2701 lines
117 KiB
Python
2701 lines
117 KiB
Python
# Karoo GP Base Class
|
|
# Define the methods and global variables used by Karoo GP
|
|
# by Kai Staats, MSc; see LICENSE.md
|
|
# Thanks to Emmanuel Dufourq and Arun Kumar for support during 2014-15 devel; TensorFlow support provided by Iurii Milovanov
|
|
# version 1.0.3
|
|
|
|
'''
|
|
A NOTE TO THE NEWBIE, EXPERT, AND BRAVE
|
|
Even if you are highly experienced in Genetic Programming, it is recommended that you review the 'Karoo User Guide' before running
|
|
this application. While your computer will not burst into flames nor will the sun collapse into a black hole if you do not, you will
|
|
likely find more enjoyment of this particular flavour of GP with a little understanding of its intent and design.
|
|
'''
|
|
|
|
import sys
|
|
import os
|
|
import csv
|
|
import time
|
|
|
|
import numpy as np
|
|
import sklearn.metrics as skm
|
|
import sklearn.cross_validation as skcv
|
|
|
|
from sympy import sympify
|
|
from datetime import datetime
|
|
from collections import OrderedDict
|
|
|
|
# TensorFlow-related imports
|
|
os.environ["TF_CPP_MIN_LOG_LEVEL"] = "1"
|
|
import tensorflow as tf
|
|
import ast
|
|
import operator as op
|
|
operators = {ast.Add: op.add, ast.Sub: op.sub, ast.Mult: op.mul, ast.Div: op.truediv, ast.Pow: op.pow, ast.BitXor: op.xor, ast.USub: op.neg}
|
|
|
|
np.set_printoptions(linewidth = 320) # set the terminal to print 320 characters before line-wrapping in order to view Trees
|
|
|
|
|
|
class Base_GP(object):
|
|
|
|
'''
|
|
This Base_BP class contains all methods for Karoo GP.
|
|
|
|
Method names are differentiated from global variable names (defined below) by the prefix 'fx_' followed by an object
|
|
and action, as in 'fx_display_tree()', with a few expections, such as 'fx_fitness_gene_pool'.
|
|
|
|
The categories (denoted by +++ banners +++) are as follows:
|
|
'karoo_gp' A single method which conducts an entire run. Employed only by karoo_gp_server.py
|
|
'fx_karoo_' Methods to Run Karoo GP
|
|
'fx_gen_' Methods to Generate a Tree
|
|
'fx_eval_' Methods to Evaluate a Tree
|
|
'fx_fitness_' Methods to Train and Test a Tree for Fitness
|
|
'fx_evolve_' Methods to Evolve a Population
|
|
'fx_display_' Methods to Display a Tree
|
|
'fx_archive_' Methods to Archive
|
|
|
|
There are no sub-classes at the time of this edit - 2015 09/21
|
|
'''
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Define Global Variables |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def __init__(self):
|
|
|
|
'''
|
|
All Karoo GP global variables are named with the prefix 'gp.' The 13 variables which begin with 'gp.pop_' are
|
|
specifically employed to define the 13 parameters for each tree as stored in the axis-1 (expand horizontally)
|
|
'gp.population' Numpy array.
|
|
|
|
### Global and local variables defined by the user in karoo_gp_main.py (in order of appearence) ###
|
|
'gp.kernel' fitness function
|
|
'gp.class_method' select the number of classes (will be automated in future version)
|
|
'tree_type' Full, Grow, or Ramped 50/50 (local variable)
|
|
'gp.tree_depth_min' minimum number of nodes
|
|
'tree_depth_base' maximum Tree depth for the initial population, where nodes = 2^(depth + 1) - 1
|
|
'gp.tree_depth_max' maximum Tree depth for the entire run; introduces potential bloat
|
|
'gp.tree_pop_max' maximum number of Trees per generation
|
|
'gp.generation_max' maximum number of generations
|
|
'gp.display' level of on-screen feedback
|
|
|
|
'gp.evolve_repro' quantity of a population generated through Reproduction
|
|
'gp.evolve_point' quantity of a population generated through Point Mutation
|
|
'gp.evolve_branch' quantity of a population generated through Branch Mutation
|
|
'gp.evolve_cross' quantity of a population generated through Crossover
|
|
|
|
'gp.tourn_size' the number of Trees chosen for each tournament
|
|
'gp.precision' the number of floating points for all applications of the round function
|
|
|
|
### Global variables used for data management ###
|
|
'gp.data_train' store train data for processing in TF
|
|
'gp.data_test' store test data for processing in TF
|
|
'gp.tf_device' set TF computation backend device (CPU or GPU)
|
|
'gp.tf_device_log' employed for TensorFlow debugging
|
|
|
|
'gp.data_train_cols' number of cols in the TRAINING data (see 'fx_karoo_data_load', below)
|
|
'gp.data_train_rows' number of rows in the TRAINING data (see 'fx_karoo_data_load', below)
|
|
'gp.data_test_cols' number of cols in the TEST data (see 'fx_karoo_data_load', below)
|
|
'gp.data_test_rows' number of rows in the TEST data (see 'fx_karoo_data_load', below)
|
|
|
|
'gp.functions' user defined functions (operators) from the associated files/[functions].csv
|
|
'gp.terminals' user defined variables (operands) from the top row of the associated [data].csv
|
|
'gp.coeff' user defined coefficients (NOT YET IN USE)
|
|
'gp.fitness_type' fitness type
|
|
'gp.datetime' date-time stamp of when the unique directory is created
|
|
'gp.path' full path to the unique directory created with each run
|
|
'gp.dataset' local path and dataset filename
|
|
|
|
### Global variables initiated and/or used by Sympy ###
|
|
'gp.algo_raw' a Sympy string which represents a flattened tree
|
|
'gp.algo_sym' a Sympy executable version of algo_raw
|
|
'gp.fittest_dict' a dictionary of the most fit trees, compiled during fitness function execution
|
|
|
|
### Variables used for evolutionary management ###
|
|
'gp.population_a' the root generation from which Trees are chosen for mutation and reproduction
|
|
'gp.population_b' the generation constructed from gp.population_a (recyled)
|
|
'gp.gene_pool' once-per-generation assessment of trees that meet min and max boundary conditions
|
|
'gp.generation_id' simple n + 1 increment
|
|
'gp.fitness_type' set in 'fx_karoo_data_load' as either a minimising or maximising function
|
|
'gp.tree' axis-1, 13 element Numpy array that defines each Tree, stored in 'gp.population'
|
|
'gp.pop_*' 13 elements which define each Tree (see 'fx_gen_tree_initialise' below)
|
|
|
|
### Fishing nets ###
|
|
You can insert a "fishing net" to search for a specific expression when you fear the evolutionary process or
|
|
something in the code may not be working. Search for "fishing net" and follow the directions.
|
|
|
|
### Error checks ###
|
|
You can quickly find all places in which error checks have been inserted by searching for "ERROR!"
|
|
'''
|
|
|
|
self.algo_raw = [] # temp store the raw expression -- CONSIDER MAKING THIS VARIABLE LOCAL
|
|
self.algo_sym = [] # temp store the sympified expression-- CONSIDER MAKING THIS VARIABLE LOCAL
|
|
self.fittest_dict = {} # temp store all Trees which share the best fitness score
|
|
self.gene_pool = [] # temp store all Tree IDs for use by Tournament
|
|
self.class_labels = 0 # temp set a variable which will be assigned the number of class labels (data_y)
|
|
|
|
return
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Run Karoo GP |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def karoo_gp(self, tree_type, tree_depth_base, filename):
|
|
|
|
'''
|
|
This method enables the engagement of the entire Karoo GP application. It is used exclusively by the server script
|
|
karoo_gp_server.py (not by the desktop script karoo_gp_main.py). Instead of returning the user to the pause menu,
|
|
this script terminates at the command-line, providing support for bash and chron job execution.
|
|
|
|
Arguments required: tree_type, tree_depth_base, filename
|
|
'''
|
|
|
|
self.karoo_banner()
|
|
start = time.time() # start the clock for the timer
|
|
|
|
# construct first generation of Trees
|
|
self.fx_karoo_data_load(tree_type, tree_depth_base, filename)
|
|
self.generation_id = 1 # set initial generation ID
|
|
self.population_a = ['Karoo GP by Kai Staats, Generation ' + str(self.generation_id)] # list to store all Tree arrays, one generation at a time
|
|
self.fx_karoo_construct(tree_type, tree_depth_base) # construct the first population of Trees
|
|
|
|
# evaluate first generation of Trees
|
|
print '\n Evaluate the first generation of Trees ...'
|
|
self.fx_fitness_gym(self.population_a) # generate expression, evaluate fitness, compare fitness
|
|
self.fx_archive_tree_write(self.population_a, 'a') # save the first generation of Trees to disk
|
|
|
|
# evolve subsequent generations of Trees
|
|
for self.generation_id in range(2, self.generation_max + 1): # loop through 'generation_max'
|
|
|
|
print '\n Evolve a population of Trees for Generation', self.generation_id, '...'
|
|
self.population_b = ['Karoo GP by Kai Staats, Evolving Generation'] # initialise population_b to host the next generation
|
|
|
|
self.fx_fitness_gene_pool() # generate the viable gene pool (compares against gp.tree_depth_min)
|
|
self.fx_karoo_reproduce() # method 1 - Reproduction
|
|
self.fx_karoo_point_mutate() # method 2 - Point Mutation
|
|
self.fx_karoo_branch_mutate() # method 3 - Branch Mutation
|
|
self.fx_karoo_crossover() # method 4 - Crossover
|
|
self.fx_eval_generation() # evaluate all Trees in a single generation
|
|
|
|
self.population_a = self.fx_evolve_pop_copy(self.population_b, ['GP Tree by Kai Staats, Generation ' + str(self.generation_id)])
|
|
|
|
# "End of line, man!" --CLU
|
|
print '\n \033[36m Karoo GP has an ellapsed time of \033[0;0m\033[31m%f\033[0;0m' % (time.time() - start), '\033[0;0m'
|
|
self.fx_archive_tree_write(self.population_b, 'f') # save the final generation of Trees to disk
|
|
self.fx_archive_params_write('Server') # save run-time parameters to disk
|
|
|
|
print '\n \033[3m Congrats!\033[0;0m Your multi-generational Karoo GP run is complete.\n'
|
|
sys.exit() # return Karoo GP to the command line to support bash and chron job execution
|
|
|
|
# return
|
|
|
|
|
|
def karoo_banner(self):
|
|
|
|
'''
|
|
This method makes Karoo GP look old-school cool!
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
os.system('clear')
|
|
|
|
print '\n\033[36m\033[1m'
|
|
print '\t ** ** ****** ***** ****** ****** ****** ******'
|
|
print '\t ** ** ** ** ** ** ** ** ** ** ** ** **'
|
|
print '\t ** ** ** ** ** ** ** ** ** ** ** ** **'
|
|
print '\t **** ******** ****** ** ** ** ** ** *** ******'
|
|
print '\t ** ** ** ** ** ** ** ** ** ** ** ** **'
|
|
print '\t ** ** ** ** ** ** ** ** ** ** ** ** **'
|
|
print '\t ** ** ** ** ** ** ** ** ** ** ** ** **'
|
|
print '\t ** ** ** ** ** ** ****** ****** ****** **'
|
|
print '\033[0;0m'
|
|
print '\t\033[36m Genetic Programming in Python - by Kai Staats, version 1.0\033[0;0m'
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_data_load(self, tree_type, tree_depth_base, filename):
|
|
|
|
'''
|
|
The data and function .csv files are loaded according to the fitness function kernel selected by the user. An
|
|
alternative dataset may be loaded at launch, by appending a command line argument. The data is then split into
|
|
both TRAINING and TEST segments in order to validate the success of the GP training run. Datasets less than
|
|
10 rows will not be split, rather copied in full to both TRAINING and TEST as it is assumed you are conducting
|
|
a system validation run, as with the built-in MATCH kernel and associated dataset.
|
|
|
|
Arguments required: tree_type, tree_depth_base, filename (of the dataset)
|
|
'''
|
|
|
|
### 1) load the associated data set, operators, operands, fitness type, and coefficients ###
|
|
|
|
full_path = os.path.realpath(__file__); cwd = os.path.dirname(full_path) # Good idea Marco :)
|
|
# cwd = os.getcwd()
|
|
|
|
data_dict = {'c':cwd + '/files/data_CLASSIFY.csv', 'r':cwd + '/files/data_REGRESS.csv', 'm':cwd + '/files/data_MATCH.csv', 'p':cwd + '/files/data_PLAY.csv'}
|
|
|
|
if len(sys.argv) == 1: # load data from the default karoo_gp/files/ directory
|
|
data_x = np.loadtxt(data_dict[self.kernel], skiprows = 1, delimiter = ',', dtype = float); data_x = data_x[:,0:-1] # load all but the right-most column
|
|
data_y = np.loadtxt(data_dict[self.kernel], skiprows = 1, usecols = (-1,), delimiter = ',', dtype = float) # load only right-most column (class labels)
|
|
header = open(data_dict[self.kernel],'r')
|
|
self.dataset = data_dict[self.kernel]
|
|
|
|
elif len(sys.argv) == 2: # load an external data file
|
|
data_x = np.loadtxt(sys.argv[1], skiprows = 1, delimiter = ',', dtype = float); data_x = data_x[:,0:-1] # load all but the right-most column
|
|
data_y = np.loadtxt(sys.argv[1], skiprows = 1, usecols = (-1,), delimiter = ',', dtype = float) # load only right-most column (class labels)
|
|
header = open(sys.argv[1],'r')
|
|
self.dataset = sys.argv[1]
|
|
|
|
elif len(sys.argv) > 2: # receive filename and additional flags from karoo_gp_server.py via argparse
|
|
data_x = np.loadtxt(filename, skiprows = 1, delimiter = ',', dtype = float); data_x = data_x[:,0:-1] # load all but the right-most column
|
|
data_y = np.loadtxt(filename, skiprows = 1, usecols = (-1,), delimiter = ',', dtype = float) # load only right-most column (class labels)
|
|
header = open(filename,'r')
|
|
self.dataset = filename
|
|
|
|
fitt_dict = {'c':'max', 'r':'min', 'm':'max', 'p':''}
|
|
self.fitness_type = fitt_dict[self.kernel] # load fitness type
|
|
|
|
func_dict = {'c':cwd + '/files/operators_CLASSIFY.csv', 'r':cwd + '/files/operators_REGRESS.csv', 'm':cwd + '/files/operators_MATCH.csv', 'p':cwd + '/files/operators_PLAY.csv'}
|
|
self.functions = np.loadtxt(func_dict[self.kernel], delimiter=',', skiprows=1, dtype = str) # load the user defined functions (operators)
|
|
self.terminals = header.readline().split(','); self.terminals[-1] = self.terminals[-1].replace('\n','') # load the user defined terminals (operands)
|
|
self.class_labels = len(np.unique(data_y)) # load the user defined labels for classification or solutions for regression
|
|
self.coeff = np.loadtxt(cwd + '/files/coefficients.csv', delimiter=',', skiprows=1, dtype = str) # load the user defined coefficients - NOT USED YET
|
|
|
|
|
|
### 2) from the dataset, extract TRAINING and TEST data ###
|
|
|
|
if len(data_x) < 11: # for small datasets we will not split them into TRAINING and TEST components
|
|
data_train = np.c_[data_x, data_y]
|
|
data_test = np.c_[data_x, data_y]
|
|
|
|
else: # if larger than 10, we run the data through the SciKit Learn's 'random split' function
|
|
x_train, x_test, y_train, y_test = skcv.train_test_split(data_x, data_y, test_size = 0.2) # 80/20 TRAIN/TEST split
|
|
data_x, data_y = [], [] # clear from memory
|
|
|
|
data_train = np.c_[x_train, y_train] # recombine each row of data with its associated label (right column)
|
|
x_train, y_train = [], [] # clear from memory
|
|
|
|
data_test = np.c_[x_test, y_test] # recombine each row of data with its associated label (right column)
|
|
x_test, y_test = [], [] # clear from memory
|
|
|
|
self.data_train_cols = len(data_train[0,:]) # qty count
|
|
self.data_train_rows = len(data_train[:,0]) # qty count
|
|
self.data_test_cols = len(data_test[0,:]) # qty count
|
|
self.data_test_rows = len(data_test[:,0]) # qty count
|
|
|
|
|
|
### 3) load TRAINING and TEST data for TensorFlow processing - tested 2017 02/02
|
|
|
|
self.data_train = data_train # Store train data for processing in TF
|
|
self.data_test = data_test # Store test data for processing in TF
|
|
self.tf_device = "/gpu:0" # Set TF computation backend device (CPU or GPU)
|
|
self.tf_device_log = False # TF device usage logging (for debugging)
|
|
|
|
|
|
### 4) create a unique directory and initialise all .csv files ###
|
|
|
|
# self.datetime = datetime.datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
|
|
self.datetime = datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
|
|
self.path = os.path.join(cwd, 'runs/', self.datetime) # generate a unique directory name
|
|
if not os.path.isdir(self.path): os.makedirs(self.path) # make a unique directory
|
|
|
|
self.filename = {} # a dictionary to hold .csv filenames
|
|
|
|
self.filename.update( {'a':self.path + '/population_a.csv'} )
|
|
target = open(self.filename['a'], 'w') # initialise the .csv file for population 'a' (foundation)
|
|
target.close()
|
|
|
|
self.filename.update( {'b':self.path + '/population_b.csv'} )
|
|
target = open(self.filename['b'], 'w') # initialise the .csv file for population 'b' (evolving)
|
|
target.close()
|
|
|
|
self.filename.update( {'f':self.path + '/population_f.csv'} )
|
|
target = open(self.filename['f'], 'w') # initialise the .csv file for the final population (test)
|
|
target.close()
|
|
|
|
self.filename.update( {'s':self.path + '/population_s.csv'} )
|
|
# do NOT initialise this .csv file, as it is retained for loading a previous run (recover)
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_data_recover(self, population):
|
|
|
|
'''
|
|
This method is used to load a saved population of Trees, as invoked through the (pause) menu where population_s
|
|
replaces population_a in the /[path]/karoo_gp/runs/ directory.
