script_normal_model_figures.py

# !/usr/bin/env python
# Created by "Thieu" at 20:50, 08/12/2021 ----------%
#       Email: nguyenthieu2102@gmail.com            %
#       Github: https://github.com/thieu1995        %
# --------------------------------------------------%

from config import Config
from utils.visual.convergence import draw_convergence_chart_models, draw_convergence_chart_epochs_popsizes


EPOCH = [1000]  # Number of generations or epoch in neural network and metaheuristics
POP_SIZE = [50]  # Number of population size in metaheuristics

elm = {
    "para": [None]
}

## Evolutionary-based group
ep_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "bout_size": [0.05],  # percentage of child agents implement tournament selection
}
es_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "lamda": [0.75],  # Percentage of child agents evolving in the next generation, default=0.75
}
ma_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pc": [0.85],  # crossover probability
    "pm": [0.15],  # mutation probability
    "p_local": [0.5],  # Probability of local search for each agent, default=0.5
    "max_local_gens": [10],  # Number of local search agent will be created during local search mechanism, default=10
}
ga_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pc": [0.85],  # crossover probability
    "pm": [0.05]  # mutation probability
}
de_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "wf": [0.85],  # weighting factor
    "cr": [0.8],  # crossover rate
    "strategy": [1]  #
}
fpa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p_s": [0.85],  # switch probability, default = 0.8
    "levy_multiplier": [0.01],  # multiplier factor of Levy-flight trajectory
}
cro_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "po": [0.85],  # The rate between free/occupied at the beginning
    "Fb": [0.9],  # BroadcastSpawner/ExistingCorals rate
    "Fa": [0.1],  # fraction of corals duplicates its self and tries to settle in a different part of the reef
    "Fd": [0.1],  # fraction of the worse health corals in reef will be applied depredation
    "Pd": [0.1],  # Probability of depredation
    "GCR": [0.1],  #
    "gamma_min": [0.02],  #
    "gamma_max": [0.2],  #
    "n_trials": [3],  # number of attempts for a larvar to set in the reef.
}

