evolution library

Classes

Agent
An Agent is a single candidate solution of double values to be hosted in a Population. Each Agent inherits its behavior from ListMixin via XList. XList is subclassable version of ListMixin. An Agent therefore can be treated like a List for the purpose of generating and improving a sinlge candidate solution or testing its fitness.
Fitness<T>
Population
A Population is a collection of candidate solutions, called Agents, which inherits its behavior from ListMixin via XList. XList is subclassable version of ListMixin. A Population therefore can be treated like a List for the purpose of generating and improving candidate solutions.
Random
XList<T>

Extensions

Times

Functions

diff(int positions, int sizeN, int bestN, int randN, int diffN, int seed, int steps, double w, double fitness(List<double>)) Agent
A version of Differential Evolution with unrestricted search space.
diff2(int positions, int sizeN, int bestN, int randN, int diffN, int seed, int steps, double w, double fitness(List<double>), double lower, double upper) Agent
A version of Differential Evolution with restricted search space.
differential(int positions, int sizeN, int bestN, int randN, int diffN, int seed, int steps, double w, double fitness(List<double>)) double
Differential Evolution. [...]
generatePopulation(int sizePopulation, int sizeAgent, Random r, dynamic fitness) Population
Generate an initial Population of Agents with all variables == 0.0.
linear(int positions, int sizeN, int bestN, int randN, int seed, int steps, double w, dynamic fitness) → void
Linear Algorithm [...]
parallelPopulations(int positions, int sizeN, int bestN, int randN, int seed, int steps, double w, dynamic fitness) → void
Parallel Algorithm [...]