Individual and Champion

class PyGMO.individual

This class represents the concept of an individual. In PyGMO, an individual is the solution to some optimization problem, enriched with some memory of its past (this usually records the best position so far occupied in the solution space by a certain individual), and a velocity, representing the variation in the search space of an individual position.

While it is important to know what an individual is, the user rarely constructs or manipulate individuals as these actions are performed by other classes that also ensure the information contained in an individual is consistent (i.e. length of the chromosomes, etc. )

__init__()

Constructs an empty individual

cur_f

Tuple containing the current fitness of the individual w.r.t a problem (can contain more than one fitness score for multi-objective optimization)

cur_x

Tuple containing the current chromosome (or decision vector) defining the individual

cur_c

Tuple containing the current constraint vector defining an individual w.r.t a problem (this will be empty in unconstrained optimizaion problems)

best_f

Tuple containing the individual fitness corresponding to best_x (the individual can move out of good points as a result of the optimization process)

best_x

Tuple storing the chromosome corresponding to the best position encountered so far since the individual creation (individual can move out of good areas as a resualt of the optimization process)

best_c

Tuple containing the constraint vector corresponding to best_x

cur_v

Velocity of an individual (that is the difference in cur_x between generations). This information is crucial in algorithms such as Particle Swarm Optimization (PSO), but it gets, by default, updated also by other algorithms.

class PyGMO.champion

This class represents the concept of a champion, i.e. the best among a set of individuals. Differently from an PyGMO.individual a champion does not have memory, nor a velocity. Similarly from the class PyGMO.individual the user rarely constructs or manipulate objects from this class as these actions are performed by other classes that also ensure the information contained in a champion is consistent (i.e. it actually is the best in a set of individuals, etc. )

__init__()

Constructs an empty champion

f

Tuple containing the fitness of the champion w.r.t a problem (can contain more than one fitness score for multi-objective optimization)

x

Tuple containing the chromosome (or decision vector) defining the champion

c

Tuple containing the constraint vector defining a champion w.r.t a problem (this will be empty in unconstrained optimizaion problems)