GEATbx: Main page  Tutorial  Algorithms  M-functions  Parameter/Options  Example functions  www.geatbx.com

# 2 General options

### NumberSubpopulation

This option defines the number of subpopulations to use.

• Default value: 4
• Type: positive integer in [1, Inf]
• Example - use 2 subpopulations:
`GeaOpt = geaoptset(GeaOpt, 'NumberSubpopulation', 2);`
• The setting of this parameter corresponds with NumberIndividuals. The number of all individuals in the population is the sum of all individuals in the subpopulations. For instance, 4 subpopulations with 20 individuals each produce a population of 80 individuals.
• When only one subpopulation is defined the population is handled as one (panmictic) population. Two and more subpopulations define the use of the regional population model (also called island model or coarse-grained model) employing migration between subpopulations (see Migration.Do).
• The setting of this option corresponds with NumberIndividuals.
• Option number in previous versions: 21 (2.x and 1.x)

### NumberIndividuals

This option defines the number of individuals per subpopulation to use.

• Default value: [25, 20, 15, 30]
• Type: positive integer in [1, Inf]
• The number of individuals can be the same for all subpopulations or different. Thus, special values for every subpopulation can be defined (multi strategy support).
• Example - set number of individuals in all subpopulations to 20:
`GeaOpt = geaoptset(GeaOpt, 'NumberIndividuals', 20);`
• Example - set number of individuals in each subpopulation to a specific number:
`GeaOpt = geaoptset(GeaOpt, 'NumberIndividuals', [20,15,25,35]);`
• A population/subpopulation should have at least 15 individuals. The larger and more complex a problem is the higher the number of individuals ought to be.
• The setting of this option corresponds with NumberSubpopulation.
• Option number in previous versions: 20 (2.x and 1.x)

### VariableFormat

This option defines the format (real, integer, binary, ordering) of the variables and the conversion (done by bindecod) between this format and the internally used representation of the variables. All conversions are handled inside the main function geamain2 (fully transparent to the user):

• Default value: 0
• Type: integer in [1, 5]
• Depending on the setting a conversion between internal representation (genotype) and the representation of the variables (phenotype) is carried out:
• 0: real values; no conversion (the EA works on real values)
• 1: real values; the EA works on binary values (conversion from binary to real)
• 2: integer values; no conversion (the EA works on integer values)
• 3: integer values; conversion from binary to integer (the EA works on binary values)
• 4: binary values; no conversion (the EA works on binary values)
• 5: ordering / permutation / scheduling problems, any representation of variables; no conversion
• Example - the problem employs integer valued variables and the EA should work on binary values, (convert the internal binary representation to the external integer representation):
`GeaOpt = geaoptset(GeaOpt, 'VariableFormat', 2);`
• The setting of this option corresponds with the recombination and mutation options. Select appropriate operators and their corresponding options: Recombination.Name, Mutation.Name.
• Depending on the internal representation of the variables the appropriate initialization functions is called inside geamain2/initpop:
• Real values (0): initialization by initrp
• Integer values (2): initialization by initip
• Binary values (1, 3, 4): initialization by initbp
• Ordering /permutation / scheduling (5): initialization by initpp
• Option number in previous versions: 19 (2.x and 1.x)
 GEATbx: Main page  Tutorial  Algorithms  M-functions  Parameter/Options  Example functions  www.geatbx.com

This document is part of version 3.8 of the GEATbx: Genetic and Evolutionary Algorithm Toolbox for use with Matlab - www.geatbx.com.
The Genetic and Evolutionary Algorithm Toolbox is not public domain.