#=============================================================================
# (c) Copyright Arthur L. Corcoran, 1992, 1993.  All rights reserved.
#
# Genetic Algorithm configuration file
#=============================================================================
#-----------------------------------------------------------------------------
# User data file
#    This information is not used by the GA, however, it is a convenient
#    way to input a data file name or other information to your application.
#-----------------------------------------------------------------------------
# user_data datafile

#-----------------------------------------------------------------------------
# Seed for random number generator
#
# Usage: rand_seed my_pid
#        rand_seed number
#
#    my_pid = use system pid as random seed
#    number = seed for random number generator, a positive integer
#
# DEFAULT: rand_seed 1
#-----------------------------------------------------------------------------
# rand_seed my_pid
# rand_seed 1

#-----------------------------------------------------------------------------
# The data type of the allele
#
# Usage: datatype [bit | int | int_perm | real]
#
#    bit      = bit string
#    int      = integers
#    int_perm = permutation of integers
#    real     = real numbers
#
# DEFAULT: int_perm
#-----------------------------------------------------------------------------
# datatype bit
# datatype int
# datatype int_perm
# datatype real

#-----------------------------------------------------------------------------
# How to initialize the pool
#
# Usage: initpool [random | from_file filename | interactive]
#
#    random      = generate at random based on 
#                     datatype, chrom_len, & pool_size
#    from_file   = read from a file
#       filename = the name of the file to read from
#    interactive = read from stdin
#
# DEFAULT: initpool random
#-----------------------------------------------------------------------------
# initpool random
# initpool from_file   initpool.dat
# initpool interactive

#-----------------------------------------------------------------------------
# Chromosome length, needed when "initpool random" selected
#
# Usage: chrom_len length
#
#    length = chromosome length, a positive integer
#
# DEFAULT: chrom_len 10
#-----------------------------------------------------------------------------
# chrom_len 25

#-----------------------------------------------------------------------------
# Pool size, needed when "initpool random" selected
#
# Usage: pool_size size
#
#    size = pool size, a positive integer
#
# DEFAULT: 100
#-----------------------------------------------------------------------------
# pool_size 200

#-----------------------------------------------------------------------------
# When to stop the GA
#
#    Convergence means when the variance = 0, or equivalently, when
#    all the fitness values in the pool are identical.
#
#    Iterations means the number of generations for the generational model
#    and the number of trials for the steady state model.  Numbers must
#    be given as positive integers.  It takes roughly pool_size/2
#    iterations of the steady state model to equal one iteration of
#    the generational model.
#
# Usage: stop_after convergence
#        stop_after number [use_convergence | ignore_convergence]
#
#    convergence           - stop when the GA converges
#    number                - stop after specified number of iterations
#       use_convergence    - will stop early if GA converges (default)
#       ignore_convergence - WILL NOT stop early even if GA converges
#
# DEFAULT: stop_after convergence
#-----------------------------------------------------------------------------
# stop_after convergence
# stop_after 500
# stop_after 500 use_convergence
# stop_after 500 ignore_convergence

#-----------------------------------------------------------------------------
# GA Type:
#
# Usage: ga [generational | steady_state]
#
#    generational = generational GA 
#    steady_state = steady-state GA
#
# WARNING: This directive has the following side effects:
#
#         GA type         Directives set as a side effect
#       ------------      -------------------------------
#       generational      selection        roulette
#                         replacement      append
#                         rp_interval      1
#
#       steady-state      selection        rank_biased
#                         replacement      by_rank
#                         rp_interval      100 
#
# DEFAULT: ga generational
#-----------------------------------------------------------------------------
# ga generational              # most commonly used
# ga steady_state              # used by Genitor

#-----------------------------------------------------------------------------
# Generation gap:
#
#    The generation gap  represents a percentage of the population to copy 
#    (clone) to the new pool at each generation.  This only makes sense in
#    a GA with two pools as in the generational model.  A gap of 0.0
#    is the traditional generational algorithm.  As the gap increases,
#    it becomes more like a steady-state algorithm.  A gap of 1.0
#    essentially disables crossover since only reproduction occurs.
#
# Usage: gap number
#
#    number = generation gap, valid range = [0.0 .. 1.0]
#
# DEFAULT: gap 0.0
#-----------------------------------------------------------------------------
# gap 0.3

#-----------------------------------------------------------------------------
# Selection method:
#
# Usage: selection [roulette | rank_biased | uniform_random]
#
#    roulette       = Roulette wheel
#    rank_biased    = Ranked, biased selection as in Genitor
#    uniform_random = Pick one at random
#
# DEFAULT: selection roulette
#-----------------------------------------------------------------------------
# selection roulette           # use with generational GA
# selection rank_biased        # use with steady-state GA
# selection uniform_random     # experimental

#-----------------------------------------------------------------------------
# Selection bias
#
# Usage: bias number
#
#    number = selection bias, valid range = [1.0 .. 2.0]
#             Only used for rank_biased selection
#
# DEFAULT: bias 1.8
#-----------------------------------------------------------------------------
# bias 1.1