|
|
|
|
Arguments required: population size
|
|
'''
|
|
|
|
with open(population, 'rb') as csv_file:
|
|
target = csv.reader(csv_file, delimiter=',')
|
|
n = 0 # track row count
|
|
|
|
for row in target:
|
|
|
|
n = n + 1
|
|
if n == 1: pass # skip first empty row
|
|
|
|
elif n == 2:
|
|
self.population_a = [row] # write header to population_a
|
|
|
|
else:
|
|
if row == []:
|
|
self.tree = np.array([[]]) # initialise Tree array
|
|
|
|
else:
|
|
if self.tree.shape[1] == 0:
|
|
self.tree = np.append(self.tree, [row], axis = 1) # append first row to Tree
|
|
|
|
else:
|
|
self.tree = np.append(self.tree, [row], axis = 0) # append subsequent rows to Tree
|
|
|
|
if self.tree.shape[0] == 13:
|
|
self.population_a.append(self.tree) # append complete Tree to population list
|
|
|
|
print self.population_a
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_construct(self, tree_type, tree_depth_base):
|
|
|
|
'''
|
|
As used by the method 'karoo_gp', this method constructs the initial population based upon the user-defined
|
|
Tree type and initial, maximum Tree depth ('tree_depth_base'). "Ramped half/half" was defined by John Koza as
|
|
a means of building maximum diversity in the initial population. There are equal numbers of Full and Grow
|
|
methods trees, and an equal spread of Trees across depths 1 to 'tree_depth_base'.
|
|
|
|
Arguments required: tree_type, tree_depth_base
|
|
'''
|
|
|
|
if self.display == 'i' or self.display == 'g':
|
|
print '\n\t Type \033[1m?\033[0;0m at any (pause) to review your options, or \033[1mENTER\033[0;0m to continue.\033[0;0m'
|
|
self.fx_karoo_pause(0)
|
|
|
|
if tree_type == 'r': # Ramped 50/50
|
|
|
|
TREE_ID = 1
|
|
for n in range(1, int((self.tree_pop_max / 2) / tree_depth_base) + 1): # split the population into equal parts
|
|
for depth in range(1, tree_depth_base + 1): # build 2 Trees ats each depth
|
|
self.fx_gen_tree_build(TREE_ID, 'f', depth) # build a Full Tree
|
|
self.fx_archive_tree_append(self.tree) # append Tree to the list 'gp.population_a'
|
|
TREE_ID = TREE_ID + 1
|
|
|
|
self.fx_gen_tree_build(TREE_ID, 'g', depth) # build a Grow Tree
|
|
self.fx_archive_tree_append(self.tree) # append Tree to the list 'gp.population_a'
|
|
TREE_ID = TREE_ID + 1
|
|
|
|
if TREE_ID < self.tree_pop_max: # eg: split 100 by 2*3 and it will produce only 96 Trees ...
|
|
for n in range(self.tree_pop_max - TREE_ID + 1): # ... so we complete the run
|
|
self.fx_gen_tree_build(TREE_ID, 'g', tree_depth_base)
|
|
self.fx_archive_tree_append(self.tree)
|
|
TREE_ID = TREE_ID + 1
|
|
|
|
else: pass
|
|
|
|
else: # Full or Grow
|
|
for TREE_ID in range(1, self.tree_pop_max + 1):
|
|
self.fx_gen_tree_build(TREE_ID, tree_type, tree_depth_base) # build the 1st generation of Trees
|
|
self.fx_archive_tree_append(self.tree)
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_reproduce(self):
|
|
|
|
'''
|
|
Through tournament selection, a single Tree from the prior generation is copied without mutation to the next
|
|
generation. This is analogous to a member of the prior generation directly entering the gene pool of the
|
|
subsequent (younger) generation.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
if self.display != 's':
|
|
if self.display == 'i': print ''
|
|
print ' Perform', self.evolve_repro, 'Reproductions ...'
|
|
if self.display == 'i': self.fx_karoo_pause(0)
|
|
|
|
for n in range(self.evolve_repro): # quantity of Trees to be copied without mutation
|
|
tourn_winner = self.fx_fitness_tournament(self.tourn_size) # perform tournament selection for each reproduction
|
|
tourn_winner = self.fx_evolve_fitness_wipe(tourn_winner) # wipe fitness data
|
|
self.population_b.append(tourn_winner) # append array to next generation population of Trees
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_point_mutate(self):
|
|
|
|
'''
|
|
Through tournament selection, a copy of a Tree from the prior generation mutates before being added to the
|
|
next generation. In this method, a single point is selected for mutation while maintaining function nodes as
|
|
functions (operators) and terminal nodes as terminals (variables). The size and shape of the Tree will remain
|
|
identical.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
if self.display != 's':
|
|
if self.display == 'i': print ''
|
|
print ' Perform', self.evolve_point, 'Point Mutations ...'
|
|
if self.display == 'i': self.fx_karoo_pause(0)
|
|
|
|
for n in range(self.evolve_point): # quantity of Trees to be generated through mutation
|
|
tourn_winner = self.fx_fitness_tournament(self.tourn_size) # perform tournament selection for each mutation
|
|
tourn_winner, node = self.fx_evolve_point_mutate(tourn_winner) # perform point mutation; return single point for record keeping
|
|
self.population_b.append(tourn_winner) # append array to next generation population of Trees
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_branch_mutate(self):
|
|
|
|
'''
|
|
Through tournament selection, a copy of a Tree from the prior generation mutates before being added to the
|
|
next generation. Unlike Point Mutation, in this method an entire branch is selected. If the evolutionary run is
|
|
designated as Full, the size and shape of the Tree will remain identical, each node mutated sequentially, where
|
|
functions remain functions and terminals remain terminals. If the evolutionary run is designated as Grow or
|
|
Ramped Half/Half, the size and shape of the Tree may grow smaller or larger, but it may not exceed
|
|
tree_depth_max as defined by the user.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
if self.display != 's':
|
|
if self.display == 'i': print ''
|
|
print ' Perform', self.evolve_branch, 'Full or Grow Mutations ...'
|
|
if self.display == 'i': self.fx_karoo_pause(0)
|
|
|
|
for n in range(self.evolve_branch): # quantity of Trees to be generated through mutation
|
|
tourn_winner = self.fx_fitness_tournament(self.tourn_size) # perform tournament selection for each mutation
|
|
branch = self.fx_evolve_branch_select(tourn_winner) # select point of mutation and all nodes beneath
|
|
|
|
# TEST & DEBUG: comment the top or bottom to force all Full or all Grow methods
|
|
|
|
if tourn_winner[1][1] == 'f': # perform Full method mutation on 'tourn_winner'
|
|
tourn_winner = self.fx_evolve_full_mutate(tourn_winner, branch)
|
|
|
|
elif tourn_winner[1][1] == 'g': # perform Grow method mutation on 'tourn_winner'
|
|
tourn_winner = self.fx_evolve_grow_mutate(tourn_winner, branch)
|
|
|
|
self.population_b.append(tourn_winner) # append array to next generation population of Trees
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_crossover(self):
|
|
|
|
'''
|
|
Through tournament selection, two trees are selected as parents to produce two offspring. Within each parent
|
|
Tree a branch is selected. Parent A is copied, with its selected branch deleted. Parent B's branch is then
|
|
copied to the former location of Parent A's branch and inserted (grafted). The size and shape of the child
|
|
Tree may be smaller or larger than either of the parents, but may not exceed 'tree_depth_max' as defined
|
|
by the user.
|
|
|
|
This process combines genetic code from two parent Trees, both of which were chosen by the tournament process
|
|
as having a higher fitness than the average population. Therefore, there is a chance their offspring will
|
|
provide an improvement in total fitness. In most GP applications, Crossover is the most commonly applied
|
|
evolutionary operator (~70-80%).
|
|
|
|
For those who like to watch, select 'db' (debug mode) at the launch of Karoo GP or at any (pause).
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
if self.display != 's':
|
|
if self.display == 'i': print ''
|
|
print ' Perform', self.evolve_cross, 'Crossovers ...'
|
|
if self.display == 'i': self.fx_karoo_pause(0)
|
|
|
|
for n in range(self.evolve_cross / 2): # quantity of Trees to be generated through Crossover, accounting for 2 children each
|
|
|
|
parent_a = self.fx_fitness_tournament(self.tourn_size) # perform tournament selection for 'parent_a'
|
|
branch_a = self.fx_evolve_branch_select(parent_a) # select branch within 'parent_a', to copy to 'parent_b' and receive a branch from 'parent_b'
|
|
|
|
parent_b = self.fx_fitness_tournament(self.tourn_size) # perform tournament selection for 'parent_b'
|
|
branch_b = self.fx_evolve_branch_select(parent_b) # select branch within 'parent_b', to copy to 'parent_a' and receive a branch from 'parent_a'
|
|
|
|
parent_c = np.copy(parent_a); branch_c = np.copy(branch_a) # else the Crossover mods affect the parent Trees, due to how Python manages '='
|
|
parent_d = np.copy(parent_b); branch_d = np.copy(branch_b) # else the Crossover mods affect the parent Trees, due to how Python manages '='
|
|
|
|
offspring_1 = self.fx_evolve_crossover(parent_a, branch_a, parent_b, branch_b) # perform Crossover
|
|
self.population_b.append(offspring_1) # append the 1st child to next generation of Trees
|
|
|
|
offspring_2 = self.fx_evolve_crossover(parent_d, branch_d, parent_c, branch_c) # perform Crossover
|
|
self.population_b.append(offspring_2) # append the 2nd child to next generation of Trees
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_pause(self, eol):
|
|
|
|
'''
|
|
Pause the program execution and output to screen until the user selects a valid option. The "eol" parameter
|
|
instructs this method to display a different screen for run-time or end-of-line, and to dive back into the
|
|
current run, or do nothing, accordingly.
|
|
|
|
Arguments required: eol
|
|
'''
|
|
|
|
options = ['?','help','i','m','g','s','db','ts','min','max','bal','id','pop','dir','l','p','t','cont','load','w','q','']
|
|
|
|
while True:
|
|
try:
|
|
pause = raw_input('\n\t\033[36m (pause) \033[0;0m')
|
|
if pause not in options: raise ValueError()
|
|
if pause == '':
|
|
if eol == 1: self.fx_karoo_pause(1) # return to pause menu as the GP run is complete
|
|
else: break # drop back into the current GP run
|
|
|
|
if pause == '?' or pause == 'help':
|
|
print '\n\t\033[36mSelect one of the following options:\033[0;0m'
|
|
print '\t\033[36m\033[1m i \t\033[0;0m Interactive display mode'
|
|
print '\t\033[36m\033[1m m \t\033[0;0m Minimal display mode'
|
|
print '\t\033[36m\033[1m g \t\033[0;0m Generation display mode'
|
|
print '\t\033[36m\033[1m s \t\033[0;0m Silent display mode'
|
|
print '\t\033[36m\033[1m db \t\033[0;0m De-Bug display mode'
|
|
print ''
|
|
print '\t\033[36m\033[1m ts \t\033[0;0m adjust the tournament size'
|
|
print '\t\033[36m\033[1m min \t\033[0;0m adjust the minimum number of nodes'
|
|
# print '\t\033[36m\033[1m max \t\033[0;0m adjust the maximum Tree depth'
|
|
print '\t\033[36m\033[1m bal \t\033[0;0m adjust the balance of genetic operators'
|
|
print ''
|
|
print '\t\033[36m\033[1m l \t\033[0;0m list Trees with leading fitness scores'
|
|
print '\t\033[36m\033[1m t \t\033[0;0m evaluate a single Tree against the test data'
|
|
print ''
|
|
print '\t\033[36m\033[1m p \t\033[0;0m print a single Tree to screen'
|
|
print '\t\033[36m\033[1m id \t\033[0;0m display the current generation ID'
|
|
print '\t\033[36m\033[1m pop \t\033[0;0m list all Trees in current population'
|
|
print '\t\033[36m\033[1m dir \t\033[0;0m display current working directory'
|
|
print ''
|
|
print '\t\033[36m\033[1m cont \t\033[0;0m continue evolution, starting with the current population'
|
|
print '\t\033[36m\033[1m load \t\033[0;0m load population_s (seed) to replace population_a (current)'
|
|
print '\t\033[36m\033[1m w \t\033[0;0m write the evolving population_b to disk'
|
|
print '\t\033[36m\033[1m q \t\033[0;0m quit Karoo GP without saving population_b'
|
|
print ''
|
|
|
|
if eol == 1: print '\t\033[0;0m Remember to archive your final population before your next run!'
|
|
else: print '\t\033[36m\033[1m ENTER\033[0;0m to continue ...'