## Swarm-based group
abc_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "n_elites": [16],  # number of bees which provided for good location and other location
    "n_others": [4],
    "patch_size": [5.0],  # patch_variables = patch_variables * patch_factor (0.985)
    "patch_reduction": [0.985],
    "n_sites": [3],  # 3 bees (employed bees, onlookers and scouts), 1 good partition
    "n_elite_sites": [1],
}
acor_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "sample_count": [50],  # Number of Newly Generated Samples
    "intent_factor": [0.5],  # Intensification Factor (Selection Pressure)
    "zeta": [1.0],  # Deviation-Distance Ratio
}
agto_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p1": [0.03],  # p in the paper)
    "p2": [0.8],  # w in the paper
    "beta": [3.0],  # coefficient in updating equation
}
alo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
ao_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
aro_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
avoa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p1": [0.6],  # probability of status transition
    "p2": [0.4],  # probability of status transition
    "p3": [0.6],  # probability of status transition
    "alpha": [0.8],  # probability of 1st best
    "gama": [2.5]  # a factor in the paper
}
ba_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "loudness": [0.8],  # (A_min, A_max): loudness, default = (1.0, 2.0)
    "pulse_rate": [0.95],  # (r_min, r_max): pulse rate / emission rate
    "pf_min": [0.],  # (pf_min, pf_max): pulse frequency
    "pf_max": [10.],
}
beesa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "selected_site_ratio": [0.5],  # (selected_site_ratio, elite_site_ratio)
    "elite_site_ratio": [0.4],
    "selected_site_bee_ratio": [0.1],  # (selected_site_bee_ratio, elite_site_bee_ratio)
    "elite_site_bee_ratio": [2.0],
    "dance_radius": [0.1],  # Bees Dance Radius
    "dance_reduction": [0.99],
}
bes_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "a_factor": [10, ],  # default: 10, determining the corner between point search in the central point, in [5, 10]
    "R_factor": [1.5],  # default: 1.5, determining the number of search cycles, in [0.5, 2]
    "alpha": [2., ],  # default: 2, parameter for controlling the changes in position, in [1.5, 2]
    "c1": [2., ],  # default: 2, in [1, 2]
    "c2": [2., ],  # c1 and c2 increase the movement intensity of bald eagles towards the best and centre points
}
bfo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "C_s": [0.1, ],  # default: 10, determining the corner between point search in the central point, in [5, 10]
    "C_e": [0.001],  # default: 1.5, determining the number of search cycles, in [0.5, 2]
    "Ped": [4, ],  # default: 2, parameter for controlling the changes in position, in [1.5, 2]
    "Ns": [4, ],  # default: 2, in [1, 2]
    "N_adapt": [2, ],  # c1 and c2 increase the movement intensity of bald eagles towards the best and centre points
    "N_split": [40, ]
}
bsa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "ff": [10],  # flight frequency - default = 10
    "pff": [0.8],  # the probability of foraging for food - default = 0.8
    "c1": [1.5],  # [c1, c2]: Cognitive accelerated coefficient, Social accelerated coefficient same as PSO
    "c2": [1.5],
    "a1": [1.0],  # [a1, a2]: The indirect and direct effect on the birds' vigilance behaviours.
    "a2": [1.0],
    "fl": [0.5, ]  # The followed coefficient- default = 0.5
}
coa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "n_coyotes": [5, ]  # number of coyotes per group
}
csa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p_a": [0.3]  # probability a
}
cso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "mixture_ratio": [0.15],  # joining seeking mode with tracing mode
    "smp": [5],  # seeking memory pool, 10 clones  (larger is better but time-consuming)
    "spc": [False],  # self-position considering
    "cdc": [0.8],  # counts of dimension to change  (larger is more diversity but slow convergence)
    "srd": [0.15],  # seeking range of the selected dimension (smaller is better but slow convergence)
    "c1": [0.4],  # same in PSO
    "w_min": [0.5],  # same in PSO
    "w_max": [0.9],
    "selected_strategy": [1],  # 0: best fitness, 1: tournament, 2: roulette wheel, else: random (decrease by quality)
}
dmoa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "peep": [2],
}
do_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
eho_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "alpha": [0.5],  # a factor that determines the influence of the best in each clan
    "beta": [0.5],  # a factor that determines the influence of the x_center
    "n_clans": [5],  # number of clans
}
fa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "max_sparks": [50],  # parameter controlling the total number of sparks generated by the pop_size fireworks
    "p_a": [0.04],  # const parameter
    "p_b": [0.8],  # const parameter
    "max_ea": [40],  # maximum explosion amplitude
    "m_sparks": [5],  # number of sparks generated in each explosion generation
}
ffa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "gamma": [0.001],  # Light Absorption Coefficient
    "beta_base": [2],  # Attraction Coefficient Base Value
    "alpha": [0.2],  # Mutation Coefficient
    "alpha_damp": [0.99],  # Mutation Coefficient Damp Rate
    "delta": [0.05],  # Mutation Step Size
    "exponent": [2],  # Exponent
}
foa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
goa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c_min": [0.00004],
    "c_max": [1.0],
}
gwo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
gwo_woa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
hba_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
hgs_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "PUP": [0.03],  # Switching updating  position probability
    "LH": [1000],  # Largest hunger / threshold
}
hho_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
ja_paras = levy_ja_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
mfo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
mpa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
mrfo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "somersault_range": [2, ]  # somersault factor that decides the somersault range of manta rays
}
msa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "n_best": [5, ],  # how many of the best moths to keep from one generation to the next
    "partition": [0.5, ],  # The proportional of first partition
    "max_step_size": [1.0, ],  # Max step size used in Levy-flight technique, default=1.0
}
nmra_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "bp": [0.75],  # breeding probability (0.75)
}
pfa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
pso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c1": [1.2],  # local coefficient
    "c2": [1.2],  # global coefficient
    "w_min": [0.4],  # weight min factor
    "w_max": [0.9],  # weight max factor
}
ppso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
cpso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c1": [2.05],  # local coefficient
    "c2": [2.05],  # global coefficient
    "w_min": [0.4],  # weight min factor
    "w_max": [0.9],  # weight max factor
}
clpso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c_local": [1.2],  # local coefficient
    "w_min": [0.4],  # weight min factor
    "w_max": [0.9],  # weight max factor
    "max_flag": [7],
}
scso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
sfo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pp": [0.2, ],  # the rate between SailFish and Sardines (N_sf = N_s * pp) = 0.25, 0.2, 0.1
    "AP": [4.0, ],  # A = 4, 6,...
    "epsilon": [0.0001, ],  # = 0.0001, 0.001
}
sho_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "h_factor": [5],  # default = 5, coefficient linearly decreased from 5 to 0
    "N_tried": [10, ],  # default = 10,
}
slo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
srsr_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
ssa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "ST": [0.8],  # ST in [0.5, 1.0], safety threshold value
    "PD": [0.2],  # number of producers
    "SD": [0.1],  # number of sparrows who perceive the danger
}
sso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
sspidera_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "r_a": [1.0],  # the rate of vibration attenuation when propagating over the spider web, default=1.0
    "p_c": [0.7],  # controls the probability of the spiders changing their dimension mask in the random walk step, default=0.7
    "p_m": [0.1]  # the probability of each value in a dimension mask to be one, default=0.1
}
sspidero_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "fp_min": [0.65],  # (fp_min, fp_max): Female Percent, default = (0.65, 0.9)
    "fp_max": [0.9],
}
woa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}

