pyIpnHeuristic 2.0.0

pyIpnHeuristic is a pure Python implementation of some heuristic algorithms

PyIpnHeuristic

pyIpnHeuristic is a pure Python implementation of some heuristic algorithms for the National Polytechnic Institute of Mexico. For more information on pyIpnHeuristic, visit the GitHub project page pyIpnHeuristic

Pip Install

pip install pyIpnHeuristic

Benchmark

Benchmark problems were taken from Liang et al. 2006. Check out the benchmark doc for a deatiled description.

Import benchmark problems as follows:

# Import the benchmark problem methods
# get_pg01, get_pg02, get_pg03, get_pg04, get_pg05,
# get_pg06, get_pg07, get_pg08, get_pg09, get_pg10,
# get_pg11, get_pg12, get_pg13, get_pg14, get_pg15,
# get_pg16, get_pg17, get_pg18, get_pg19, get_pg20,
# get_pg21, get_pg22, get_pg23, get_pg24
from pyIpnHeuristic.benchmark import get_pg01

# get_pg[*problem]() returns problem parameters as:
#    {
#        objective_function": <class 'function'>,
#        "gx": [<class 'function'>, <class 'function'>, ...], # Soft Restrictions
#        "hx": [<class 'function'>, <class 'function'>, ...], # Hard Restrictions
#        "ranges": [[inf(x1), sup(x1)], ..., [inf(xd), sup(xd)]], # List of Ranges for each variable
#        "markdown": "PROBLEM X", # Markdown Problem description 
#        "x": x_best, # Best values
#        "fx": fx_best # Best solution value
#    }
problem = get_pg01()

objective_function = problem.get("objective_function")
gx = problem.get("gx")
hx = problem.get("hx")
ranges = problem.get("ranges")
markdown = problem.get("markdown")
x_best = problem.get("x")
fx_best = problem.get("fx")

Released Algorithms

• Harmony Search

from pyIpnHeuristic.harmonySearch import HarmonySearch
    
harmonySearch = HarmonySearch(
    objective_function, # e.g, def f(*x): return x[0]**2 + x[1] **2
    soft_constrains=g, # e.g, [def g(*x): return x[0], def g(*x): return x[0]**2]
    hard_constrains=h, # e.g, [def h(*x): return x[1], def h(*x): return x[1]**2]
    ranges=ranges, # e.g, [[0, 1], [0, 1]]
    population_size=population_size, # e.g, 4
    smooth=False, # If True, hard restrictions will be treated as soft restrictions
    epsilon=10**-4, # If smmooth is True, then h(x) = 0 --> |h(x)| - epsilon <= 0
    # Harmony Search Parameters
    hcmr=hcmr,
    par=par,
    alpha=alpha
)

harmonySearch.search(
    iterations=T, # Number of iterations to be made
    save_history=True # Save the results for each iteration. Default False
)

• Modified Harmony Search

from pyIpnHeuristic.modifiedHarmonySearch import ModifiedHarmonySearch
    
harmonySearch = ModifiedHarmonySearch(
    objective_function, # e.g, def f(*x): return x[0]**2 + x[1] **2
    soft_constrains=g, # e.g, [def g(*x): return x[0], def g(*x): return x[0]**2]
    hard_constrains=h, # e.g, [def h(*x): return x[1], def h(*x): return x[1]**2]
    ranges=ranges, # e.g, [[0, 1], [0, 1]]
    population_size=population_size, # e.g, 4
    smooth=False, # If True, hard restrictions will be treated as soft restrictions
    epsilon=10**-4, # If smmooth is True, then h(x) = 0 --> |h(x)| - epsilon <= 0
    # Harmony Search Parameters
    hcmr=hcmr,
    par=par,
    alpha=alpha
)

harmonySearch.search(
    iterations=T, # Number of iterations to be made
    save_history=True # Save the results for each iteration. Default False
)

• Differential Evolution:
  • DE/rand/1/ bin
  • DE/best/1
  • DE/current-to-best/1
  • DE/best/2
  • DE/rand/2

from pyIpnHeuristic.differentialEvolution import DifferentialEvolution

differentialEvolution = DifferentialEvolution(
    objective_function, # e.g, def f(*x): return x[0]**2 + x[1] **2
    soft_constrains=g, # e.g, [def g(*x): return x[0], def g(*x): return x[0]**2]
    hard_constrains=h, # e.g, [def h(*x): return x[1], def h(*x): return x[1]**2]
    ranges=ranges, # e.g, [[0, 1], [0, 1]]
    population_size=population_size, # e.g, 4
    smooth=False, # If True, hard restrictions will be treated as soft restrictions
    epsilon=10**-4, # If smooth is True, then h(x) = 0 --> |h(x)| - epsilon <= 0
    # Differential Evolution Parameters
    type=de_type, # Types: "rand-1", "best-1", "current-to-best-1", "best-2, "rand-2", default: "rand-1"
    f=0.9,
    cr=0.05,
)

differentialEvolution.search(
    iterations=T, # Number of iterations to be made
    save_history=True # Save the results for each iteration. Default False
)

• Particle Swarm Optimization

from pyIpnHeuristic.particleSwarmOptimization import ParticleSwarmOptimization

particleSwarmOptimization = ParticleSwarmOptimization(
    objective_function, # e.g, def f(*x): return x[0]**2 + x[1] **2
    soft_constrains=g, # e.g, [def g(*x): return x[0], def g(*x): return x[0]**2]
    hard_constrains=h, # e.g, [def h(*x): return x[1], def h(*x): return x[1]**2]
    ranges=ranges, # e.g, [[0, 1], [0, 1]]
    population_size=population_size, # e.g, 4
    smooth=False, # If True, hard restrictions will be treated as soft restrictions
    epsilon=10**-4, # If smmooth is True, then h(x) = 0 --> |h(x)| - epsilon <= 0
    # Particle Swarm Parameters
    w=0.3,
    c1=0.1,
    c2=1.9
)

particleSwarmOptimization.search(
    iterations=T, # Number of iterations to be made
    save_history=True # Save the results for each iteration. Default False
)

• Artificial Bee Colony

from pyIpnHeuristic.artificialBeeColony import ArtificialBeeColony

artificialBeeColony = ArtificialBeeColony(
    objective_function, # e.g, def f(*x): return x[0]**2 + x[1] **2
    soft_constrains=g, # e.g, [def g(*x): return x[0], def g(*x): return x[0]**2]
    hard_constrains=h, # e.g, [def h(*x): return x[1], def h(*x): return x[1]**2]
    ranges=ranges, # e.g, [[0, 1], [0, 1]]
    population_size=population_size, # e.g, 4
    smooth=False, # If True, hard restrictions will be treated as soft restrictions
    epsilon=10**-4, # If smmooth is True, then h(x) = 0 --> |h(x)| - epsilon <= 0
    # Artificial Bee Colony Parameters
    mr=0.3,
    max_trials=3,
)

artificialBeeColony.search(
    iterations=T, # Number of iterations to be made
    save_history=True # Save the results for each iteration. Default False
)