|
|
|
|
elif pause == 'i': self.display = 'i'; print '\t Interactive display mode engaged (for control freaks)'
|
|
elif pause == 'm': self.display = 'm'; print '\t Minimal display mode engaged (for recovering control freaks)'
|
|
elif pause == 'g': self.display = 'g'; print '\t Generation display mode engaged (for GP gurus)'
|
|
elif pause == 's': self.display = 's'; print '\t Silent display mode engaged (for zen masters)'
|
|
elif pause == 'db': self.display = 'db'; print '\t De-Bug display mode engaged (for vouyers)'
|
|
|
|
|
|
elif pause == 'ts': # adjust the tournament size
|
|
menu = range(2,self.tree_pop_max + 1) # set to total population size only for the sake of experimentation
|
|
while True:
|
|
try:
|
|
print '\n\t The current tournament size is:', self.tourn_size
|
|
query = int(raw_input('\t Adjust the tournament size (suggest 10): '))
|
|
if query not in menu: raise ValueError()
|
|
self.tourn_size = query; break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 2 including', str(self.tree_pop_max) + ".", 'Try again ...\033[0;0m'
|
|
|
|
|
|
elif pause == 'min': # adjust the global, minimum number of nodes per Tree
|
|
menu = range(3,1001) # we must have at least 3 nodes, as in: x * y; 1000 is an arbitrary number
|
|
while True:
|
|
try:
|
|
print '\n\t The current minimum number of nodes is:', self.tree_depth_min
|
|
query = int(raw_input('\t Adjust the minimum number of nodes for all Trees (min 3): '))
|
|
if query not in menu: raise ValueError()
|
|
self.tree_depth_min = query; break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 3 including 1000. Try again ...\033[0;0m'
|
|
|
|
|
|
# NEED TO ADD: adjustable tree_depth_max
|
|
#elif pause == 'max': # adjust the global, adjusted maximum Tree depth
|
|
#
|
|
# menu = range(1,11)
|
|
# while True:
|
|
# try:
|
|
# print '\n\t The current \033[3madjusted\033[0;0m maximum Tree depth is:', self.tree_depth_max
|
|
# query = int(raw_input('\n\t Adjust the global maximum Tree depth to (1 ... 10): '))
|
|
# if query not in menu: raise ValueError()
|
|
# if query < self.tree_depth_max:
|
|
# print '\n\t\033[32m This value is less than the current value.\033[0;0m'
|
|
# conf = raw_input('\n\t Are you ok with this? (y/n) ')
|
|
# if conf == 'n': break
|
|
# except ValueError: print '\n\t\033[32m Enter a number from 1 including 10. Try again ...\033[0;0m'
|
|
|
|
|
|
elif pause == 'bal': # adjust the balance of genetic operators'
|
|
print '\n\t The current balance of genetic operators is:'
|
|
print '\t\t Reproduction:', self.evolve_repro; tmp_repro = self.evolve_repro
|
|
print '\t\t Point Mutation:', self.evolve_point; tmp_point = self.evolve_point
|
|
print '\t\t Branch Mutation:', self.evolve_branch; tmp_branch = self.evolve_branch
|
|
print '\t\t Crossover:', self.evolve_cross, '\n'; tmp_cross = self.evolve_cross
|
|
|
|
menu = range(0,1000) # 0 to 1000 expresssed as an integer
|
|
|
|
while True:
|
|
try:
|
|
query = raw_input('\t Enter quantity of Trees to be generated by Reproduction: ')
|
|
if query not in str(menu): raise ValueError()
|
|
elif query == '': break
|
|
tmp_repro = int(float(query)); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 0 including 1000. Try again ...\033[0;0m'
|
|
|
|
while True:
|
|
try:
|
|
query = raw_input('\t Enter quantity of Trees to be generated by Point Mutation: ')
|
|
if query not in str(menu): raise ValueError()
|
|
elif query == '': break
|
|
tmp_point = int(float(query)); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 0 including 1000. Try again ...\033[0;0m'
|
|
|
|
while True:
|
|
try:
|
|
query = raw_input('\t Enter quantity of Trees to be generated by Branch Mutation: ')
|
|
if query not in str(menu): raise ValueError()
|
|
elif query == '': break
|
|
tmp_branch = int(float(query)); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 0 including 1000. Try again ...\033[0;0m'
|
|
|
|
while True:
|
|
try:
|
|
query = raw_input('\t Enter quantity of Trees to be generated by Crossover: ')
|
|
if query not in str(menu): raise ValueError()
|
|
elif query == '': break
|
|
tmp_cross = int(float(query)); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 0 including 1000. Try again ...\033[0;0m'
|
|
|
|
if tmp_repro + tmp_point + tmp_branch + tmp_cross != self.tree_pop_max: print '\n\t The sum of the above does not equal', self.tree_pop_max, 'Try again ...'
|
|
else:
|
|
print '\n\t The revised balance of genetic operators is:'
|
|
self.evolve_repro = tmp_repro; print '\t\t Reproduction:', self.evolve_repro
|
|
self.evolve_point = tmp_point; print '\t\t Point Mutation:', self.evolve_point
|
|
self.evolve_branch = tmp_branch; print '\t\t Branch Mutation:', self.evolve_branch
|
|
self.evolve_cross = tmp_cross; print '\t\t Crossover:', self.evolve_cross
|
|
|
|
|
|
elif pause == 'l': # display dictionary of Trees with the best fitness score
|
|
print '\n\t The leading Trees and their associated expressions are:'
|
|
for n in sorted(self.fittest_dict): print '\t ', n, ':', self.fittest_dict[n]
|
|
|
|
|
|
elif pause == 't': # evaluate a Tree against the TEST data
|
|
if self.generation_id > 1:
|
|
menu = range(1, len(self.population_b))
|
|
while True:
|
|
try:
|
|
query = raw_input('\n\t Select a Tree in population_b to test: ')
|
|
if query not in str(menu) or query == '0': raise ValueError()
|
|
elif query == '': break
|
|
|
|
self.fx_eval_poly(self.population_b[int(query)]) # generate the raw and sympified equation for the given Tree using SymPy
|
|
|
|
# get simplified expression and process it by TF - tested 2017 02/02
|
|
expr = str(self.algo_sym) # might change this to algo_raw for more correct expression evaluation
|
|
result = self.fx_fitness_eval(expr, self.data_test, get_labels=True)
|
|
|
|
print '\n\t\033[36mTree', query, 'yields (raw):', self.algo_raw, '\033[0;0m'
|
|
print '\t\033[36mTree', query, 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m\n'
|
|
|
|
# test user selected Trees using TF - tested 2017 02/02
|
|
if self.kernel == 'c': self.fx_fitness_test_classify(result); break
|
|
elif self.kernel == 'r': self.fx_fitness_test_regress(result); break
|
|
elif self.kernel == 'm': self.fx_fitness_test_match(result); break
|
|
# elif self.kernel == '[other]': self.fx_fitness_test_[other](result); break
|
|
|
|
except ValueError: print '\n\t\033[32m Enter a number from 1 including', str(len(self.population_b) - 1) + ".", 'Try again ...\033[0;0m'
|
|
|
|
else: print '\n\t\033[32m Karoo GP does not enable evaluation of the foundation population. Be patient ...\033[0;0m'
|
|
|
|
|
|
elif pause == 'p': # print a Tree to screen -- NEED TO ADD: SymPy graphical print option
|
|
if self.generation_id == 1:
|
|
menu = range(1,len(self.population_a))
|
|
while True:
|
|
try:
|
|
query = raw_input('\n\t Select a Tree to print: ')
|
|
if query not in str(menu) or query == '0': raise ValueError()
|
|
elif query == '': break
|
|
self.fx_display_tree(self.population_a[int(query)]); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 1 including', str(len(self.population_a) - 1) + ".", 'Try again ...\033[0;0m'
|
|
|
|
elif self.generation_id > 1:
|
|
menu = range(1,len(self.population_b))
|
|
while True:
|
|
try:
|
|
query = raw_input('\n\t Select a Tree to print: ')
|
|
if query not in str(menu) or query == '0': raise ValueError()
|
|
elif query == '': break
|
|
self.fx_display_tree(self.population_b[int(query)]); break
|
|
except ValueError: print '\n\t\033[32m Enter a number from 1 including', str(len(self.population_b) - 1) + ".", 'Try again ...\033[0;0m'
|
|
|
|
else: print '\n\t\033[36m There is nor forest for which to see the Trees.\033[0;0m'
|
|
|
|
|
|
elif pause == 'id': print '\n\t The current generation is:', self.generation_id
|
|
|
|
|
|
elif pause == 'pop': # list Trees in the current population
|
|
print ''
|
|
if self.generation_id == 1:
|
|
for tree_id in range(1, len(self.population_a)):
|
|
self.fx_eval_poly(self.population_a[tree_id]) # extract the expression
|
|
print '\t\033[36m Tree', self.population_a[tree_id][0][1], 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m'
|
|
|
|
elif self.generation_id > 1:
|
|
for tree_id in range(1, len(self.population_b)):
|
|
self.fx_eval_poly(self.population_b[tree_id]) # extract the expression
|
|
print '\t\033[36m Tree', self.population_b[tree_id][0][1], 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m'
|
|
|
|
else: print '\n\t\033[36m There is nor forest for which to see the Trees.\033[0;0m'
|
|
|
|
|
|
elif pause == 'dir': print '\n\t The current working directory is:', self.path
|
|
|
|
|
|
elif pause == 'cont': # continue evolution, starting with the current population
|
|
menu = range(1,101)
|
|
while True:
|
|
try:
|
|
query = raw_input('\n\t How many more generations would you like to add? (1-100): ')
|
|
if query not in str(menu) or query == '0': raise ValueError()
|
|
elif query == '': break
|
|
self.generation_max = self.generation_max + int(query)
|
|
next_gen_start = self.generation_id + 1
|
|
self.fx_karoo_continue(next_gen_start) # continue evolving, starting with the last population
|
|
except ValueError: print '\n\t\033[32m Enter a number from 1 including 100. Try again ...\033[0;0m'
|
|
|
|
|
|
elif pause == 'load': # load population_s to replace population_a
|
|
while True:
|
|
try:
|
|
query = raw_input('\n\t Overwrite the current population with population_s? ')
|
|
if query not in ['y','n']: raise ValueError()
|
|
if query == 'y': self.fx_karoo_data_recover(self.filename['s']); break
|
|
elif query == 'n': break
|
|
except ValueError: print '\n\t\033[32m Enter (y)es or (n)o. Try again ...\033[0;0m'
|
|
|
|
|
|
elif pause == 'w': # write the evolving population_b to disk
|
|
if self.generation_id > 1:
|
|
self.fx_archive_tree_write(self.population_b, 'b')
|
|
print '\t\033[36m All current members of the evolving population_b saved to .csv\033[0;0m'
|
|
|
|
else: print '\n\t\033[36m The evolving population_b does not yet exist\033[0;0m'
|
|
|
|
|
|
elif pause == 'q':
|
|
if eol == 0: # if the GP run is not at the final generation
|
|
query = raw_input('\n\t \033[32mThe current population_b will be lost!\033[0;0m\n\n\t Are you certain you want to quit? (y/n)')
|
|
if query == 'y':
|
|
self.fx_archive_params_write('Desktop') # save run-time parameters to disk
|
|
sys.exit() # quit the script without saving population_b
|
|
else: break
|
|
|
|
else: # if the GP run is complete
|
|
query = raw_input('\n\t Are you certain you want to quit? (y/n)')
|
|
if query == 'y':
|
|
print '\n\t \033[32mYour Trees and runtime parameters are archived in karoo_gp/runs/\033[0;0m'
|
|
self.fx_archive_params_write('Desktop') # save run-time parameters to disk
|
|
sys.exit()
|
|
else: self.fx_karoo_pause(1)
|
|
|
|
except ValueError: print '\t\033[32m Select from the options given. Try again ...\033[0;0m'
|
|
except KeyboardInterrupt: print '\n\t\033[32m Enter q to quit\033[0;0m'
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_continue(self, next_gen_start):
|
|
|
|
'''
|
|
This method enables the launch of another full run of Karoo GP, but starting with a seed generation
|
|
instead of with a randomly generated first population. This can be used at the end of a standard run to
|
|
continue the evoluationary process, or after having recovered a set of trees from a prior run.
|
|
|
|
Arguments required: next_gen_start
|
|
'''
|
|
|
|
for self.generation_id in range(next_gen_start, self.generation_max + 1): # evolve additional generations of Trees
|
|
|
|
print '\n Evolve a population of Trees for Generation', self.generation_id, '...'
|
|
self.population_b = ['Karoo GP by Kai Staats, Evolving Generation'] # initialise population_b to host the next generation
|
|
|
|
self.fx_fitness_gene_pool() # generate the viable gene pool (compares against gp.tree_depth_min)
|
|
self.fx_karoo_reproduce() # method 1 - Reproduction
|
|
self.fx_karoo_point_mutate() # method 2 - Point Mutation
|
|
self.fx_karoo_branch_mutate() # method 3 - Branch Mutation
|
|
self.fx_karoo_crossover() # method 4 - Crossover
|
|
self.fx_eval_generation() # evaluate all Trees in a single generation
|
|
|
|
self.population_a = self.fx_evolve_pop_copy(self.population_b, ['GP Tree by Kai Staats, Generation ' + str(self.generation_id)])
|
|
|
|
# "End of line, man!" --CLU
|
|
target = open(self.filename['f'], 'w') # reset the .csv file for the final population
|
|
target.close()
|
|
|
|
self.fx_archive_tree_write(self.population_b, 'f') # save the final generation of Trees to disk
|
|
self.fx_karoo_eol()
|
|
|
|
return
|
|
|
|
|
|
def fx_karoo_eol(self):
|
|
|
|
'''
|
|
The very last method to run in Karoo GP.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
print '\n\033[3m "It is not the strongest of the species that survive, nor the most intelligent,\033[0;0m'
|
|
print '\033[3m but the one most responsive to change."\033[0;0m --Charles Darwin'
|
|
print ''
|
|
print '\033[3m Congrats!\033[0;0m Your multi-generational Karoo GP run is complete.\n'
|
|
print '\033[36m Type \033[1m?\033[0;0m\033[36m to review your options or \033[1mq\033[0;0m\033[36m to quit.\033[0;0m\n'
|
|
self.fx_karoo_pause(1)
|
|
|
|
return
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Generate a new Tree |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_gen_tree_initialise(self, TREE_ID, tree_type, tree_depth_base):
|
|
|
|
'''
|
|
Assign 13 global variables to the array 'tree'.
|
|
|
|
Build the array 'tree' with 13 rows and initally, just 1 column of labels. This array will grow as each new
|
|
node is appended. The values of this array are stored as string characters. Numbers will be forced to integers
|
|
at the point of execution.
|
|
|
|
This method is called by 'fx_gen_tree_build'.
|
|
|
|
Arguments required: TREE_ID, tree_type, tree_depth_base
|
|
'''
|
|
|
|
self.pop_TREE_ID = TREE_ID # pos 0: a unique identifier for each tree
|
|
self.pop_tree_type = tree_type # pos 1: a global constant based upon the initial user setting
|
|
self.pop_tree_depth_base = tree_depth_base # pos 2: a global variable which conveys 'tree_depth_base' as unique to each new Tree
|
|
self.pop_NODE_ID = 1 # pos 3: unique identifier for each node; this is the INDEX KEY to this array
|
|
self.pop_node_depth = 0 # pos 4: depth of each node when committed to the array
|
|
self.pop_node_type = '' # pos 5: root, function, or terminal
|
|
self.pop_node_label = '' # pos 6: operator [+, -, *, ...] or terminal [a, b, c, ...]
|
|
self.pop_node_parent = '' # pos 7: parent node
|
|
self.pop_node_arity = '' # pos 8: number of nodes attached to each non-terminal node
|
|
self.pop_node_c1 = '' # pos 9: child node 1
|
|
self.pop_node_c2 = '' # pos 10: child node 2
|
|
self.pop_node_c3 = '' # pos 11: child node 3 (assumed max of 3 with boolean operator 'if')
|
|
self.pop_fitness = '' # pos 12: fitness score following Tree evaluation
|
|
|
|
self.tree = np.array([ ['TREE_ID'],['tree_type'],['tree_depth_base'],['NODE_ID'],['node_depth'],['node_type'],['node_label'],['node_parent'],['node_arity'],['node_c1'],['node_c2'],['node_c3'],['fitness'] ])
|
|
|
|
return
|
|
|
|
|
|
### Root Node ###
|
|
|
|
def fx_gen_root_node_build(self):
|
|
|
|
'''
|
|
Build the Root node for the initial population.
|
|
|
|
This method is called by 'fx_gen_tree_build'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.fx_gen_function_select() # select the operator for root
|
|
|
|
if self.pop_node_arity == 1: # 1 child
|
|
self.pop_node_c1 = 2
|
|
self.pop_node_c2 = ''
|
|
self.pop_node_c3 = ''
|
|
|
|
elif self.pop_node_arity == 2: # 2 children
|
|
self.pop_node_c1 = 2
|
|
self.pop_node_c2 = 3
|
|
self.pop_node_c3 = ''
|
|
|
|
elif self.pop_node_arity == 3: # 3 children
|
|
self.pop_node_c1 = 2
|
|
self.pop_node_c2 = 3
|
|
self.pop_node_c3 = 4
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_gen_root_node_build: pop_node_arity =', self.pop_node_arity, '\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
self.pop_node_type = 'root'
|
|
|
|
self.fx_gen_node_commit()
|
|
|
|
return
|
|
|
|
|
|
### Function Nodes ###
|
|
|
|
def fx_gen_function_node_build(self):
|
|
|
|
'''
|
|
Build the Function nodes for the intial population.