## Physics-based group
sa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "max_sub_iter": [5],  # Maximum Number of Sub-Iteration (within fixed temperature), default=5
    "t0": [1000],
    "t1": [1],
    "move_count": [5],  # Move Count per Individual Solution, default=5
    "mutation_rate": [0.1],  # Mutation Rate, default=0.1
    "mutation_step_size": [0.1],  # Mutation Step Size, default=0.1
    "mutation_step_size_damp": [0.99],
}
wdo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "RT": [3],  # RT coefficient
    "g_c": [0.2],  # gravitational constant
    "alp": [0.4],  # constants in the update equation
    "c_e": [0.4],  # coriolis effect
    "max_v": [0.3],  # maximum allowed speed
}
mvo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "wep_min": [0.2],  # Wormhole Existence Probability (min in Eq.(3.3) paper, default = 0.2
    "wep_max": [1.0],  # Wormhole Existence Probability (max in Eq.(3.3) paper, default = 1.0
}
two_paras = otwo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
efo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "r_rate": [0.3],  # default = 0.3     # Like mutation parameter in GA but for one variable, mutation probability
    "ps_rate": [0.85],  # default = 0.85    # Like crossover parameter in GA, crossover probability
    "p_field": [0.1],  # default = 0.1     # portion of population, positive field
    "n_field": [0.45],  # default = 0.45    # portion of population, negative field
}
nro_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
hgso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "n_clusters": [2]  # Number of clusters
}
aso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "alpha": [10, ],  # Depth weight
    "beta": [0.2, ],  # Multiplier weight
}
eo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
archoa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c1": [2, ],  # Default belongs [1, 2]
    "c2": [6, ],  # Default belongs [2, 4, 6]
    "c3": [2, ],  # Default belongs [1, 2]
    "c4": [0.5, ],  # Default belongs [0.5, 1]
    "acc_max": [0.9, ],  # Default 0.9
    "acc_min": [0.1, ],  # Default 0.1
}
bro_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "threshold": [3]
}

## Human-based group
bso_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "m_clusters": [5, ],  # m: number of clusters
    "p1": [0.2, ],  # probability
    "p2": [0.8, ],  # probability
    "p3": [0.4, ],  # probability
    "p4": [0.5, ],  # probability
    "slope": [20, ],  # k: factor that changing logsig() function's slope
}
ca_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "accepted_rate": [0.15],  # probability of accepted rate, default: 0.15
}
chio_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "brr": [0.15, ],  # Basic reproduction rate, default=0.15
    "max_age": [10, ],  # Maximum infected cases age, default=10
}
fbio_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
gska_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pb": [0.1, ],  # percent of the best   0.1%, 0.8%, 0.1%
    "kf": [0.5],
    "kr": [0.9, ],  # knowledge ratio
    "kg": [5],
}
ica_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "empire_count": [5, ],  # Number of Empires (also Imperialists)
    "assimilation_coeff": [1.5, ],  # Assimilation Coefficient (beta in the paper)
    "revolution_prob": [0.05, ],  # Revolution Probability
    "revolution_rate": [0.1, ],  # Revolution Rate (mu)
    "revolution_step_size": [0.1, ],  # Revolution Step Size (sigma)
    "zeta": [0.1, ],  # Colonies Coefficient in Total Objective Value of Empires
}
lco_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "r1": [2.35]  # step size coefficient
}
qsa_paras = improved_qsa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
saro_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "se": [0.5, ],
    "mu": [15, ],
}
ssdo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
tlo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
spbo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}

