Breeding Terrains with Genetic Terrain Programming: The Evolution of Terrain Generators

Table 3

Parameters for a GTP run.

Objective	Generate realistic or aesthetic terrains
Function set	Functions from Table 1, all operating on matrices with float numbers
Terminal set	Terminals from Table 2 chosen randomly
Selection and fitness	Decided by the designer accordingly to desired terrain features or aesthetic appeal
Population	Fixed size with 12 individuals; initial depth limit 6; no tree size limits; random initialisation
Parameters	If 2 individuals are selected: 90% subtree crossover and 10% mutation; if just one individual is selected: 50% mutation (without crossover)
Operators	Three mutation operators are used with equal probability: (1) Replace mutation where a random node is replaced with a new random tree generated by the grow method; (2) Shrink mutation where a random subtree (S) is chosen from the parent tree and replaced by a random subtree of S; (3) Swap mutation where two random subtrees are chosen from the parent tree and swapped, whenever possible the two subtrees do not intersect. One crossover operator is used: subtree crossover where random nodes are chosen from both parent trees, and the respective branches are swapped creating two offsprings
Termination	Can be stopped at any time by the designer, the “best” individual is chosen by the designer