forked from sydney-machine-learning/MEHH_RCPSP
-
Notifications
You must be signed in to change notification settings - Fork 0
/
params_map_elites.py
30 lines (28 loc) · 1.44 KB
/
params_map_elites.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
import os
n_runs=31
nb_features = 3 # The number of features to take into account in the container
nb_bins = [10,10,10]
features_domain = [(4, 127),(0,30),(1.65,2.00)] # The domain (min/max values) of the features
fitness_domain = [(0., 1.0)] # The domain (min/max values) of the fitness
init_batch_size = 1024 # The number of evaluations of the initial batch ('batch' = population)
batch_size = 1024 # The number of evaluations in each subsequent batch
nb_iterations = 25 # The number of iterations (i.e. times where a new batch is evaluated)
cxpb = 0.8
mutation_pb = 0.2 # The probability of mutating each value of a genome
max_items_per_bin = 1 # The number of items in each bin of the grid
verbose = True
show_warnings = True # Display warning and error messages. Set to True if you want to check if some individuals were out-of-bounds
SELECTION_POOL_SIZE=7 # Number of individuals for tournament
HEIGHT_LIMIT = 7 # Height Limit for tree
GEN_MIN_HEIGHT=2
GEN_MAX_HEIGHT=5
"""Eval mode
0: Evaluate only best individual on train set
1: Evaluate best individual on validation set
2: Evaluate all individuals on grid
"""
eval_mode=[0,1]
occupied=0 # Set number for storing results
filename="./map_elites.py.py"
with open(filename) as infile:
exec(infile.read())