## Bio-based group
bbo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p_m": [0.01],
    "elites": [2]  # Number of elites solution
}
eoa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p_c": [0.9],  # default = 0.9, crossover probability
    "p_m": [0.01],  # default = 0.01 initial mutation probability
    "n_best": [2],  # default = 2, how many of the best earthworm to keep from one generation to the next
    "alpha": [0.98],  # default = 0.98, similarity factor
    "beta": [0.9],  # default = 1, the initial proportional factor
    "gamma": [0.9],  # default = 0.9, a constant that is similar to cooling factor of a cooling schedule in the simulated annealing.
}
iwo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "seed_min": [2],  # (Min, Max) Number of Seeds
    "seed_max": [10],
    "exponent": [2],  # Variance Reduction Exponent
    "sigma_start": [1.0],  # (Initial, Final) Value of Standard Deviation
    "sigma_end": [0.01],
}
sbo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "alpha": [0.94, ],  # the greatest step size
    "p_m": [0.05, ],  # mutation probability
    "psw": [0.02]  # percent of the difference between the upper and lower limit (Eq. 7)
}
sma_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "p_t": [0.03],  # probability threshold
}
vcs_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "lamda": [0.5],  # Number of the best will keep (percentage %)
    "sigma": [1.5],  # Weight factor
}
who_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "n_explore_step": [3],  # default = 3, number of exploration step
    "n_exploit_step": [3],  # default = 3, number of exploitation step
    "eta": [0.15],  # default = 0.15, learning rate
    "p_hi": [0.9],  # default = 0.9, the probability of wildebeest move to another position based on herd instinct
    "local_alpha": [0.9],  # default = (0.9, 0.3), (alpha 1, beta 1) - control local movement
    "local_beta": [0.3],
    "global_alpha": [0.2],  # default = (0.2, 0.8), (alpha 2, beta 2) - control global movement
    "global_beta": [0.8],
    "delta_w": [2.0],  # default = (2.0, 2.0) , (delta_w, delta_c) - (dist to worst, dist to best)
    "delta_c": [2.0],
}
bmo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pl": [5],  # [1, pop_size - 1], barnacle’s threshold
}
sos_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
soa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "fc": [2],  # [1, 5], freequency of employing variable A (A linear decreased from fc to 0), default = 2
}
tsa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}

## System-based group
gco_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "cr": [0.7],  # Same as DE algorithm  # default: 0.7, crossover-rate
    "wf": [1.25],  # Same as DE algorithm  # default: 1.25, weighting factor
}
wca_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "nsr": [4],  # Number of rivers + sea (sea = 1)
    "wc": [2.0],  # Coefficient
}
aeo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
aaeo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}

## Math-based group
aoa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "alpha": [5],  # fixed parameter, sensitive exploitation parameter, Default: 5,
    "miu": [0.5],  # fixed parameter , control parameter to adjust the search process, Default: 0.5,
    "moa_min": [0.2],  # range min of Math Optimizer Accelerated, Default: 0.2,
    "moa_max": [0.9],  # range max of Math Optimizer Accelerated, Default: 0.9,
}
cgo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
gbo_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "pr": [0.5],  # Probability Parameter, default = 0.5
    "beta_min": [0.2],
    "beta_max": [1.2],  # Fixed parameter (no name in the paper), default = (0.2, 1.2)
}
hc_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "neighbour_size": [10],  # fixed parameter, sensitive exploitation parameter, Default: 10
}
pss_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "acceptance_rate": [0.9],  # the probability of accepting a solution in the normal range, default = 0.9
    "sampling_method": ["LHS"],  # 'LHS': Latin-Hypercube or 'MC': 'MonteCarlo', default = "LHS"
}
sca_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
info_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
run_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
}
circlesa_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c_factor": [0.8]
}
hs_paras = {
    "epoch": EPOCH, "pop_size": POP_SIZE,
    "c_r": [0.95],  # Harmony Memory Consideration Rate, default = 0.15
    "pa_r": [0.05],  # Pitch Adjustment Rate, default=0.5

}

models = [
    # ## Open this comment only when run the script: get_visual_loss.py
    # {"name": "ELM", "class": "Elm", "param_grid": elm},         # get statistics for ELM models