|
|
|
|
This method is called by 'fx_gen_tree_build'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
for i in range(1, self.pop_tree_depth_base): # increment depth, from 1 through 'tree_depth_base' - 1
|
|
|
|
self.pop_node_depth = i # increment 'node_depth'
|
|
|
|
parent_arity_sum = 0
|
|
prior_sibling_arity = 0 # reset for 'c_buffer' in 'children_link'
|
|
prior_siblings = 0 # reset for 'c_buffer' in 'children_link'
|
|
|
|
for j in range(1, len(self.tree[3])): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if int(self.tree[4][j]) == self.pop_node_depth - 1: # find parent nodes which reside at the prior depth
|
|
parent_arity_sum = parent_arity_sum + int(self.tree[8][j]) # sum arities of all parent nodes at the prior depth
|
|
|
|
# (do *not* merge these 2 "j" loops or it gets all kinds of messed up)
|
|
|
|
for j in range(1, len(self.tree[3])): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if int(self.tree[4][j]) == self.pop_node_depth - 1: # find parent nodes which reside at the prior depth
|
|
|
|
for k in range(1, int(self.tree[8][j]) + 1): # increment through each degree of arity for each parent node
|
|
self.pop_node_parent = int(self.tree[3][j]) # set the parent 'NODE_ID' ...
|
|
prior_sibling_arity = self.fx_gen_function_gen(parent_arity_sum, prior_sibling_arity, prior_siblings) # ... generate a Function ndoe
|
|
prior_siblings = prior_siblings + 1 # sum sibling nodes (current depth) who will spawn their own children (cousins? :)
|
|
|
|
return
|
|
|
|
|
|
def fx_gen_function_gen(self, parent_arity_sum, prior_sibling_arity, prior_siblings):
|
|
|
|
'''
|
|
Generate a single Function node for the initial population.
|
|
|
|
This method is called by 'fx_gen_function_node_build'.
|
|
|
|
Arguments required: parent_arity_sum, prior_sibling_arity, prior_siblings
|
|
'''
|
|
|
|
if self.pop_tree_type == 'f': # user defined as (f)ull
|
|
self.fx_gen_function_select() # retrieve a function
|
|
self.fx_gen_child_link(parent_arity_sum, prior_sibling_arity, prior_siblings) # establish links to children
|
|
|
|
elif self.pop_tree_type == 'g': # user defined as (g)row
|
|
rnd = np.random.randint(2)
|
|
|
|
if rnd == 0: # randomly selected as Function
|
|
self.fx_gen_function_select() # retrieve a function
|
|
self.fx_gen_child_link(parent_arity_sum, prior_sibling_arity, prior_siblings) # establish links to children
|
|
|
|
elif rnd == 1: # randomly selected as Terminal
|
|
self.fx_gen_terminal_select() # retrieve a terminal
|
|
self.pop_node_c1 = ''
|
|
self.pop_node_c2 = ''
|
|
self.pop_node_c3 = ''
|
|
|
|
self.fx_gen_node_commit() # commit new node to array
|
|
prior_sibling_arity = prior_sibling_arity + self.pop_node_arity # sum the arity of prior siblings
|
|
|
|
return prior_sibling_arity
|
|
|
|
|
|
def fx_gen_function_select(self):
|
|
|
|
'''
|
|
Define a single Function (operator extracted from the associated functions.csv) for the initial population.
|
|
|
|
This method is called by 'fx_gen_function_gen' and 'fx_gen_root_node_build'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.pop_node_type = 'func'
|
|
rnd = np.random.randint(0, len(self.functions[:,0])) # call the previously loaded .csv which contains all operators
|
|
self.pop_node_label = self.functions[rnd][0]
|
|
self.pop_node_arity = int(self.functions[rnd][1])
|
|
|
|
return
|
|
|
|
|
|
### Terminal Nodes ###
|
|
|
|
def fx_gen_terminal_node_build(self):
|
|
|
|
'''
|
|
Build the Terminal nodes for the intial population.
|
|
|
|
This method is called by 'fx_gen_tree_build'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.pop_node_depth = self.pop_tree_depth_base # set the final node_depth (same as 'gp.pop_node_depth' + 1)
|
|
|
|
for j in range(1, len(self.tree[3]) ): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if int(self.tree[4][j]) == self.pop_node_depth - 1: # find parent nodes which reside at the prior depth
|
|
|
|
for k in range(1,(int(self.tree[8][j]) + 1)): # increment through each degree of arity for each parent node
|
|
self.pop_node_parent = int(self.tree[3][j]) # set the parent 'NODE_ID' ...
|
|
self.fx_gen_terminal_gen() # ... generate a Terminal node
|
|
|
|
return
|
|
|
|
|
|
def fx_gen_terminal_gen(self):
|
|
|
|
'''
|
|
Generate a single Terminal node for the initial population.
|
|
|
|
This method is called by 'fx_gen_terminal_node_build'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.fx_gen_terminal_select() # retrieve a terminal
|
|
self.pop_node_c1 = ''
|
|
self.pop_node_c2 = ''
|
|
self.pop_node_c3 = ''
|
|
|
|
self.fx_gen_node_commit() # commit new node to array
|
|
|
|
return
|
|
|
|
|
|
def fx_gen_terminal_select(self):
|
|
|
|
'''
|
|
Define a single Terminal (variable extracted from the top row of the associated TRAINING data)
|
|
|
|
This method is called by 'fx_gen_terminal_gen' and 'fx_gen_function_gen'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.pop_node_type = 'term'
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
self.pop_node_label = self.terminals[rnd]
|
|
self.pop_node_arity = 0
|
|
|
|
return
|
|
|
|
|
|
### The Lovely Children ###
|
|
|
|
def fx_gen_child_link(self, parent_arity_sum, prior_sibling_arity, prior_siblings):
|
|
|
|
'''
|
|
Link each parent node to its children in the intial population.
|
|
|
|
This method is called by 'fx_gen_function_gen'.
|
|
|
|
Arguments required: parent_arity_sum, prior_sibling_arity, prior_siblings
|
|
'''
|
|
|
|
c_buffer = 0
|
|
|
|
for n in range(1, len(self.tree[3]) ): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if int(self.tree[4][n]) == self.pop_node_depth - 1: # find all nodes that reside at the prior (parent) 'node_depth'
|
|
|
|
c_buffer = self.pop_NODE_ID + (parent_arity_sum + prior_sibling_arity - prior_siblings) # One algo to rule the world!
|
|
|
|
if self.pop_node_arity == 0: # terminal in a Grow Tree
|
|
self.pop_node_c1 = ''
|
|
self.pop_node_c2 = ''
|
|
self.pop_node_c3 = ''
|
|
|
|
elif self.pop_node_arity == 1: # 1 child
|
|
self.pop_node_c1 = c_buffer
|
|
self.pop_node_c2 = ''
|
|
self.pop_node_c3 = ''
|
|
|
|
elif self.pop_node_arity == 2: # 2 children
|
|
self.pop_node_c1 = c_buffer
|
|
self.pop_node_c2 = c_buffer + 1
|
|
self.pop_node_c3 = ''
|
|
|
|
elif self.pop_node_arity == 3: # 3 children
|
|
self.pop_node_c1 = c_buffer
|
|
self.pop_node_c2 = c_buffer + 1
|
|
self.pop_node_c3 = c_buffer + 2
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_gen_child_link: pop_node_arity =', self.pop_node_arity, '\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
return
|
|
|
|
|
|
def fx_gen_node_commit(self):
|
|
|
|
'''
|
|
Commit the values of a new node (root, function, or terminal) to the array 'tree'.
|
|
|
|
This method is called by 'fx_gen_root_node_build' and 'fx_gen_function_gen' and 'fx_gen_terminal_gen'.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.tree = np.append(self.tree, [ [self.pop_TREE_ID],[self.pop_tree_type],[self.pop_tree_depth_base],[self.pop_NODE_ID],[self.pop_node_depth],[self.pop_node_type],[self.pop_node_label],[self.pop_node_parent],[self.pop_node_arity],[self.pop_node_c1],[self.pop_node_c2],[self.pop_node_c3],[self.pop_fitness] ], 1)
|
|
|
|
self.pop_NODE_ID = self.pop_NODE_ID + 1
|
|
|
|
return
|
|
|
|
|
|
def fx_gen_tree_build(self, TREE_ID, tree_type, tree_depth_base):
|
|
|
|
'''
|
|
This method combines 4 sub-methods into a single method for ease of deployment. It is designed to executed
|
|
within a loop such that an entire population is built. However, it may also be run from the command line,
|
|
passing a single TREE_ID to the method.
|
|
|
|
'tree_type' is either (f)ull or (g)row. Note, however, that when the user selects 'ramped 50/50' at launch,
|
|
it is still (f) or (g) which are passed to this method.
|
|
|
|
This method is called by a 'fx_evolve_crossover' and 'fx_evolve_grow_mutate' and 'fx_karoo_construct'.
|
|
|
|
Arguments required: TREE_ID, tree_type, tree_depth_base
|
|
'''
|
|
|
|
self.fx_gen_tree_initialise(TREE_ID, tree_type, tree_depth_base) # initialise a new Tree
|
|
self.fx_gen_root_node_build() # build the Root node
|
|
self.fx_gen_function_node_build() # build the Function nodes
|
|
self.fx_gen_terminal_node_build() # build the Terminal nodes
|
|
|
|
return # each Tree is written to 'gp.tree'
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Evaluate a Tree |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_eval_poly(self, tree):
|
|
|
|
'''
|
|
Evaluate a Tree and generate its multivariate expression (both raw and Sympified).
|
|
|
|
We need to extract the variables from the expression. However, these variables are no longer correlated
|
|
to the original variables listed across the top of each column of data.csv. Therefore, we must re-assign
|
|
the respective values for each subsequent row in the data .csv, for each Tree's unique expression.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
self.algo_raw = self.fx_eval_label(tree, 1) # pass the root 'node_id', then flatten the Tree to a string
|
|
self.algo_sym = sympify(self.algo_raw) # convert string to a functional expression (the coolest line in Karoo! :)
|
|
|
|
return
|
|
|
|
|
|
def fx_eval_label(self, tree, node_id):
|
|
|
|
'''
|
|
Evaluate all or part of a Tree (starting at node_id) and return a raw mutivariate expression ('algo_raw').
|
|
|
|
In the main code, this method is called once per Tree, but may be called at any time to prepare an expression
|
|
for any full or partial (branch) Tree contained in 'population'.
|
|
|
|
Pass the starting node for recursion via the local variable 'node_id' where the local variable 'tree' is a
|
|
copy of the Tree you desire to evaluate.
|
|
|
|
Arguments required: tree, node_id
|
|
'''
|
|
|
|
# if tree[6, node_id] == 'not': tree[6, node_id] = ', not' # temp until this can be fixed at data_load
|
|
|
|
node_id = int(node_id)
|
|
|
|
if tree[8, node_id] == '0': # arity of 0 for the pattern '[term]'
|
|
return '(' + tree[6, node_id] + ')' # 'node_label' (function or terminal)
|
|
|
|
else:
|
|
if tree[8, node_id] == '1': # arity of 1 for the explicit pattern 'not [term]'
|
|
return self.fx_eval_label(tree, tree[9, node_id]) + tree[6, node_id] # original code
|
|
|
|
elif tree[8, node_id] == '2': # arity of 2 for the pattern '[func] [term] [func]'
|
|
return self.fx_eval_label(tree, tree[9, node_id]) + tree[6, node_id] + self.fx_eval_label(tree, tree[10, node_id])
|
|
|
|
elif tree[8, node_id] == '3': # arity of 3 for the explicit pattern 'if [term] then [term] else [term]'
|
|
return tree[6, node_id] + self.fx_eval_label(tree, tree[9, node_id]) + ' then ' + self.fx_eval_label(tree, tree[10, node_id]) + ' else ' + self.fx_eval_label(tree, tree[11, node_id])
|
|
|
|
|
|
def fx_eval_id(self, tree, node_id):
|
|
|
|
'''
|
|
Evaluate all or part of a Tree and return a list of all 'NODE_ID's.
|
|
|
|
This method generates a list of all 'NODE_ID's from the given Node and below. It is used primarily to generate
|
|
'branch' for the multi-generational mutation of Trees.
|
|
|
|
Pass the starting node for recursion via the local variable 'node_id' where the local variable 'tree' is a copy
|
|
of the Tree you desire to evaluate.
|
|
|
|
Arguments required: tree, node_id
|
|
'''
|
|
|
|
node_id = int(node_id)
|
|
|
|
if tree[8, node_id] == '0': # arity of 0 for the pattern '[NODE_ID]'
|
|
return tree[3, node_id] # 'NODE_ID'
|
|
|
|
else:
|
|
if tree[8, node_id] == '1': # arity of 1 for the pattern '[NODE_ID], [NODE_ID]'
|
|
return tree[3, node_id] + ', ' + self.fx_eval_id(tree, tree[9, node_id])
|
|
|
|
elif tree[8, node_id] == '2': # arity of 2 for the pattern '[NODE_ID], [NODE_ID], [NODE_ID]'
|
|
return tree[3, node_id] + ', ' + self.fx_eval_id(tree, tree[9, node_id]) + ', ' + self.fx_eval_id(tree, tree[10, node_id])
|
|
|
|
elif tree[8, node_id] == '3': # arity of 3 for the pattern '[NODE_ID], [NODE_ID], [NODE_ID], [NODE_ID]'
|
|
return tree[3, node_id] + ', ' + self.fx_eval_id(tree, tree[9, node_id]) + ', ' + self.fx_eval_id(tree, tree[10, node_id]) + ', ' + self.fx_eval_id(tree, tree[11, node_id])
|
|
|
|
|
|
def fx_eval_generation(self):
|
|
|
|
'''
|
|
This method invokes the evaluation of an entire generation of Trees, as engaged by karoo_gp_server.py and the
|
|
'cont' function of karoo_go_main.py. It automatically evaluates population_b before invoking the copy of _b to _a.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
if self.display != 's':
|
|
if self.display == 'i': print ''
|
|
print '\n Evaluate all Trees in Generation', self.generation_id
|
|
if self.display == 'i': self.fx_karoo_pause(0)
|
|
|
|
self.fx_evolve_tree_renum(self.population_b) # population renumber
|
|
self.fx_fitness_gym(self.population_b) # run 'fx_eval', 'fx_fitness', 'fx_fitness_store', and fitness record
|
|
self.fx_archive_tree_write(self.population_b, 'a') # archive current population as foundation for next generation
|
|
|
|
if self.display != 's':
|
|
print '\n Copy gp.population_b to gp.population_a\n'
|
|
|
|
return
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Train and Test a Tree |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_fitness_gym(self, population):
|
|
|
|
'''
|
|
Part 1 evaluates each expression against the data, line for line. This is the most time consuming and
|
|
computationally expensive part of genetic programming. When GPUs are available, the performance can increase
|
|
by many orders of magnitude.
|
|
|
|
Part 2 evaluates every Tree in each generation to determine which have the best, overall fitness score. This
|
|
could be the highest or lowest depending upon if the fitness function is maximising (higher is better) or
|
|
minimising (lower is better). The total fitness score is then saved with each Tree in the external .csv file.
|
|
|
|
Part 3 compares the fitness of each Tree to the prior best fit in order to track those that improve with each
|
|
comparison. For matching functions, all the Trees will have the same fitness score, but they may present more
|
|
than one solution. For minimisation and maximisation functions, the final Tree should present the best overall
|
|
fitness for that generation. It is important to note that Part 3 does *not* in any way influence the Tournament
|
|
Selection which is a stand-alone process.