#-----------------------------------------------------------------------------
# Crossover method:
#
# Usage: crossover [simple | uniform | order1 | order2 | position | cycle |
#                   pmx | uox | rox | asexual]
#
#    simple    = children get alternate "halves" of parents
#    uniform   = alleles swapped uniformly
#    order1    = order based
#    order2    = order based
#    position  = order based
#    cycle     = order based
#    pmx       = order based
#    uox       = uniform order 
#    rox       = relative order
#    asexual   = swap two alleles
#
# DEFAULT: crossover order1
#-----------------------------------------------------------------------------
# crossover simple
# crossover uniform
# crossover order1            # use ony with integer permutations
# crossover order2            # use ony with integer permutations
  crossover position          # use ony with integer permutations
# crossover cycle             # use ony with integer permutations
# crossover pmx               # use ony with integer permutations
# crossover uox               # use ony with integer permutations
# crossover rox               # use ony with integer permutations
# crossover asexual

#-----------------------------------------------------------------------------
# Crossover Rate
#
# Usage: x_rate number
#
#    number = crossover rate (percentage), valid range = [0.0 .. 1.0]
#             A crossover rate of 0.0 disables crossover
#
# DEFAULT: x_rate 1.0
#-----------------------------------------------------------------------------
# x_rate 0.6

#-----------------------------------------------------------------------------
# Mutation method:
#
# Usage: mutation [simple_invert | simple_random | swap]
#
#    simple_invert = invert a bit
#    simple_random = random bit value
#    swap          = swap two alleles 
#
# DEFAULT: mutation swap
#-----------------------------------------------------------------------------
# mutation simple_invert      # use only with bits
# mutation simple_random      # use only with bits
# mutation swap               # use with any datatype

#-----------------------------------------------------------------------------
# Mutation Rate
#
# Usage: mu_rate number
#
#    number = mutation rate (percentage), valid range = [0.0 .. 1.0]
#             A mutation rate of 0.0 disables mutation
#
# DEFAULT: mu_rate 0.0
#-----------------------------------------------------------------------------
# mu_rate 0.1

#-----------------------------------------------------------------------------
# Replacement method:
#
# Usage: replacement [append | by_rank | first_weaker | weakest]
#
#    append       = append to new pool, as in generational GA
#    by_rank      = insert in sorted order, as in Genitor
#    first_weaker = replace first weaker found in linear scan of pool
#    weakest      = replace weakest member of the pool
#
# DEFAULT: replacement append
#-----------------------------------------------------------------------------
# replacement append           # use with roulette (generational GA)
# replacement by_rank          # use with rank_biased (steady-state GA)
# replacement first_weaker     # experimental
# replacement weakest          # experimental

#-----------------------------------------------------------------------------
# Objective of GA:
#
# Usage: objective [minimize | maximize]
#
#    minimize = minimize evaluation function
#    maximize = maximize evaluation function
#
# DEFAULT: objective minimize
#-----------------------------------------------------------------------------
# objective minimize
# objective maximize

#-----------------------------------------------------------------------------
# Elitism
#
#    Elitism has two actions.  For a generational GA, elitism makes two copies
#    of the best performer in the old pool and and places them in the new 
#    pool, thus ensuring the most fit chromosome survives.  The other action
#    works with both models.  In this case, elitism picks the best two 
#    chromosomes from the parents and children.  Thus, if a child is not as
#    fit as either parent, it will not be placed in the new pool.  Selecting
#    elitism in LibGA performs both actions.
#
# Usage: elitism [true | false]
#
#    true  = ensure best members survive until next generation
#    false = no guarantee best will survive 
#
# DEFAULT: elitism true
#-----------------------------------------------------------------------------
# elitism true
# elitism false

#-----------------------------------------------------------------------------
# Report type
#
# Usage: rp_type [none | minimal | short | long]
#
#    none     = output nothing 
#    minimal  = output configuration and final result 
#    short    = output minimal + statistics only
#    long     = output short + dump pool
#
# DEFAULT: rp_type short
#-----------------------------------------------------------------------------
# rp_type none
# rp_type minimal
# rp_type short
# rp_type long

#-----------------------------------------------------------------------------
# Report interval
#
# Usage: rp_interval number
#
#    number = interval between reports, a positive integer
#
# DEFAULT: rp_interval 1 
#-----------------------------------------------------------------------------
# rp_interval 10

#-----------------------------------------------------------------------------
# Output report filename
#
# Usage: rp_file file_name [file_mode]
#
#    file_name = name of report file
#    file_mode = optional file mode for fopen()
#       a      = append (DEFAULT)
#       w      = overwrite
#
# DEFAULT: (write to stdout)
#-----------------------------------------------------------------------------
# rp_file ga.out
# rp_file ga.out a
# rp_file ga.out w