    ## Evolutionary-based
    # {"name": "EP-ELM", "class": "EpFlnn", "param_grid": ep_paras},  # Evolutionary Programming (EP)
    # {"name": "ES-ELM", "class": "EsFlnn", "param_grid": es_paras},  # Evolution Strategies (ES)
    # {"name": "MA-ELM", "class": "MaFlnn", "param_grid": ma_paras},  # Memetic Algorithm (MA)
    {"name": "GA-ELM", "class": "GaFlnn", "param_grid": ga_paras},  # Genetic Algorithm (GA)
    # {"name": "DE-ELM", "class": "DeFlnn", "param_grid": de_paras},  # Differential Evolution (DE)
    # {"name": "FPA-ELM", "class": "FpaFlnn", "param_grid": fpa_paras},  # Flower Pollination Algorithm (FPA)
    # {"name": "CRO-ELM", "class": "CroFlnn", "param_grid": cro_paras},  # Coral Reefs Optimization (CRO)

    # ## Swarm-based
    # {"name": "PSO-ELM", "class": "PsoFlnn", "param_grid": pso_paras},  # Particle Swarm Optimization (PSO)
    # {"name": "BeesA-ELM", "class": "BeesaFlnn", "param_grid": beesa_paras},  # Bees Algorithm (BeesA)
    # {"name": "CSO-ELM", "class": "CsoFlnn", "param_grid": cso_paras},  # Cat Swarm Optimization (CSO)
    # {"name": "ABC-ELM", "class": "AbcFlnn", "param_grid": abc_paras},  # Artificial Bee Colony (ABC)
    # {"name": "ACOR-ELM", "class": "AcorFlnn", "param_grid": acor_paras},  # Ant Colony Optimization Continuous (ACOR)
    # {"name": "CSA-ELM", "class": "CsaFlnn", "param_grid": csa_paras},  # Cuckoo Search Algorithm (CSA)
    # {"name": "FFA-ELM", "class": "FfaFlnn", "param_grid": ffa_paras},  # Firefly Algorithm (FireflyA)
    # {"name": "FA-ELM", "class": "FaFlnn", "param_grid": fa_paras},  # Fireworks Algorithm (FA)
    # {"name": "BA-ELM", "class": "BaFlnn", "param_grid": ba_paras},  # Bat Algorithm (BA)
    # {"name": "FOA-ELM", "class": "FoaFlnn", "param_grid": foa_paras},  # Fruit-fly Optimization Algorithm (FOA)
    # {"name": "SSpiderO-ELM", "class": "SspideroFlnn", "param_grid": sspidero_paras},  # Social Spider Optimization (SSpiderO)
    # {"name": "GWO-ELM", "class": "GwoFlnn", "param_grid": gwo_paras},  # Grey Wolf Optimizer (GWO)
    # {"name": "SSpiderA-ELM", "class": "SspideraFlnn", "param_grid": sspidera_paras},  # Social Spider Optimization (SSpiderO)
    # {"name": "ALO-ELM", "class": "AloFlnn", "param_grid": alo_paras},  # Ant Lion Optimizer (ALO)
    # {"name": "MFO-ELM", "class": "MfoFlnn", "param_grid": mfo_paras},  # Moth-flame optimization (MFO)
    # {"name": "EHO-ELM", "class": "EhoFlnn", "param_grid": eho_paras},  #  Elephant Herding Optimization (EHO)
    # {"name": "JA-ELM", "class": "JaFlnn", "param_grid": ja_paras},  # Jaya Algorithm (JA)
    # {"name": "WOA-ELM", "class": "WoaFlnn", "param_grid": woa_paras},  # Whale Optimization Algorithm (WOA)
    # {"name": "DO-ELM", "class": "DoFlnn", "param_grid": do_paras},  # Dragonfly Optimization (DO)
    # {"name": "BSA-ELM", "class": "BsaFlnn", "param_grid": bsa_paras},  # Bird Swarm Algorithm (BSO)
    # {"name": "SHO-ELM", "class": "ShoFlnn", "param_grid": sho_paras},  # Spotted Hyena Optimizer (SHO)