|
|
|
|
Arguments required: population
|
|
'''
|
|
|
|
fitness_best = 0
|
|
self.fittest_dict = {}
|
|
time_sum = 0
|
|
|
|
for tree_id in range(1, len(population)):
|
|
|
|
### PART 1 - GENERATE MULTIVARIATE EXPRESSION FOR EACH TREE ###
|
|
self.fx_eval_poly(population[tree_id]) # extract the expression
|
|
if self.display not in ('s'): print '\t\033[36mTree', population[tree_id][0][1], 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m'
|
|
|
|
|
|
### PART 2 - EVALUATE FITNESS FOR EACH TREE AGAINST TRAINING DATA ###
|
|
fitness = 0
|
|
|
|
expr = str(self.algo_sym) # get sympified expression and process it with TF - tested 2017 02/02
|
|
result = self.fx_fitness_eval(expr, self.data_train)
|
|
fitness = result['fitness'] # extract fitness score
|
|
|
|
if self.display == 'i':
|
|
print '\t \033[36m with fitness sum:\033[1m', fitness, '\033[0;0m\n'
|
|
|
|
self.fx_fitness_store(population[tree_id], fitness) # store Fitness with each Tree
|
|
|
|
|
|
### PART 3 - COMPARE FITNESS OF ALL TREES IN CURRENT GENERATION ###
|
|
if self.kernel == 'c': # display best fit Trees for the CLASSIFY kernel
|
|
if fitness >= fitness_best: # find the Tree with Maximum fitness score
|
|
fitness_best = fitness # set best fitness score
|
|
self.fittest_dict.update({tree_id:self.algo_sym}) # add to dictionary if fitness >= prior
|
|
|
|
elif self.kernel == 'r': # display best fit Trees for the REGRESSION kernel
|
|
if fitness_best == 0: fitness_best = fitness # set the baseline first time through
|
|
if fitness <= fitness_best: # find the Tree with Minimum fitness score
|
|
fitness_best = fitness # set best fitness score
|
|
self.fittest_dict.update({tree_id:self.algo_sym}) # add to dictionary if fitness <= prior
|
|
|
|
elif self.kernel == 'm': # display best fit Trees for the MATCH kernel
|
|
if fitness == self.data_train_rows: # find the Tree with a perfect match for all data rows
|
|
fitness_best = fitness # set best fitness score
|
|
self.fittest_dict.update({tree_id:self.algo_sym}) # add to dictionary if all rows match
|
|
|
|
# elif self.kernel == '[other]': # display best fit Trees for the [other] kernel
|
|
# if fitness [>=, <=] fitness_best: # find the Tree with [Maximum or Minimum] fitness score
|
|
# fitness_best = fitness # set best fitness score
|
|
# self.fittest_dict.update({tree_id:self.algo_sym}) # add to dictionary
|
|
|
|
print '\n\033[36m ', len(self.fittest_dict.keys()), 'trees\033[1m', np.sort(self.fittest_dict.keys()), '\033[0;0m\033[36moffer the highest fitness scores.\033[0;0m'
|
|
if self.display == 'g': self.fx_karoo_pause(0)
|
|
|
|
return
|
|
|
|
|
|
def fx_fitness_eval(self, expr, data, get_labels = False): # used to be fx_fitness_eval
|
|
|
|
'''
|
|
Computes tree expression using TensorFlow (TF) returning results and fitness scores.
|
|
|
|
This method orchestrates most of the TF routines by parsing input string expression and converting it into TF
|
|
operation graph which then is processed in an isolated TF session to compute the results and corresponding fitness
|
|
values.
|
|
|
|
'self.tf_device' - controls which device will be used for computations (CPU or GPU).
|
|
'self.tf_device_log' - controls device placement logging (debug only).
|
|
|
|
Args:
|
|
'expr' - a string containing math expression to be computed on the data. Variable names should match corresponding
|
|
terminal names in 'self.terminals'. Only algebraic operations are currently supported (+, -, *, /, **).
|
|
|
|
'data' - an 'n by m' matrix of the data points containing n observations and m features each. Variable order should
|
|
match corresponding order of terminals in 'self.terminals'.
|
|
|
|
'get_labels' - a boolean flag which controls whether classification labels should be extracted from the results.
|
|
This is applied only to the CLASSIFY kernel and defaults to 'False'.
|
|
|
|
Returns:
|
|
A dict mapping keys to the following outputs:
|
|
'result' - an array of the results of applying given expression to the data
|
|
'labels' - an array of the labels extracted from the results; defined only for CLASSIFY kernel, None otherwise
|
|
'solution' - an array of the solution values extracted from the data (variable 's' in the dataset)
|
|
'pairwise_fitness' - an array of the element-wise results of applying corresponding fitness kernel function
|
|
'fitness' - aggregated scalar fitness score
|
|
|
|
Arguments required: expr, data
|
|
'''
|
|
|
|
# Initialize TensorFlow session
|
|
tf.reset_default_graph() # Reset TF internal state and cache (after previous processing)
|
|
config = tf.ConfigProto(log_device_placement=self.tf_device_log, allow_soft_placement=True)
|
|
config.gpu_options.allow_growth = True
|
|
|
|
with tf.Session(config=config) as sess:
|
|
with sess.graph.device(self.tf_device):
|
|
|
|
# Load data into TF
|
|
tensors = {}
|
|
for i in range(len(self.terminals)):
|
|
var = self.terminals[i]
|
|
tensors[var] = tf.constant(data[:, i], dtype=tf.float32)
|
|
|
|
# Transform string expression into TF operation graph
|
|
result = self.fx_fitness_expr_parse(expr, tensors)
|
|
|
|
labels = tf.no_op() # a placeholder, applies only to CLASSIFY kernel
|
|
solution = tensors['s'] # solution value is assumed to be stored in 's' terminal
|
|
|
|
# Add fitness computation into TF graph
|
|
if self.kernel == 'c': # CLASSIFY kernels
|
|
if get_labels: labels = tf.map_fn(self.fx_fitness_labels_map, result, dtype=[tf.int32, tf.string], swap_memory=True)
|
|
pairwise_fitness = self.fx_fitness_train_classify(result, tf.cast(solution, tf.float32))
|
|
|
|
elif self.kernel == 'r': # REGRESSION kernel
|
|
pairwise_fitness = self.fx_fitness_train_regress(result, tf.cast(solution, tf.float32))
|
|
|
|
elif self.kernel == 'm': # MATCH kernel
|
|
pairwise_fitness = self.fx_fitness_train_match(result, solution)
|
|
|
|
# elif self.kernel == '[other]': # [OTHER] kernel
|
|
# pairwise_fitness = self.fx_fitness_train_[other](result ?, solution ?)
|
|
|
|
else: raise Exception('Kernel type is wrong or missing. You entered {}'.format(self.kernel))
|
|
|
|
fitness = tf.reduce_sum(pairwise_fitness)
|
|
|
|
# Process TF graph and collect the results
|
|
result, labels, solution, fitness, pairwise_fitness = sess.run([result, labels, solution, fitness, pairwise_fitness])
|
|
|
|
return {'result': result, 'labels': labels, 'solution': solution, 'fitness': float(fitness), 'pairwise_fitness': pairwise_fitness}
|
|
|
|
|
|
def fx_fitness_expr_parse(self, expr, tensors):
|
|
|
|
'''
|
|
Extract expression tree from the string algo_sym and transform into TensorFlow (TF) graph.
|
|
|
|
Arguments required: expr, tensors
|
|
'''
|
|
|
|
tree = ast.parse(expr, mode='eval').body
|
|
|
|
return self.fx_fitness_node_parse(tree, tensors)
|
|
|
|
|
|
def fx_fitness_node_parse(self, node, tensors):
|
|
|
|
'''
|
|
Recursively transforms parsed expression tree into TensorFlow (TF) graph.
|
|
|
|
Arguments required: node, tensors
|
|
'''
|
|
|
|
if isinstance(node, ast.Name): # <tensor_name>
|
|
return tensors[node.id]
|
|
|
|
elif isinstance(node, ast.Num): # <number>
|
|
shape = tensors[tensors.keys()[0]].get_shape()
|
|
return tf.constant(node.n, shape=shape, dtype=tf.float32)
|
|
|
|
elif isinstance(node, ast.BinOp): # <left> <operator> <right>
|
|
return operators[type(node.op)](self.fx_fitness_node_parse(node.left, tensors), self.fx_fitness_node_parse(node.right, tensors))
|
|
|
|
elif isinstance(node, ast.UnaryOp): # <operator> <operand> e.g., -1
|
|
return operators[type(node.op)](self.fx_fitness_node_parse(node.operand, tensors))
|
|
|
|
else: raise TypeError(node)
|
|
|
|
|
|
def fx_fitness_labels_map(self, result):
|
|
|
|
'''
|
|
Creates label extraction TensorFlow (TF) sub-graph for CLASSIFY kernel defined as a sequence of boolean conditions.
|
|
Outputs an array of tuples containing label extracted from the result and corresponding boolean condition triggered.
|
|
|
|
The original (pre-TensorFlow) code is as follows:
|
|
|
|
skew = (self.class_labels / 2) - 1 # '-1' keeps a binary classification splitting over the origin
|
|
if solution == 0 and result <= 0 - skew; fitness = 1: # check for first class (the left-most bin)
|
|
elif solution == self.class_labels - 1 and result > solution - 1 - skew; fitness = 1: # check for last class (the right-most bin)
|
|
elif solution - 1 - skew < result <= solution - skew; fitness = 1: # check for class bins between first and last
|
|
else: fitness = 0 # no class match
|
|
|
|
See 'fx_fitness_train_classify' for a description of the multi-class classifier.
|
|
|
|
Arguments required: result
|
|
'''
|
|
|
|
skew = (self.class_labels / 2) - 1
|
|
label_rules = {self.class_labels - 1: (tf.constant(self.class_labels - 1), tf.constant(' > {}'.format(self.class_labels - 2 - skew)))}
|
|
|
|
for class_label in range(self.class_labels - 2, 0, -1):
|
|
cond = (class_label - 1 - skew < result) & (result <= class_label - skew)
|
|
label_rules[class_label] = tf.cond(cond, lambda: (tf.constant(class_label), tf.constant(' <= {}'.format(class_label - skew))), lambda: label_rules[class_label + 1])
|
|
|
|
zero_rule = tf.cond(result <= 0 - skew, lambda: (tf.constant(0), tf.constant(' <= {}'.format(0 - skew))), lambda: label_rules[1])
|
|
|
|
return zero_rule
|
|
|
|
|
|
def fx_fitness_train_classify(self, result, solution): # CLASSIFICATION kernel
|
|
|
|
'''
|
|
Creates element-wise fitness computation TensorFlow (TF) sub-graph for CLASSIFY kernel.
|
|
|
|
This method uses the 'sympified' (SymPy) expression ('algo_sym') created in 'fx_eval_poly' and the data set
|
|
loaded at run-time to evaluate the fitness of the selected kernel.
|
|
|
|
This multiclass classifer compares each row of a given Tree to the known solution, comparing estimated values
|
|
(labels) generated by Karoo GP against the correct labels. This method is able to work with any number of class
|
|
labels, from 2 to n. The left-most bin includes -inf. The right-most bin includes +inf. Those inbetween are
|
|
by default confined to the spacing of 1.0 each, as defined by:
|
|
|
|
(solution - 1) < result <= solution
|
|
|
|
The skew adjusts the boundaries of the bins such that they fall on both the negative and positive sides of the
|
|
origin. At the time of this writing, an odd number of class labels will generate an extra bin on the positive
|
|
side of origin as it has not yet been determined the effect of enabling the middle bin to include both a
|
|
negative and positive space.
|
|
|
|
Arguments required: result, solution
|
|
'''
|
|
|
|
skew = (self.class_labels / 2) - 1
|
|
rule11 = tf.equal(solution, 0)
|
|
rule12 = tf.less_equal(result, 0 - skew)
|
|
rule13 = tf.logical_and(rule11, rule12)
|
|
rule21 = tf.equal(solution, self.class_labels - 1)
|
|
rule22 = tf.greater(result, solution - 1 - skew)
|
|
rule23 = tf.logical_and(rule21, rule22)
|
|
rule31 = tf.less(solution - 1 - skew, result)
|
|
rule32 = tf.less_equal(result, solution - skew)
|
|
rule33 = tf.logical_and(rule31, rule32)
|
|
|
|
return tf.cast(tf.logical_or(tf.logical_or(rule13, rule23), rule33), tf.int32)
|
|
|
|
|
|
def fx_fitness_train_regress(self, result, solution): # REGRESSION kernel
|
|
|
|
'''
|
|
Creates element-wise fitness computation TensorFlow (TF) sub-graph for REGRESSION kernel.
|
|
|
|
This is a minimisation function which seeks a result which is closest to the solution.
|
|
|
|
[need to write more]
|
|
|
|
Arguments required: result, solution
|
|
'''
|
|
|
|
return tf.abs(solution - result)
|
|
|
|
|
|
def fx_fitness_train_match(self, result, solution): # MATCH kernel
|
|
|
|
'''
|
|
Creates element-wise fitness computation TensorFlow (TF) sub-graph for MATCH kernel.
|
|
|
|
This is a maximization function which seeks an exact solution (a perfect match).
|
|
|
|
[need to write more]
|
|
|
|
Arguments required: result, solution
|
|
'''
|
|
|
|
return tf.cast(tf.equal(solution, result), tf.int32)
|
|
|
|
|
|
# def fx_fitness_train_[other](self, result, solution): # [OTHER] kernel
|
|
|
|
# '''
|
|
# Creates element-wise fitness computation TensorFlow (TF) sub-graph for [other] kernel.
|
|
|
|
# This is a [minimisation or maximization] function which [insert description].
|
|
|
|
# return tf.[?]([insert formula])
|
|
# '''
|
|
|
|
|
|
def fx_fitness_store(self, tree, fitness):
|
|
|
|
'''
|
|
Records the fitness and length of the raw algorithm (multivariate expression) to the Numpy array. Parsimony can
|
|
be used to apply pressure to the evolutionary process to select from a set of trees with the same fitness function
|
|
the one(s) with the simplest (shortest) multivariate expression.
|
|
|
|
Arguments required: tree, fitness
|
|
'''
|
|
|
|
fitness = float(fitness)
|
|
fitness = round(fitness, self.precision)
|
|
|
|
tree[12][1] = fitness # store the fitness with each tree
|
|
tree[12][2] = len(str(self.algo_raw)) # store the length of the raw algo for parsimony
|
|
# if len(tree[3]) > 4: # if the Tree array is wide enough -- SEE SCRATCHPAD
|
|
|
|
return
|
|
|
|
|
|
def fx_fitness_tournament(self, tourn_size):
|
|
|
|
'''
|
|
Select one Tree by means of a Tournament in which 'tourn_size' contenders are randomly selected and then
|
|
compared for their respective fitness (as determined in 'fx_fitness_gym'). The tournament is engaged for each
|
|
of the four types of inter-generational evolution: reproduction, point mutation, branch (full and grow)
|
|
mutation, and crossover (sexual reproduction).
|
|
|
|
The original Tournament Selection drew directly from the foundation generation (gp.generation_a). However,
|
|
with the introduction of a minimum number of nodes as defined by the user ('gp.tree_depth_min'),
|
|
'gp.gene_pool' provides only from those Trees which meet all criteria.
|
|
|
|
With upper (max depth) and lower (min nodes) invoked, one may enjoy interesting results. Stronger boundary
|
|
parameters (a reduced gap between the min and max number of nodes) may invoke more compact solutions, but also
|
|
runs the risk of elitism, even total population die-off where a healthy population once existed.
|
|
|
|
Arguments required: tourn_size
|
|
'''
|
|
|
|
tourn_test = 0
|
|
# short_test = 0 # an incomplete parsimony test (seeking shortest solution)
|
|
|
|
if self.display == 'i': print '\n\tEnter the tournament ...'
