Module `nkcs.constants`

Parameters for NKCS coevolutionary experimentation.

Expand source code

#!/usr/bin/python3
#
# Copyright (C) 2019--2024 Richard Preen <rpreen@gmail.com>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
#

"""Parameters for NKCS coevolutionary experimentation."""

from typing import List, TextIO


class Constants:
    """Global constants."""

    F: int = 10  # number of different NKCS experiments to run
    E: int = 10  # number of experiments per NKCS to run
    G: int = 10  # number of generations to run genetic algorithm
    N: int = 20  # number of genes (nodes/characteristics)
    K: int = 2  # number of connections to other (internal) genes
    C: int = 2  # number of external genes affecting each gene
    S: int = 2  # number of species
    P: int = 20  # number of individuals in each species population
    T_SIZE: int = 3  # size of selection and replacement tournament
    P_MUT: float = 1 / N  # per gene probability of mutation
    P_CROSS: float = 0.8  # probability of performing crossover
    REPLACE: str = "worst"  # replace = {"worst", "tournament"}
    M: int = 1000  # number of children per parent to test via model
    MAX_ARCHIVE: int = 5000  #: max evaluated individuals for model training
    H: int = 20  # number of hidden neurons for MLP
    N_MODELS: int = 10  # num surrogate models (for averaging, and std dev)
    PLOT: bool = True  # plot graph
    ACQUISITION: str = "ea"  # acquisition = {"ea", "ei", "uc", "pi", "mean"}
    MODEL: str = "gp"  #  {"gp", "mlp", "svr", "linear", "tree", "gradient"}
    MAX_EVALS: int = P * S * G  #: number of evaluations per experiment
    NKCS_TOPOLOGY: str = "standard"  # topology = {"line", "standard"}
    NUM_THREADS: int = 8  # number of CPU threads for model building
    EXPERIMENT_SAVE: bool = False  # whether to save landscapes and populations
    EXPERIMENT_LOAD: bool = False  # whether to load landscapes and populations


def cons_to_string() -> str:
    """Return a string representation of the constants."""
    string: str = (
        f"{Constants.NKCS_TOPOLOGY}"
        f"_f{Constants.F}"
        f"_e{Constants.E}"
        f"_g{Constants.G}"
        f"_p{Constants.P}"
        f"_s{Constants.S}"
        f"_n{Constants.N}"
        f"_k{Constants.K}"
        f"_c{Constants.C}"
        f"_tsize{Constants.T_SIZE}"
        f"_pmut{Constants.P_MUT}"
        f"_pcross{Constants.P_CROSS}"
        f"_replace{Constants.REPLACE}"
        f"_evals{Constants.MAX_EVALS}"
    )
    if Constants.ACQUISITION != "ea":
        string += (
            f"_m{Constants.M}"
            f"_h{Constants.H}"
            f"_maxa{Constants.MAX_ARCHIVE}"
            f"_nmodels{Constants.N_MODELS}"
            f"_model{Constants.MODEL}"
        )
    return string


def save_constants(fp: TextIO) -> None:
    """Write constants to a data file."""
    fp.write(f"{Constants.ACQUISITION},")
    fp.write(f"{Constants.F},")
    fp.write(f"{Constants.E},")
    fp.write(f"{Constants.G},")
    fp.write(f"{Constants.P},")
    fp.write(f"{Constants.S},")
    fp.write(f"{Constants.N},")
    fp.write(f"{Constants.K},")
    fp.write(f"{Constants.C},")
    fp.write(f"{Constants.T_SIZE},")
    fp.write(f"{Constants.P_MUT},")
    fp.write(f"{Constants.P_CROSS},")
    fp.write(f"{Constants.REPLACE},")
    fp.write(f"{Constants.M},")
    fp.write(f"{Constants.H},")
    fp.write(f"{Constants.MAX_ARCHIVE},")
    fp.write(f"{Constants.N_MODELS},")
    fp.write(f"{Constants.NKCS_TOPOLOGY},")
    fp.write(f"{Constants.MODEL}")
    fp.write("\n")


def read_constants(row: List[str]) -> None:
    """Read constants from a row."""
    Constants.ACQUISITION = str(row[0])
    Constants.F = int(row[1])
    Constants.E = int(row[2])
    Constants.G = int(row[3])
    Constants.P = int(row[4])
    Constants.S = int(row[5])
    Constants.N = int(row[6])
    Constants.K = int(row[7])
    Constants.C = int(row[8])
    Constants.T_SIZE = int(row[9])
    Constants.P_MUT = float(row[10])
    Constants.P_CROSS = float(row[11])
    Constants.REPLACE = str(row[12])
    Constants.M = int(row[13])
    Constants.H = int(row[14])
    Constants.MAX_ARCHIVE = int(row[15])
    Constants.N_MODELS = int(row[16])
    Constants.NKCS_TOPOLOGY = str(row[17])
    Constants.MODEL = str(row[18])

Functions

def cons_to_string() ‑> str

Return a string representation of the constants.