    # {"name": "SSO-ELM", "class": "SsoFlnn", "param_grid": sso_paras},  # Salp Swarm Optimization (SSO)
    # {"name": "SRSR-ELM", "class": "SrsrFlnn", "param_grid": srsr_paras},  # Swarm Robotics Search And Rescue (SRSR)
    # {"name": "GOA-ELM", "class": "GoaFlnn", "param_grid": goa_paras},  # Grasshopper Optimization Algorithm (GOA)
    # {"name": "COA-ELM", "class": "CoaFlnn", "param_grid": coa_paras},  # Coyote Optimization Algorithm (COA)
    # {"name": "MSA-ELM", "class": "MsaFlnn", "param_grid": msa_paras},  # Moth Search Algorithm (MSA)
    # {"name": "SLO-ELM", "class": "SloFlnn", "param_grid": slo_paras},  # Sea Lion Optimization Algorithm (SLO)
    # {"name": "NMRA-ELM", "class": "NmraFlnn", "param_grid": nmra_paras},  # Naked Mole-rat Algorithm (NMRA)
    # {"name": "PFA-ELM", "class": "PfaFlnn", "param_grid": pfa_paras},  # Pathfinder Algorithm (PFA)
    # {"name": "SFO-ELM", "class": "SfoFlnn", "param_grid": sfo_paras},  # Sailfish Optimizer (SFO)
    # {"name": "HHO-ELM", "class": "HhoFlnn", "param_grid": hho_paras},  # Harris Hawks Optimization (HHO)
    # {"name": "MRFO-ELM", "class": "MrfoFlnn", "param_grid": mrfo_paras},  # Manta Ray Foraging Optimization (MRFO)
    # {"name": "BES-ELM", "class": "BesFlnn", "param_grid": bes_paras},  # Bald Eagle Search (BES)
    # {"name": "SSA-ELM", "class": "SsaFlnn", "param_grid": ssa_paras},  # Sparrow Search Algorithm (SpaSA)
    # {"name": "HGS-ELM", "class": "HgsFlnn", "param_grid": hgs_paras},  # Hunger Games Search (HGS)
    # {"name": "AO-ELM", "class": "AoFlnn", "param_grid": ao_paras},  # Aquila Optimization (AO)
    #
    # ## Physics-based
    # {"name": "SA-ELM", "class": "SaFlnn", "param_grid": sa_paras},  # Simulated Annealing (SA)
    # {"name": "WDO-ELM", "class": "WdoFlnn", "param_grid": wdo_paras},  # Wind Driven Optimization (WDO)
    # {"name": "MVO-ELM", "class": "MvoFlnn", "param_grid": mvo_paras},  # Multi-Verse Optimizer (MVO)
    # {"name": "TWO-ELM", "class": "TwoFlnn", "param_grid": two_paras},  # Tug of War Optimization (TWO)
    # {"name": "EFO-ELM", "class": "EfoFlnn", "param_grid": efo_paras},  # Electromagnetic Field Optimization (EFO)
    # {"name": "NRO-ELM", "class": "NroFlnn", "param_grid": nro_paras},  # Nuclear Reaction Optimization (NRO)
    # {"name": "HGSO-ELM", "class": "HgsoFlnn", "param_grid": hgso_paras},  # Henry Gas Solubility Optimization (HGSO)
    # {"name": "ASO-ELM", "class": "AsoFlnn", "param_grid": aso_paras},  # Atom Search Optimization (ASO)
    # {"name": "EO-ELM", "class": "EoFlnn", "param_grid": eo_paras},  # Equilibrium Optimizer (EO)
    # {"name": "ArchOA-ELM", "class": "ArchoaFlnn", "param_grid": archoa_paras},  # Archimedes Optimization Algorithm (ArchOA)
    #
    # # Human-based added
    # {"name": "CA-ELM", "class": "CaFlnn", "param_grid": tlo_paras},  # Culture Algorithm (CA)
    # {"name": "ICA-ELM", "class": "IcaFlnn", "param_grid": ica_paras},  # Imperialist Competitive Algorithm (ICA)
    # {"name": "TLO-ELM", "class": "TloFlnn", "param_grid": tlo_paras},  # Teaching Learning-based Optimization (TLO)