|
|
|
|
for n in range(tourn_size):
|
|
# tree_id = np.random.randint(1, self.tree_pop_max + 1) # former method of selection from the unfiltered population
|
|
rnd = np.random.randint(len(self.gene_pool)) # select one Tree at random from the gene pool
|
|
tree_id = int(self.gene_pool[rnd])
|
|
|
|
fitness = float(self.population_a[tree_id][12][1]) # extract the fitness from the array
|
|
fitness = round(fitness, self.precision) # force 'result' and 'solution' to the same number of floating points
|
|
|
|
if self.fitness_type == 'max': # if the fitness function is Maximising
|
|
|
|
# first time through, 'tourn_test' will be initialised below
|
|
|
|
if fitness > tourn_test: # if the current Tree's 'fitness' is greater than the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '>', tourn_test, 'and leads\033[0;0m'
|
|
tourn_lead = tree_id # set 'TREE_ID' for the new leader
|
|
tourn_test = fitness # set 'fitness' of the new leader
|
|
# short_test = int(self.population_a[tree_id][12][2]) # set len(algo_raw) of new leader
|
|
|
|
elif fitness == tourn_test: # if the current Tree's 'fitness' is equal to the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '=', tourn_test, 'and leads\033[0;0m'
|
|
tourn_lead = tree_id # in case there is no variance in this tournament
|
|
# tourn_test remains unchanged
|
|
|
|
# NEED TO ADD: option for parsimony
|
|
# if int(self.population_a[tree_id][12][2]) < short_test:
|
|
# short_test = int(self.population_a[tree_id][12][2]) # set len(algo_raw) of new leader
|
|
# print '\t\033[36m with improved parsimony score of:\033[1m', short_test, '\033[0;0m'
|
|
|
|
elif fitness < tourn_test: # if the current Tree's 'fitness' is less than the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '<', tourn_test, 'and is ignored\033[0;0m'
|
|
# tourn_lead remains unchanged
|
|
# tourn_test remains unchanged
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_fitness_tournament: fitness =', fitness, 'and tourn_test =', tourn_test, '\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
|
|
elif self.fitness_type == 'min': # if the fitness function is Minimising
|
|
|
|
if tourn_test == 0: # first time through, 'tourn_test' is given a baseline value
|
|
tourn_test = fitness
|
|
|
|
if fitness < tourn_test: # if the current Tree's 'fitness' is less than the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '<', tourn_test, 'and leads\033[0;0m'
|
|
tourn_lead = tree_id # set 'TREE_ID' for the new leader
|
|
tourn_test = fitness # set 'fitness' of the new leader
|
|
|
|
elif fitness == tourn_test: # if the current Tree's 'fitness' is equal to the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '=', tourn_test, 'and leads\033[0;0m'
|
|
tourn_lead = tree_id # in case there is no variance in this tournament
|
|
# tourn_test remains unchanged
|
|
|
|
elif fitness > tourn_test: # if the current Tree's 'fitness' is greater than the priors'
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has fitness', fitness, '>', tourn_test, 'and is ignored\033[0;0m'
|
|
# tourn_lead remains unchanged
|
|
# tourn_test remains unchanged
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_fitness_tournament: fitness =', fitness, 'and tourn_test =', tourn_test, '\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
|
|
tourn_winner = np.copy(self.population_a[tourn_lead]) # copy full Tree so as to not inadvertantly modify the original tree
|
|
|
|
if self.display == 'i': print '\n\t\033[36mThe winner of the tournament is Tree:\033[1m', tourn_winner[0][1], '\033[0;0m'
|
|
|
|
return tourn_winner
|
|
|
|
|
|
def fx_fitness_gene_pool(self):
|
|
|
|
'''
|
|
With the introduction of the minimum number of nodes parameter (gp.tree_depth_min), the means by which the
|
|
lower node count is enforced is through the creation of a gene pool from those Trees which contain equal or
|
|
greater nodes to the user defined limit.
|
|
|
|
When the minimum node count is human guided, it can help keep the solution from defaulting to a local minimum,
|
|
as with 't/t' in the Kepler problem. However, the ramification of this limitation on the evolutionary process
|
|
has not been fully studied.
|
|
|
|
This method is automatically invoked with every Tournament Selection ('fx_fitness_tournament').
|
|
|
|
At this time, the gene pool does *not* limit the number of times any given Tree may be selected for mutation or
|
|
reproduction nor does it take into account parsimony (seeking the simplest multivariate expression).
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
self.gene_pool = []
|
|
if self.display == 'i': print '\n Prepare a viable gene pool ...'; self.fx_karoo_pause(0)
|
|
|
|
for tree_id in range(1, len(self.population_a)):
|
|
|
|
self.fx_eval_poly(self.population_a[tree_id]) # extract the expression
|
|
|
|
if len(self.population_a[tree_id][3])-1 >= self.tree_depth_min and self.algo_sym != 1: # check if Tree meets the requirements
|
|
if self.display == 'i': print '\t\033[36m Tree', tree_id, 'has >=', self.tree_depth_min, 'nodes and is added to the gene pool\033[0;0m'
|
|
self.gene_pool.append(self.population_a[tree_id][0][1])
|
|
|
|
if len(self.gene_pool) > 0 and self.display == 'i': print '\n\t The total population of the gene pool is', len(self.gene_pool); self.fx_karoo_pause(0)
|
|
|
|
elif len(self.gene_pool) <= 0: # the evolutionary constraints were too tight, killing off the entire population
|
|
# self.generation_id = self.generation_id - 1 # revert the increment of the 'generation_id'
|
|
# self.generation_max = self.generation_id # catch the unused "cont" values in the 'fx_karoo_pause' method
|
|
print "\n\t\033[31m\033[3m 'They're dead Jim. They're all dead!'\033[0;0m There are no Trees in the gene pool. You should archive your populations and (q)uit."; self.fx_karoo_pause(0)
|
|
|
|
return
|
|
|
|
|
|
def fx_fitness_test_classify(self, result):
|
|
|
|
'''
|
|
Print the Precision-Recall and Confusion Matrix for a CLASSIFICATION run against the test data.
|
|
|
|
From scikit-learn.org/stable/auto_examples/model_selection/plot_precision_recall.html
|
|
Precision (P) = true_pos / true_pos + false_pos
|
|
Recall (R) = true_pos / true_pos + false_neg
|
|
harmonic mean of Precision and Recall (F1) = 2(P x R) / (P + R)
|
|
|
|
From scikit-learn.org/stable/modules/generated/sklearn.metrics.classification_report.html
|
|
y_pred = result, the estimated target values (labels) generated by Karoo GP
|
|
y_true = solution, the correct target values (labels) associated with the data
|
|
|
|
Arguments required: result
|
|
'''
|
|
|
|
for i in range(len(result['result'])):
|
|
print '\t\033[36m Data row {} predicts class:\033[1m {} ({} label) as {:.2f}{}\033[0;0m'.format(i, int(result['labels'][0][i]), int(result['solution'][i]), result['result'][i], result['labels'][1][i])
|
|
|
|
print '\n Fitness score: {}'.format(result['fitness'])
|
|
print '\n Precision-Recall report:\n', skm.classification_report(result['solution'], result['labels'][0])
|
|
print ' Confusion matrix:\n', skm.confusion_matrix(result['solution'], result['labels'][0])
|
|
|
|
return
|
|
|
|
|
|
def fx_fitness_test_regress(self, result):
|
|
|
|
'''
|
|
Print the Fitness score and Mean Squared Error for a REGRESSION run against the test data.
|
|
'''
|
|
|
|
for i in range(len(result['result'])):
|
|
print '\t\033[36m Data row {} predicts value:\033[1m {:.2f} ({:.2f} True)\033[0;0m'.format(i, result['result'][i], result[ 'solution'][i])
|
|
|
|
MSE, fitness = skm.mean_squared_error(result['result'], result['solution']), result['fitness']
|
|
print '\n\t Regression fitness score: {}'.format(fitness)
|
|
print '\t Mean Squared Error: {}'.format(MSE)
|
|
|
|
return
|
|
|
|
|
|
def fx_fitness_test_match(self, result):
|
|
|
|
'''
|
|
Print the accuracy for a MATCH kernel run against the test data.
|
|
'''
|
|
|
|
for i in range(len(result['result'])):
|
|
print '\t\033[36m Data row {} predicts value:\033[1m {} ({} label)\033[0;0m'.format(i, int(result['result'][i]), int(result['solution'][i]))
|
|
|
|
print '\n\tMatching fitness score: {}'.format(result['fitness'])
|
|
|
|
return
|
|
|
|
|
|
# def fx_fitness_test_[other](self, result):
|
|
|
|
# '''
|
|
# Print the [statistical measure] for a [OTHER] kernel run against the test data.
|
|
# '''
|
|
|
|
# for i in range(len(result['result'])):
|
|
# print '\t\033[36m Data row {} predicts value:\033[1m {} ({} label)\033[0;0m'.format(i, int(result['result'][i]), int(result['solution'][i]))
|
|
|
|
# print '\n\tFitness score: {}'.format(result['fitness'])
|
|
|
|
# return
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Evolve a Population |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_evolve_point_mutate(self, tree):
|
|
|
|
'''
|
|
Mutate a single point in any Tree (Grow or Full).
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
node = np.random.randint(1, len(tree[3])) # randomly select a point in the Tree (including root)
|
|
if self.display == 'i': print '\t\033[36m with', tree[5][node], 'node\033[1m', tree[3][node], '\033[0;0m\033[36mchosen for mutation\n\033[0;0m'
|
|
elif self.display == 'db': print '\n\n\033[33m *** Point Mutation *** \033[0;0m\n\n\033[36m This is the unaltered tourn_winner:\033[0;0m\n', tree
|
|
|
|
if tree[5][node] == 'root':
|
|
rnd = np.random.randint(0, len(self.functions[:,0])) # call the previously loaded .csv which contains all operators
|
|
tree[6][node] = self.functions[rnd][0] # replace function (operator)
|
|
|
|
elif tree[5][node] == 'func':
|
|
rnd = np.random.randint(0, len(self.functions[:,0])) # call the previously loaded .csv which contains all operators
|
|
tree[6][node] = self.functions[rnd][0] # replace function (operator)
|
|
|
|
elif tree[5][node] == 'term':
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
tree[6][node] = self.terminals[rnd] # replace terminal (variable)
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_evolve_point_mutate, node_type =', tree[5][node], '\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
tree = self.fx_evolve_fitness_wipe(tree) # wipe fitness data
|
|
|
|
if self.display == 'db': print '\n\033[36m This is tourn_winner after node\033[1m', node, '\033[0;0m\033[36mmutation and updates:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
return tree, node # 'node' is returned only to be assigned to the 'tourn_trees' record keeping
|
|
|
|
|
|
def fx_evolve_full_mutate(self, tree, branch):
|
|
|
|
'''
|
|
Mutate a branch of a Full method Tree.
|
|
|
|
The full mutate method is straight-forward. A branch was generated and passed to this method. As the size and
|
|
shape of the Tree must remain identical, each node is mutated sequentially (copied from the new Tree to replace
|
|
the old, node for node), where functions remain functions and terminals remain terminals.
|
|
|
|
Arguments required: tree, branch
|
|
'''
|
|
|
|
if self.display == 'db': print '\n\n\033[33m *** Full Mutation: function to function *** \033[0;0m\n\n\033[36m This is the unaltered tourn_winner:\033[0;0m\n', tree
|
|
|
|
for n in range(len(branch)):
|
|
|
|
# 'root' is not made available for Full mutation as this would build an entirely new Tree
|
|
|
|
if tree[5][branch[n]] == 'func':
|
|
if self.display == 'i': print '\t\033[36m from\033[1m', tree[5][branch[n]], '\033[0;0m\033[36mto\033[1m func \033[0;0m'
|
|
|
|
rnd = np.random.randint(0, len(self.functions[:,0])) # call the previously loaded .csv which contains all operators
|
|
tree[6][branch[n]] = self.functions[rnd][0] # replace function (operator)
|
|
|
|
elif tree[5][branch[n]] == 'term':
|
|
if self.display == 'i': print '\t\033[36m from\033[1m', tree[5][branch[n]], '\033[0;0m\033[36mto\033[1m term \033[0;0m'
|
|
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
tree[6][branch[n]] = self.terminals[rnd] # replace terminal (variable)
|
|
|
|
tree = self.fx_evolve_fitness_wipe(tree) # wipe fitness data
|
|
|
|
if self.display == 'db': print '\n\033[36m This is tourn_winner after nodes\033[1m', branch, '\033[0;0m\033[36mwere mutated and updated:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_grow_mutate(self, tree, branch):
|
|
|
|
'''
|
|
Mutate a branch of a Grow method Tree.
|
|
|
|
A branch is selected within a given tree.
|
|
|
|
If the point of mutation ('branch_top') resides at 'tree_depth_max', we do not need to grow a new tree. As the
|
|
methods for building trees always assume root (node 0) to be a function, we need only mutate this terminal node
|
|
to another terminal node, and this branch mutate method is complete.
|
|
|
|
If the top of that branch is a terminal which does not reside at 'tree_depth_max', then it may either remain a
|
|
terminal (in which case a new value is randomly assigned) or it may mutate into a function. If it becomes a
|
|
function, a new branch (mini-tree) is generated to be appended to that nodes current location. The same is true
|
|
for function-to-function mutation. Either way, the new branch will be only as deep as allowed by the distance
|
|
from it's branch_top to the bottom of the tree.
|
|
|
|
If however a function mutates into a terminal, the entire branch beneath the function is deleted from the array
|
|
and the entire array is updated, to fix parent/child links, associated arities, and node IDs.
|
|
|
|
Arguments required: tree, branch
|
|
'''
|
|
|
|
branch_top = int(branch[0]) # replaces 2 instances, below; tested 2016 07/09
|
|
branch_depth = self.tree_depth_max - int(tree[4][branch_top]) # 'tree_depth_max' - depth at 'branch_top' to set max potential size of new branch - 2016 07/10
|
|
|
|
if branch_depth < 0: # this has never occured ... yet
|
|
print '\n\t\033[31m ERROR! In fx_evolve_grow_mutate: branch_depth < 0\033[0;0m'
|
|
print '\t branch_depth =', branch_depth; self.fx_karoo_pause(0)
|
|
|
|
elif branch_depth == 0: # the point of mutation ('branch_top') chosen resides at the maximum allowable depth, so mutate term to term
|
|
|
|
if self.display == 'i': print '\t\033[36m max depth branch node\033[1m', tree[3][branch_top], '\033[0;0m\033[36mmutates from \033[1mterm\033[0;0m \033[36mto \033[1mterm\033[0;0m\n'
|
|
if self.display == 'db': print '\n\n\033[33m *** Grow Mutation: terminal to terminal *** \033[0;0m\n\n\033[36m This is the unaltered tourn_winner:\033[0;0m\n', tree
|
|
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
tree[6][branch_top] = self.terminals[rnd] # replace terminal (variable)
|
|
|
|
if self.display == 'db': print '\n\033[36m This is tourn_winner after terminal\033[1m', branch_top, '\033[0;0m\033[36mmutation, branch deletion, and updates:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
else: # the point of mutation ('branch_top') chosen is at least one degree of depth from the maximum allowed
|
|
|
|
# type_mod = '[func or term]' # TEST & DEBUG: force to 'func' or 'term' and comment the next 3 lines
|
|
rnd = np.random.randint(2)
|
|
if rnd == 0: type_mod = 'func' # randomly selected as Function
|
|
elif rnd == 1: type_mod = 'term' # randomly selected as Terminal
|
|
|
|
if type_mod == 'term': # mutate 'branch_top' to a terminal and delete all nodes beneath (no subsequent nodes are added to this branch)
|
|
|
|
if self.display == 'i': print '\t\033[36m branch node\033[1m', tree[3][branch_top], '\033[0;0m\033[36mmutates from\033[1m', tree[5][branch_top], '\033[0;0m\033[36mto\033[1m term \n\033[0;0m'
|
|
if self.display == 'db': print '\n\n\033[33m *** Grow Mutation: branch_top to terminal *** \033[0;0m\n\n\033[36m This is the unaltered tourn_winner:\033[0;0m\n', tree
|
|
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
tree[5][branch_top] = 'term' # replace type ('func' to 'term' or 'term' to 'term')
|
|
tree[6][branch_top] = self.terminals[rnd] # replace label
|
|
|
|
tree = np.delete(tree, branch[1:], axis = 1) # delete all nodes beneath point of mutation ('branch_top')
|
|
tree = self.fx_evolve_node_arity_fix(tree) # fix all node arities
|
|
tree = self.fx_evolve_child_link_fix(tree) # fix all child links
|
|
tree = self.fx_evolve_node_renum(tree) # renumber all 'NODE_ID's
|
|
|
|
if self.display == 'db': print '\n\033[36m This is tourn_winner after terminal\033[1m', branch_top, '\033[0;0m\033[36mmutation, branch deletion, and updates:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
|
|
if type_mod == 'func': # mutate 'branch_top' to a function (a new 'gp.tree' will be copied, node by node, into 'tourn_winner')
|
|
|
|
if self.display == 'i': print '\t\033[36m branch node\033[1m', tree[3][branch_top], '\033[0;0m\033[36mmutates from\033[1m', tree[5][branch_top], '\033[0;0m\033[36mto\033[1m func \n\033[0;0m'
|
|
if self.display == 'db': print '\n\n\033[33m *** Grow Mutation: branch_top to function *** \033[0;0m\n\n\033[36m This is the unaltered tourn_winner:\033[0;0m\n', tree
|
|
|
|
self.fx_gen_tree_build('mutant', self.pop_tree_type, branch_depth) # build new Tree ('gp.tree') with a maximum depth which matches 'branch'
|
|
|
|
if self.display == 'db': print '\n\033[36m This is the new Tree to be inserted at node\033[1m', branch_top, '\033[0;0m\033[36min tourn_winner:\033[0;0m\n', self.tree; self.fx_karoo_pause(0)
|
|
|
|
# because we already know the maximum depth to which this branch can grow, there is no need to prune after insertion
|
|
tree = self.fx_evolve_branch_top_copy(tree, branch) # copy root of new 'gp.tree' to point of mutation ('branch_top') in 'tree' ('tourn_winner')
|
|
tree = self.fx_evolve_branch_body_copy(tree) # copy remaining nodes in new 'gp.tree' to 'tree' ('tourn_winner')
|
|
|
|
tree = self.fx_evolve_fitness_wipe(tree) # wipe fitness data
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_crossover(self, parent, branch_x, offspring, branch_y):
|
|
|
|
'''
|
|
Refer to the method 'fx_karoo_crossover' for a full description of the genetic operator Crossover.
