1,
CreationType = OrganismCreationType.NEW_WITH_GENES
});
// Clears the temporary child genes list for re-use.
_childGenes.Clear();
// Return the newly created organism.
return(organism);
}
/// <summary>
/// Initializes an instance of the <see cref="Organism"/> class.
/// With an initial generation of 0.
/// </summary>
/// <param name="trainingRoomSettings">The training room settings.</param>
/// <param name="innovationFunction">The innovation function.</param>
public Organism(TrainingRoomSettings trainingRoomSettings, Func <uint, uint, uint> innovationFunction) : this(trainingRoomSettings, innovationFunction, 0)
{
}