    # {"name": "BSO-ELM", "class": "BsoFlnn", "param_grid": bso_paras},  # Brain Storm Optimization (BSO)
    # {"name": "QSA-ELM", "class": "QsaFlnn", "param_grid": qsa_paras},  # Queuing search algorithm (QSA)
    # {"name": "SARO-ELM", "class": "SaroFlnn", "param_grid": saro_paras},  # Search And Rescue Optimization (SARO)
    # {"name": "LCO-ELM", "class": "LcoFlnn", "param_grid": lco_paras},  # Life Choice-based Optimization (LCO)
    # {"name": "SSDO-ELM", "class": "SsdoFlnn", "param_grid": ssdo_paras},  # Social Ski-Driver Optimization (SSDO)
    # {"name": "GSKA-ELM", "class": "GskaFlnn", "param_grid": gska_paras},  # Gaining Sharing Knowledge-based Algorithm (GSKA)
    # {"name": "CHIO-ELM", "class": "ChioFlnn", "param_grid": chio_paras},  # Coronavirus Herd Immunity Optimization (CHIO)
    # {"name": "FBIO-ELM", "class": "FbioFlnn", "param_grid": fbio_paras},  # Forensic-Based Investigation Optimization (FBIO)
    # {"name": "BRO-ELM", "class": "BroFlnn", "param_grid": bro_paras},  # Battle Royale Optimization (BRO)
    #
    # ## Bio-based
    # {"name": "IWO-ELM", "class": "IwoFlnn", "param_grid": iwo_paras},  # Invasive weed colonization (IWO)
    # {"name": "BBO-ELM", "class": "BboFlnn", "param_grid": bbo_paras},  # Biogeography-based optimization (BBO)
    # {"name": "VCS-ELM", "class": "VcsFlnn", "param_grid": vcs_paras},  # Virus Colony Search (VCS)
    # {"name": "SBO-ELM", "class": "SboFlnn", "param_grid": sbo_paras},  # Satin Bowerbird Optimizer (SBO)
    # {"name": "EOA-ELM", "class": "EoaFlnn", "param_grid": eoa_paras},  # Earthworm Optimisation Algorithm (EOA)
    # {"name": "WHO-ELM", "class": "WhoFlnn", "param_grid": who_paras},  # Wildebeest Herd Optimization (WHO)
    # {"name": "SMA-ELM", "class": "SmaFlnn", "param_grid": sma_paras},  # Slime Mould Algorithm (SMA)
    #
    # ## System-based
    # {"name": "GCO-ELM", "class": "GcoFlnn", "param_grid": gco_paras},  # Germinal Center Optimization (GCO)
    # {"name": "WCA-ELM", "class": "WcaFlnn", "param_grid": wca_paras},  # Water Cycle Algorithm (WCA)
    # {"name": "AEO-ELM", "class": "AeoFlnn", "param_grid": aeo_paras},  # Artificial Ecosystem-based Optimization (AEO)
    #
    # ## Math-based
    # {"name": "HC-ELM", "class": "HcFlnn", "param_grid": sca_paras},  # Hill Climbing (HC)
    # {"name": "SCA-ELM", "class": "ScaFlnn", "param_grid": sca_paras},  # Sine Cosine Algorithm (SCA)
    # {"name": "GBO-ELM", "class": "GboFlnn", "param_grid": gbo_paras},  # Gradient-Based Optimizer (GBO)
    # {"name": "AOA-ELM", "class": "AoaFlnn", "param_grid": aoa_paras},  # Arithmetic Optimization Algorithm (AOA)
    #
    # {"name": "CGO-ELM", "class": "CgoFlnn", "param_grid": sca_paras},  # Chaos Game Optimization (CGO)
    # {"name": "PSS-ELM", "class": "PssFlnn", "param_grid": pss_paras},  # Pareto-like Sequential Sampling (PSS)
    #
    # ## Music-based group
    # {"name": "HS-ELM", "class": "HsFlnn", "param_grid": hs_paras},  # Harmony Search (HS)
]