|
|
|
|
This method is called twice to produce 2 offspring per pair of parent Trees. Note that in the method
|
|
'karoo_fx_crossover' the parent/branch relationships are swapped from the first run to the second, such that
|
|
this method receives swapped components to produce the alternative offspring. Therefore 'parent_b' is first
|
|
passed to 'offspring' which will receive 'branch_a'. With the second run, 'parent_a' is passed to 'offspring' which
|
|
will receive 'branch_b'.
|
|
|
|
Arguments required: parent, branch_x, offspring, branch_y (parents_a / _b, branch_a / _b from 'fx_karoo_crossover')
|
|
'''
|
|
|
|
crossover = int(branch_x[0]) # pointer to the top of the 1st parent branch passed from 'fx_karoo_crossover'
|
|
branch_top = int(branch_y[0]) # pointer to the top of the 2nd parent branch passed from 'fx_karoo_crossover'
|
|
|
|
if self.display == 'db': print '\n\n\033[33m *** Crossover *** \033[0;0m'
|
|
|
|
if len(branch_x) == 1: # if the branch from the parent contains only one node (terminal)
|
|
|
|
if self.display == 'i': print '\t\033[36m terminal crossover from \033[1mparent', parent[0][1], '\033[0;0m\033[36mto \033[1moffspring', offspring[0][1], '\033[0;0m\033[36mat node\033[1m', branch_top, '\033[0;0m'
|
|
|
|
if self.display == 'db':
|
|
print '\n\033[36m In a copy of one parent:\033[0;0m\n', offspring
|
|
print '\n\033[36m ... we remove nodes\033[1m', branch_y, '\033[0;0m\033[36mand replace node\033[1m', branch_top, '\033[0;0m\033[36mwith a terminal from branch_x\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
offspring[5][branch_top] = 'term' # replace type
|
|
offspring[6][branch_top] = parent[6][crossover] # replace label with that of a particular node in 'branch_x'
|
|
offspring[8][branch_top] = 0 # set terminal arity
|
|
|
|
offspring = np.delete(offspring, branch_y[1:], axis = 1) # delete all nodes beneath point of mutation ('branch_top')
|
|
offspring = self.fx_evolve_child_link_fix(offspring) # fix all child links
|
|
offspring = self.fx_evolve_node_renum(offspring) # renumber all 'NODE_ID's
|
|
|
|
if self.display == 'db': print '\n\033[36m This is the resulting offspring:\033[0;0m\n', offspring; self.fx_karoo_pause(0)
|
|
|
|
|
|
else: # we are working with a branch from 'parent' >= depth 1 (min 3 nodes)
|
|
|
|
if self.display == 'i': print '\t\033[36m branch crossover from \033[1mparent', parent[0][1], '\033[0;0m\033[36mto \033[1moffspring', offspring[0][1], '\033[0;0m\033[36mat node\033[1m', branch_top, '\033[0;0m'
|
|
|
|
# self.fx_gen_tree_build('test', 'f', 2) # TEST & DEBUG: disable the next 'self.tree ...' line
|
|
self.tree = self.fx_evolve_branch_copy(parent, branch_x) # generate stand-alone 'gp.tree' with properties of 'branch_x'
|
|
|
|
if self.display == 'db':
|
|
print '\n\033[36m From one parent:\033[0;0m\n', parent
|
|
print '\n\033[36m ... we copy branch_x\033[1m', branch_x, '\033[0;0m\033[36mas a new, sub-tree:\033[0;0m\n', self.tree; self.fx_karoo_pause(0)
|
|
|
|
if self.display == 'db':
|
|
print '\n\033[36m ... and insert it into a copy of the second parent in place of the selected branch\033[1m', branch_y,':\033[0;0m\n', offspring; self.fx_karoo_pause(0)
|
|
|
|
offspring = self.fx_evolve_branch_top_copy(offspring, branch_y) # copy root of 'branch_y' ('gp.tree') to 'offspring'
|
|
offspring = self.fx_evolve_branch_body_copy(offspring) # copy remaining nodes in 'branch_y' ('gp.tree') to 'offspring'
|
|
offspring = self.fx_evolve_tree_prune(offspring, self.tree_depth_max) # prune to the max Tree depth + adjustment - tested 2016 07/10
|
|
|
|
offspring = self.fx_evolve_fitness_wipe(offspring) # wipe fitness data
|
|
|
|
return offspring
|
|
|
|
|
|
def fx_evolve_branch_select(self, tree):
|
|
|
|
'''
|
|
Select all nodes in the 'tourn_winner' Tree at and below the randomly selected starting point.
|
|
|
|
While Grow mutation uses this method to select a region of the 'tourn_winner' to delete, Crossover uses this
|
|
method to select a region of the 'tourn_winner' which is then converted to a stand-alone tree. As such, it is
|
|
imperative that the nodes be in the correct order, else all kinds of bad things happen.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
branch = np.array([]) # the array is necessary in order to len(branch) when 'branch' has only one element
|
|
branch_top = np.random.randint(2, len(tree[3])) # randomly select a non-root node
|
|
branch_eval = self.fx_eval_id(tree, branch_top) # generate tuple of 'branch_top' and subseqent nodes
|
|
branch_symp = sympify(branch_eval) # convert string into something useful
|
|
branch = np.append(branch, branch_symp) # append list to array
|
|
|
|
branch = np.sort(branch) # sort nodes in branch for Crossover.
|
|
|
|
if self.display == 'i': print '\t \033[36mwith nodes\033[1m', branch, '\033[0;0m\033[36mchosen for mutation\033[0;0m'
|
|
|
|
return branch
|
|
|
|
|
|
def fx_evolve_branch_top_copy(self, tree, branch):
|
|
|
|
'''
|
|
Copy the point of mutation ('branch_top') from 'gp.tree' to 'tree'.
|
|
|
|
This method works with 3 inputs: local 'tree' is being modified; local 'branch' is a section of 'tree' which
|
|
will be removed; and global 'gp.tree' (recycling from initial population generation) is the new Tree to be
|
|
copied into 'tree', replacing 'branch'.
|
|
|
|
This method is used in both Grow Mutation and Crossover.
|
|
|
|
Arguments required: tree, branch
|
|
'''
|
|
|
|
branch_top = int(branch[0])
|
|
|
|
tree[5][branch_top] = 'func' # update type ('func' to 'term' or 'term' to 'term'); this modifies gp.tree[5[1] from 'root' to 'func'
|
|
tree[6][branch_top] = self.tree[6][1] # copy node_label from new tree
|
|
tree[8][branch_top] = self.tree[8][1] # copy node_arity from new tree
|
|
|
|
tree = np.delete(tree, branch[1:], axis = 1) # delete all nodes beneath point of mutation ('branch_top')
|
|
|
|
c_buffer = self.fx_evolve_c_buffer(tree, branch_top) # generate c_buffer for point of mutation ('branch_top')
|
|
tree = self.fx_evolve_child_insert(tree, branch_top, c_buffer) # insert new nodes
|
|
tree = self.fx_evolve_node_renum(tree) # renumber all 'NODE_ID's
|
|
|
|
if self.display == 'db':
|
|
print '\n\t ... inserted node 1 of', len(self.tree[3])-1
|
|
print '\n\033[36m This is the Tree after a new node is inserted:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_branch_body_copy(self, tree):
|
|
|
|
'''
|
|
Copy the body of 'gp.tree' to 'tree', one node at a time.
|
|
|
|
This method works with 3 inputs: local 'tree' is being modified; local 'branch' is a section of 'tree' which
|
|
will be removed; and global 'gp.tree' (recycling from initial population generation) is the new Tree to be
|
|
copied into 'tree', replacing 'branch'.
|
|
|
|
This method is used in both Grow Mutation and Crossover.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
node_count = 2 # set node count for 'gp.tree' to 2 as the new root has already replaced 'branch_top' in 'fx_evolve_branch_top_copy'
|
|
|
|
while node_count < len(self.tree[3]): # increment through all nodes in the new Tree ('gp.tree'), starting with node 2
|
|
|
|
for j in range(1, len(tree[3])): # increment through all nodes in tourn_winner ('tree')
|
|
|
|
if self.display == 'db': print '\tScanning tourn_winner node_id:', j
|
|
|
|
if tree[5][j] == '':
|
|
tree[5][j] = self.tree[5][node_count] # copy 'node_type' from branch to tree
|
|
tree[6][j] = self.tree[6][node_count] # copy 'node_label' from branch to tree
|
|
tree[8][j] = self.tree[8][node_count] # copy 'node_arity' from branch to tree
|
|
|
|
if tree[5][j] == 'term':
|
|
tree = self.fx_evolve_child_link_fix(tree) # fix all child links
|
|
tree = self.fx_evolve_node_renum(tree) # renumber all 'NODE_ID's
|
|
|
|
if tree[5][j] == 'func':
|
|
c_buffer = self.fx_evolve_c_buffer(tree, j) # generate 'c_buffer' for point of mutation ('branch_top')
|
|
tree = self.fx_evolve_child_insert(tree, j, c_buffer) # insert new nodes
|
|
tree = self.fx_evolve_child_link_fix(tree) # fix all child links
|
|
tree = self.fx_evolve_node_renum(tree) # renumber all 'NODE_ID's
|
|
|
|
if self.display == 'db':
|
|
print '\n\t ... inserted node', node_count, 'of', len(self.tree[3])-1
|
|
print '\n\033[36m This is the Tree after a new node is inserted:\033[0;0m\n', tree; self.fx_karoo_pause(0)
|
|
|
|
node_count = node_count + 1 # exit loop when 'node_count' reaches the number of columns in the array 'gp.tree'
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_branch_copy(self, tree, branch):
|
|
|
|
'''
|
|
This method prepares a stand-alone Tree as a copy of the given branch.
|
|
|
|
This method is used with Crossover.
|
|
|
|
Arguments required: tree, branch
|
|
'''
|
|
|
|
new_tree = np.array([ ['TREE_ID'],['tree_type'],['tree_depth_base'],['NODE_ID'],['node_depth'],['node_type'],['node_label'],['node_parent'],['node_arity'],['node_c1'],['node_c2'],['node_c3'],['fitness'] ])
|
|
|
|
# tested 2015 06/08
|
|
for n in range(len(branch)):
|
|
|
|
node = branch[n]
|
|
branch_top = int(branch[0])
|
|
|
|
TREE_ID = 'copy'
|
|
tree_type = tree[1][1]
|
|
tree_depth_base = int(tree[4][branch[-1]]) - int(tree[4][branch_top]) # subtract depth of 'branch_top' from the last in 'branch'
|
|
NODE_ID = tree[3][node]
|
|
node_depth = int(tree[4][node]) - int(tree[4][branch_top]) # subtract the depth of 'branch_top' from the current node depth
|
|
node_type = tree[5][node]
|
|
node_label = tree[6][node]
|
|
node_parent = '' # updated by 'fx_evolve_parent_link_fix', below
|
|
node_arity = tree[8][node]
|
|
node_c1 = '' # updated by 'fx_evolve_child_link_fix', below
|
|
node_c2 = ''
|
|
node_c3 = ''
|
|
fitness = ''
|
|
|
|
new_tree = np.append(new_tree, [ [TREE_ID],[tree_type],[tree_depth_base],[NODE_ID],[node_depth],[node_type],[node_label],[node_parent],[node_arity],[node_c1],[node_c2],[node_c3],[fitness] ], 1)
|
|
|
|
new_tree = self.fx_evolve_node_renum(new_tree)
|
|
new_tree = self.fx_evolve_child_link_fix(new_tree)
|
|
new_tree = self.fx_evolve_parent_link_fix(new_tree)
|
|
new_tree = self.fx_archive_tree_clean(new_tree)
|
|
|
|
return new_tree
|
|
|
|
|
|
def fx_evolve_c_buffer(self, tree, node):
|
|
|
|
'''
|
|
This method serves the very important function of determining the links from parent to child for any given
|
|
node. The single, simple formula [parent_arity_sum + prior_sibling_arity - prior_siblings] perfectly determines
|
|
the correct position of the child node, already in place or to be inserted, no matter the depth nor complexity
|
|
of the tree.
|
|
|
|
This method is currently called from the evolution methods, but will soon (I hope) be called from the first
|
|
generation Tree generation methods (above) such that the same method may be used repeatedly.
|
|
|
|
Arguments required: tree, node
|
|
'''
|
|
|
|
parent_arity_sum = 0
|
|
prior_sibling_arity = 0
|
|
prior_siblings = 0
|
|
|
|
for n in range(1, len(tree[3])): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if int(tree[4][n]) == int(tree[4][node])-1: # find parent nodes at the prior depth
|
|
if tree[8][n] != '': parent_arity_sum = parent_arity_sum + int(tree[8][n]) # sum arities of all parent nodes at the prior depth
|
|
|
|
if int(tree[4][n]) == int(tree[4][node]) and int(tree[3][n]) < int(tree[3][node]): # find prior siblings at the current depth
|
|
if tree[8][n] != '': prior_sibling_arity = prior_sibling_arity + int(tree[8][n]) # sum prior sibling arity
|
|
prior_siblings = prior_siblings + 1 # sum quantity of prior siblings
|
|
|
|
c_buffer = node + (parent_arity_sum + prior_sibling_arity - prior_siblings) # One algo to rule the world!
|
|
|
|
return c_buffer
|
|
|
|
|
|
def fx_evolve_child_link(self, tree, node, c_buffer):
|
|
|
|
'''
|
|
Link each parent node to its children.
|
|
|
|
Arguments required: tree, node, c_buffer
|
|
'''
|
|
|
|
if int(tree[3][node]) == 1: c_buffer = c_buffer + 1 # if root (node 1) is passed through this method
|
|
|
|
if tree[8][node] != '':
|
|
|
|
if int(tree[8][node]) == 0: # if arity = 0
|
|
tree[9][node] = ''
|
|
tree[10][node] = ''
|
|
tree[11][node] = ''
|
|
|
|
elif int(tree[8][node]) == 1: # if arity = 1
|
|
tree[9][node] = c_buffer
|
|
tree[10][node] = ''
|
|
tree[11][node] = ''
|
|
|
|
elif int(tree[8][node]) == 2: # if arity = 2
|
|
tree[9][node] = c_buffer
|
|
tree[10][node] = c_buffer + 1
|
|
tree[11][node] = ''
|
|
|
|
elif int(tree[8][node]) == 3: # if arity = 3
|
|
tree[9][node] = c_buffer
|
|
tree[10][node] = c_buffer + 1
|
|
tree[11][node] = c_buffer + 2
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_evolve_child_link: node', node, 'has arity', tree[8][node]; self.fx_karoo_pause(0)
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_child_link_fix(self, tree):
|
|
|
|
'''
|
|
In a given Tree, fix 'node_c1', 'node_c2', 'node_c3' for all nodes.