Expand source code

def cons_to_string() -> str:
    """Return a string representation of the constants."""
    string: str = (
        f"{Constants.NKCS_TOPOLOGY}"
        f"_f{Constants.F}"
        f"_e{Constants.E}"
        f"_g{Constants.G}"
        f"_p{Constants.P}"
        f"_s{Constants.S}"
        f"_n{Constants.N}"
        f"_k{Constants.K}"
        f"_c{Constants.C}"
        f"_tsize{Constants.T_SIZE}"
        f"_pmut{Constants.P_MUT}"
        f"_pcross{Constants.P_CROSS}"
        f"_replace{Constants.REPLACE}"
        f"_evals{Constants.MAX_EVALS}"
    )
    if Constants.ACQUISITION != "ea":
        string += (
            f"_m{Constants.M}"
            f"_h{Constants.H}"
            f"_maxa{Constants.MAX_ARCHIVE}"
            f"_nmodels{Constants.N_MODELS}"
            f"_model{Constants.MODEL}"
        )
    return string

def read_constants(row: List[str]) ‑> None

Read constants from a row.

Expand source code

def read_constants(row: List[str]) -> None:
    """Read constants from a row."""
    Constants.ACQUISITION = str(row[0])
    Constants.F = int(row[1])
    Constants.E = int(row[2])
    Constants.G = int(row[3])
    Constants.P = int(row[4])
    Constants.S = int(row[5])
    Constants.N = int(row[6])
    Constants.K = int(row[7])
    Constants.C = int(row[8])
    Constants.T_SIZE = int(row[9])
    Constants.P_MUT = float(row[10])
    Constants.P_CROSS = float(row[11])
    Constants.REPLACE = str(row[12])
    Constants.M = int(row[13])
    Constants.H = int(row[14])
    Constants.MAX_ARCHIVE = int(row[15])
    Constants.N_MODELS = int(row[16])
    Constants.NKCS_TOPOLOGY = str(row[17])
    Constants.MODEL = str(row[18])

def save_constants(fp: ) ‑> None

Write constants to a data file.

Expand source code

def save_constants(fp: TextIO) -> None:
    """Write constants to a data file."""
    fp.write(f"{Constants.ACQUISITION},")
    fp.write(f"{Constants.F},")
    fp.write(f"{Constants.E},")
    fp.write(f"{Constants.G},")
    fp.write(f"{Constants.P},")
    fp.write(f"{Constants.S},")
    fp.write(f"{Constants.N},")
    fp.write(f"{Constants.K},")
    fp.write(f"{Constants.C},")
    fp.write(f"{Constants.T_SIZE},")
    fp.write(f"{Constants.P_MUT},")
    fp.write(f"{Constants.P_CROSS},")
    fp.write(f"{Constants.REPLACE},")
    fp.write(f"{Constants.M},")
    fp.write(f"{Constants.H},")
    fp.write(f"{Constants.MAX_ARCHIVE},")
    fp.write(f"{Constants.N_MODELS},")
    fp.write(f"{Constants.NKCS_TOPOLOGY},")
    fp.write(f"{Constants.MODEL}")
    fp.write("\n")

Classes

class Constants

Global constants.

Expand source code

class Constants:
    """Global constants."""

    F: int = 10  # number of different NKCS experiments to run
    E: int = 10  # number of experiments per NKCS to run
    G: int = 10  # number of generations to run genetic algorithm
    N: int = 20  # number of genes (nodes/characteristics)
    K: int = 2  # number of connections to other (internal) genes
    C: int = 2  # number of external genes affecting each gene
    S: int = 2  # number of species
    P: int = 20  # number of individuals in each species population
    T_SIZE: int = 3  # size of selection and replacement tournament
    P_MUT: float = 1 / N  # per gene probability of mutation
    P_CROSS: float = 0.8  # probability of performing crossover
    REPLACE: str = "worst"  # replace = {"worst", "tournament"}
    M: int = 1000  # number of children per parent to test via model
    MAX_ARCHIVE: int = 5000  #: max evaluated individuals for model training
    H: int = 20  # number of hidden neurons for MLP
    N_MODELS: int = 10  # num surrogate models (for averaging, and std dev)
    PLOT: bool = True  # plot graph
    ACQUISITION: str = "ea"  # acquisition = {"ea", "ei", "uc", "pi", "mean"}
    MODEL: str = "gp"  #  {"gp", "mlp", "svr", "linear", "tree", "gradient"}
    MAX_EVALS: int = P * S * G  #: number of evaluations per experiment
    NKCS_TOPOLOGY: str = "standard"  # topology = {"line", "standard"}
    NUM_THREADS: int = 8  # number of CPU threads for model building
    EXPERIMENT_SAVE: bool = False  # whether to save landscapes and populations
    EXPERIMENT_LOAD: bool = False  # whether to load landscapes and populations

Class variables

var ACQUISITION : str
var C : int
var E : int
var EXPERIMENT_LOAD : bool
var EXPERIMENT_SAVE : bool
var F : int
var G : int
var H : int
var K : int
var M : int
var MAX_ARCHIVE : int: max evaluated individuals for model training
var MAX_EVALS : int: number of evaluations per experiment
var MODEL : str
var N : int
var NKCS_TOPOLOGY : str
var NUM_THREADS : int
var N_MODELS : int
var P : int
var PLOT : bool
var P_CROSS : float
var P_MUT : float
var REPLACE : str
var S : int
var T_SIZE : int