models_elm = [
    # Evolutionary-based
    {"name": "GA-ELM", "class": "GaElm", "param_grid": ga_paras},  # Genetic Algorithm (GA)
    {"name": "CRO-ELM", "class": "CroElm", "param_grid": cro_paras},  # Coral Reefs Optimization (CRO)

    ## Swarm-based
    {"name": "AGTO-ELM", "class": "AgtoElm", "param_grid": agto_paras},  # Artificial Gorilla Troops Optimization (AGTO)
    {"name": "DMOA-ELM", "class": "DmoaElm", "param_grid": dmoa_paras},  # Dwarf Mongoose Optimization Algorithm (DMOA)
    {"name": "HGS-ELM", "class": "HgsElm", "param_grid": hgs_paras},  # Hunger Games Search (HGS)
    {"name": "WOA-ELM", "class": "WoaElm", "param_grid": woa_paras},  # Whale Optimization Algorithm (WOA)

    # ## Physics-based
    {"name": "NRO-ELM", "class": "NroElm", "param_grid": nro_paras},  # Nuclear Reaction Optimization (NRO)
    {"name": "HGSO-ELM", "class": "HgsoElm", "param_grid": hgso_paras},  # Henry Gas Solubility Optimization (HGSO)
    {"name": "ASO-ELM", "class": "AsoElm", "param_grid": aso_paras},  # Atom Search Optimization (ASO)

    # # Human-based added
    {"name": "GSKA-ELM", "class": "GskaElm", "param_grid": gska_paras},  # Gaining Sharing Knowledge-based Algorithm (GSKA)
    {"name": "LCO-ELM", "class": "LcoElm", "param_grid": lco_paras},  # Life Choice-based Optimization (LCO)

    # ## Bio-based
    {"name": "SMA-ELM", "class": "SmaElm", "param_grid": sma_paras},  # Slime Mould Algorithm (SMA)
    {"name": "SOA-ELM", "class": "SoaElm", "param_grid": soa_paras},  # Seagull Optimization Algorithm (SOA)
    {"name": "TSA-ELM", "class": "TsaElm", "param_grid": tsa_paras},  # Tunicate Swarm Algorithm (TSA)

    # ## System-based
    {"name": "AEO-ELM", "class": "AeoElm", "param_grid": aeo_paras},  # Artificial Ecosystem-based Optimization (AEO)

    # ## Music-based group
    {"name": "HS-ELM", "class": "HsElm", "param_grid": hs_paras},  # Harmony Search (HS)

    # ## Math-based
    {"name": "GBO-ELM", "class": "GboElm", "param_grid": gbo_paras},  # Gradient-Based Optimizer (GBO)
    {"name": "PSS-ELM", "class": "PssElm", "param_grid": pss_paras},  # Pareto-like Sequential Sampling (PSS)
    {"name": "INFO-ELM", "class": "InfoElm", "param_grid": info_paras},  # weIghted meaN oF vectOrs (INFO)
    {"name": "RUN-ELM", "class": "RunElm", "param_grid": run_paras},  # RUNge Kutta optimizer (RUN)
]


## Noted: The name can change to anything, but the class can not change.
## For example: current name is "FireflyA-CFNN" can change to "Firefly-Algorithm-CFNN"
## This name will be used in drawing figures, so shorter name make it better clear


def draw_convergence_chart_of_compared_models(compared_models):
    for obj in Config.OBJ_FUNCS:
        for act_name in Config.ELM_ACT_NAMES:
            for size_hidden in Config.ELM_HIDDEN_SIZES:
                ## Draw convergence chart of set compared model in each network and each trial
                for trial in range(0, Config.N_TRIALS):
                    pathread = f"{Config.DATA_RESULTS}/{act_name}-{size_hidden}-{obj}/trial-{trial}"
                    pathsave = f"{Config.DATA_RESULTS}/{act_name}-{size_hidden}-{obj}/{Config.FOLDER_VISUALIZE}/trial-{trial}"
                    suf_file_read = Config.FILENAME_LOSS_TRAIN
                    cols_header = Config.FILE_LOSS_HEADER
                    sur_file_save_train = "train-convergence"
                    sur_file_save_test = "valid-convergence" if Config.VALIDATION_USED else "test-convergence"
                    draw_convergence_chart_models(compared_models, pathread, pathsave, suf_file_read, cols_header, obj, sur_file_save_train, sur_file_save_test)


def draw_convergence_chart_of_all_models(models):
    for obj in Config.OBJ_FUNCS:
        for trial in range(0, Config.N_TRIALS):
            for act_name in Config.ELM_ACT_NAMES:
                for size_hidden in Config.ELM_HIDDEN_SIZES:
                    pathread = f"{Config.DATA_RESULTS}/{act_name}-{size_hidden}-{obj}/trial-{trial}"
                    pathsave = f"{Config.DATA_RESULTS}/{act_name}-{size_hidden}-{obj}/{Config.FOLDER_VISUALIZE}/trial-{trial}"
                    suf_file_read = Config.FILENAME_LOSS_TRAIN
                    cols_header = Config.FILE_LOSS_HEADER
                    validation_used = Config.VALIDATION_USED
                    draw_convergence_chart_epochs_popsizes(models, pathread, pathsave, suf_file_read, cols_header, obj, validation_used,Config.LEGEND_EPOCH)


draw_convergence_chart_of_compared_models(models_elm)

# draw_convergence_chart_of_all_models(models_elm)