|
|
|
|
This is required anytime the size of the array 'gp.tree' has been modified, as with both Grow and Full mutation.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
# tested 2015 06/04
|
|
for node in range(1, len(tree[3])):
|
|
|
|
c_buffer = self.fx_evolve_c_buffer(tree, node) # generate c_buffer for each node
|
|
tree = self.fx_evolve_child_link(tree, node, c_buffer) # update child links for each node
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_child_insert(self, tree, node, c_buffer):
|
|
|
|
'''
|
|
Insert child nodes.
|
|
|
|
Arguments required: tree, node, c_buffer
|
|
'''
|
|
|
|
if int(tree[8][node]) == 0: # if arity = 0
|
|
print '\n\t\033[31m ERROR! In fx_evolve_child_insert: node', node, 'has arity 0\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
elif int(tree[8][node]) == 1: # if arity = 1
|
|
tree = np.insert(tree, c_buffer, '', axis=1) # insert node for 'node_c1'
|
|
tree[3][c_buffer] = c_buffer # node ID
|
|
tree[4][c_buffer] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer] = int(tree[3][node]) # parent ID
|
|
|
|
elif int(tree[8][node]) == 2: # if arity = 2
|
|
tree = np.insert(tree, c_buffer, '', axis=1) # insert node for 'node_c1'
|
|
tree[3][c_buffer] = c_buffer # node ID
|
|
tree[4][c_buffer] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer] = int(tree[3][node]) # parent ID
|
|
|
|
tree = np.insert(tree, c_buffer + 1, '', axis=1) # insert node for 'node_c2'
|
|
tree[3][c_buffer + 1] = c_buffer + 1 # node ID
|
|
tree[4][c_buffer + 1] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer + 1] = int(tree[3][node]) # parent ID
|
|
|
|
elif int(tree[8][node]) == 3: # if arity = 3
|
|
tree = np.insert(tree, c_buffer, '', axis=1) # insert node for 'node_c1'
|
|
tree[3][c_buffer] = c_buffer # node ID
|
|
tree[4][c_buffer] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer] = int(tree[3][node]) # parent ID
|
|
|
|
tree = np.insert(tree, c_buffer + 1, '', axis=1) # insert node for 'node_c2'
|
|
tree[3][c_buffer + 1] = c_buffer + 1 # node ID
|
|
tree[4][c_buffer + 1] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer + 1] = int(tree[3][node]) # parent ID
|
|
|
|
tree = np.insert(tree, c_buffer + 2, '', axis=1) # insert node for 'node_c3'
|
|
tree[3][c_buffer + 2] = c_buffer + 2 # node ID
|
|
tree[4][c_buffer + 2] = int(tree[4][node]) + 1 # node_depth
|
|
tree[7][c_buffer + 2] = int(tree[3][node]) # parent ID
|
|
|
|
else: print '\n\t\033[31m ERROR! In fx_evolve_child_insert: node', node, 'arity > 3\033[0;0m'; self.fx_karoo_pause(0)
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_parent_link_fix(self, tree):
|
|
|
|
'''
|
|
In a given Tree, fix 'parent_id' for all nodes.
|
|
|
|
This is automatically handled in all mutations except with Crossover due to the need to copy branches 'a' and
|
|
'b' to their own trees before inserting them into copies of the parents.
|
|
|
|
Technically speaking, the 'node_parent' value is not used by any methods. The parent ID can be completely out
|
|
of whack and the expression will work perfectly. This is maintained for the sole purpose of granting the user
|
|
a friendly, makes-sense interface which can be read in both directions.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
### THIS METHOD MAY NOT BE REQUIRED AS SORTING 'branch' SEEMS TO HAVE FIXED 'parent_id' ###
|
|
|
|
# tested 2015 06/05
|
|
for node in range(1, len(tree[3])):
|
|
|
|
if tree[9][node] != '':
|
|
child = int(tree[9][node])
|
|
tree[7][child] = node
|
|
|
|
if tree[10][node] != '':
|
|
child = int(tree[10][node])
|
|
tree[7][child] = node
|
|
|
|
if tree[11][node] != '':
|
|
child = int(tree[11][node])
|
|
tree[7][child] = node
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_node_arity_fix(self, tree):
|
|
|
|
'''
|
|
In a given Tree, fix 'node_arity' for all nodes labeled 'term' but with arity 2.
|
|
|
|
This is required after a function has been replaced by a terminal, as may occur with both Grow mutation and
|
|
Crossover.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
# tested 2015 05/31
|
|
for n in range(1, len(tree[3])): # increment through all nodes (exclude 0) in array 'tree'
|
|
|
|
if tree[5][n] == 'term': # check for discrepency
|
|
tree[8][n] = '0' # set arity to 0
|
|
tree[9][n] = '' # wipe 'node_c1'
|
|
tree[10][n] = '' # wipe 'node_c2'
|
|
tree[11][n] = '' # wipe 'node_c3'
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_node_renum(self, tree):
|
|
|
|
'''
|
|
Renumber all 'NODE_ID' in a given tree.
|
|
|
|
This is required after a new generation is evolved as the NODE_ID numbers are carried forward from the previous
|
|
generation but are no longer in order.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
for n in range(1, len(tree[3])):
|
|
|
|
tree[3][n] = n # renumber all Trees in given population
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_fitness_wipe(self, tree):
|
|
|
|
'''
|
|
Remove all fitness data from a given tree.
|
|
|
|
This is required after a new generation is evolved as the fitness of the same Tree prior to its mutation will
|
|
no longer apply.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
tree[12][1:] = '' # wipe fitness data
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_tree_prune(self, tree, depth):
|
|
|
|
'''
|
|
This method reduces the depth of a Tree. Used with Crossover, the input value 'branch' can be a partial Tree
|
|
(branch) or a full tree, and it will operate correctly. The input value 'depth' becomes the new maximum depth,
|
|
where depth is defined as the local maximum + the user defined adjustment.
|
|
|
|
Arguments required: tree, depth
|
|
'''
|
|
|
|
nodes = []
|
|
|
|
# tested 2015 06/08
|
|
for n in range(1, len(tree[3])):
|
|
|
|
if int(tree[4][n]) == depth and tree[5][n] == 'func':
|
|
rnd = np.random.randint(0, len(self.terminals) - 1) # call the previously loaded .csv which contains all terminals
|
|
tree[5][n] = 'term' # mutate type 'func' to 'term'
|
|
tree[6][n] = self.terminals[rnd] # replace label
|
|
|
|
elif int(tree[4][n]) > depth: # record nodes deeper than the maximum allowed Tree depth
|
|
nodes.append(n)
|
|
|
|
else: pass # as int(tree[4][n]) < depth and will remain untouched
|
|
|
|
tree = np.delete(tree, nodes, axis = 1) # delete nodes deeper than the maximum allowed Tree depth
|
|
tree = self.fx_evolve_node_arity_fix(tree) # fix all node arities
|
|
|
|
return tree
|
|
|
|
|
|
def fx_evolve_tree_renum(self, population):
|
|
|
|
'''
|
|
Renumber all 'TREE_ID' in a given population.
|
|
|
|
This is required after a new generation is evolved as the TREE_ID numbers are carried forward from the previous
|
|
generation but are no longer in order.
|
|
|
|
Arguments required: population
|
|
'''
|
|
|
|
for tree_id in range(1, len(population)):
|
|
|
|
population[tree_id][0][1] = tree_id # renumber all Trees in given population
|
|
|
|
return population
|
|
|
|
|
|
def fx_evolve_pop_copy(self, pop_a, title):
|
|
|
|
'''
|
|
Copy one population to another.
|
|
|
|
Simply copying a list of arrays generates a pointer to the original list. Therefore we must append each array
|
|
to a new, empty array and then build a list of those new arrays.
|
|
|
|
Arguments required: pop_a, title
|
|
'''
|
|
|
|
pop_b = [title] # an empty list stores a copy of the prior generation
|
|
|
|
for tree in range(1, len(pop_a)): # increment through each Tree in the current population
|
|
|
|
tree_copy = np.copy(pop_a[tree]) # copy each array in the current population
|
|
pop_b.append(tree_copy) # add each copied Tree to the new population list
|
|
|
|
return pop_b
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Display a Tree |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_display_tree(self, tree):
|
|
|
|
'''
|
|
Display all or part of a Tree on-screen.
|
|
|
|
This method displays all sequential node_ids from 'start' node through bottom, within the given tree.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
ind = ''
|
|
print '\n\033[1m\033[36m Tree ID', int(tree[0][1]), '\033[0;0m'
|
|
|
|
for depth in range(0, self.tree_depth_max + 1): # increment through all possible Tree depths - tested 2016 07/09
|
|
print '\n', ind,'\033[36m Tree Depth:', depth, 'of', tree[2][1], '\033[0;0m'
|
|
|
|
for node in range(1, len(tree[3])): # increment through all nodes (redundant, I know)
|
|
if int(tree[4][node]) == depth:
|
|
print ''
|
|
print ind,'\033[1m\033[36m NODE:', tree[3][node], '\033[0;0m'
|
|
print ind,' type:', tree[5][node]
|
|
print ind,' label:', tree[6][node], '\tparent node:', tree[7][node]
|
|
print ind,' arity:', tree[8][node], '\tchild node(s):', tree[9][node], tree[10][node], tree[11][node]
|
|
|
|
ind = ind + '\t'
|
|
|
|
print ''
|
|
self.fx_eval_poly(tree) # generate the raw and sympified equation for the entire Tree
|
|
print '\t\033[36mTree', tree[0][1], 'yields (raw):', self.algo_raw, '\033[0;0m'
|
|
print '\t\033[36mTree', tree[0][1], 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m'
|
|
|
|
return
|
|
|
|
|
|
def fx_display_branch(self, tree, start):
|
|
|
|
'''
|
|
Display a Tree branch on-screen.
|
|
|
|
This method displays all sequential node_ids from 'start' node through bottom, within the given branch.
|
|
|
|
This method is not used by Karoo GP at this time.
|
|
|
|
Arguments required: tree, start
|
|
'''
|
|
|
|
branch = np.array([]) # the array is necessary in order to len(branch) when 'branch' has only one element
|
|
branch_eval = self.fx_eval_id(tree, start) # generate tuple of given 'branch'
|
|
branch_symp = sympify(branch_eval) # convert string from tuple to list
|
|
branch = np.append(branch, branch_symp) # append list to array
|
|
ind = ''
|
|
|
|
# for depth in range(int(tree[4][start]), int(tree[2][1]) + self.tree_depth_max + 1): # increment through all Tree depths - tested 2016 07/09
|
|
for depth in range(int(tree[4][start]), self.tree_depth_max + 1): # increment through all Tree depths - tested 2016 07/09
|
|
print '\n', ind,'\033[36m Tree Depth:', depth, 'of', tree[2][1], '\033[0;0m'
|
|
|
|
for n in range(0, len(branch)): # increment through all nodes listed in the branch
|
|
node = branch[n]
|
|
|
|
if int(tree[4][node]) == depth:
|
|
print ''
|
|
print ind,'\033[1m\033[36m NODE:', node, '\033[0;0m'
|
|
print ind,' type:', tree[5][node]
|
|
print ind,' label:', tree[6][node], '\tparent node:', tree[7][node]
|
|
print ind,' arity:', tree[8][node], '\tchild node(s):', tree[9][node], tree[10][node], tree[11][node]
|
|
|
|
ind = ind + '\t'
|
|
|
|
print ''
|
|
self.fx_eval_poly(tree) # generate the raw and sympified equation for the entire Tree
|
|
print '\t\033[36mTree', tree[0][1], 'yields (raw):', self.algo_raw, '\033[0;0m'
|
|
print '\t\033[36mTree', tree[0][1], 'yields (sym):\033[1m', self.algo_sym, '\033[0;0m'
|
|
|
|
return
|
|
|
|
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
# Methods to Archive |
|
|
#++++++++++++++++++++++++++++++++++++++++++
|
|
|
|
def fx_archive_tree_clean(self, tree):
|
|
|
|
'''
|
|
This method aesthetically cleans the Tree array, removing redundant data.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
tree[0][2:] = '' # A little clean-up to make things look pretty :)
|
|
tree[1][2:] = '' # Ignore the man behind the curtain!
|
|
tree[2][2:] = '' # Yes, I am a bit OCD ... but you *know* you appreciate clean arrays.
|
|
|
|
return tree
|
|
|
|
|
|
def fx_archive_tree_append(self, tree):
|
|
|
|
'''
|
|
Append Tree array to the foundation Population.
|
|
|
|
Arguments required: tree
|
|
'''
|
|
|
|
self.fx_archive_tree_clean(tree) # clean 'tree' prior to storing
|
|
self.population_a.append(tree) # append 'tree' to population list
|
|
|
|
return
|
|
|
|
|
|
def fx_archive_tree_write(self, population, key):
|
|
|
|
'''
|
|
Save population_* to disk.
|
|
|
|
Arguments required: population, key
|
|
'''
|
|
|
|
with open(self.filename[key], 'a') as csv_file:
|
|
target = csv.writer(csv_file, delimiter=',')
|
|
if self.generation_id != 1: target.writerows(['']) # empty row before each generation
|
|
target.writerows([['Karoo GP by Kai Staats', 'Generation:', str(self.generation_id)]])
|
|
|
|
for tree in range(1, len(population)):
|
|
target.writerows(['']) # empty row before each Tree
|
|
for row in range(0, 13): # increment through each row in the array Tree
|
|
target.writerows([population[tree][row]])
|
|
|
|
|
|
def fx_archive_params_write(self, app): # tested 2017 02/13
|
|
|
|
'''
|
|
Save run-time configuration parameters to disk.
|
|
|
|
Arguments required: none
|
|
'''
|
|
|
|
file = open(self.path + '/log_config.txt', 'w')
|
|
file.write('Karoo GP ' + app)
|
|
file.write('\n launched: ' + str(self.datetime))
|
|
file.write('\n dataset: ' + str(self.dataset))
|
|
file.write('\n')
|
|
file.write('\n kernel: ' + str(self.kernel))
|
|
file.write('\n precision: ' + str(self.precision))
|
|
file.write('\n')
|
|
# file.write('tree type: ' + tree_type)
|
|
# file.write('tree depth base: ' + str(tree_depth_base))
|
|
file.write('\n tree depth max: ' + str(self.tree_depth_max))
|
|
file.write('\n min node count: ' + str(self.tree_depth_min))
|
|
file.write('\n')
|
|
file.write('\n genetic operator Reproduction: ' + str(self.evolve_repro))
|
|
file.write('\n genetic operator Point Mutation: ' + str(self.evolve_point))
|
|
file.write('\n genetic operator Branch Mutation: ' + str(self.evolve_branch))
|
|
file.write('\n genetic operator Crossover: ' + str(self.evolve_cross))
|
|
file.write('\n')
|
|
file.write('\n tournament size: ' + str(self.tourn_size))
|
|
file.write('\n population: ' + str(self.tree_pop_max))
|
|
file.write('\n number of generations: ' + str(self.generation_id))
|
|
file.write('\n\n')
|
|
file.close()
|
|
|
|
|
|
file = open(self.path + '/log_test.txt', 'w')
|
|
file.write('Karoo GP ' + app)
|
|
file.write('\n launched: ' + str(self.datetime))
|
|
file.write('\n dataset: ' + str(self.dataset))
|
|
file.write('\n')
|
|
|
|
if len(self.fittest_dict) > 0:
|
|
|
|
fitness_best = 0
|
|
fittest_tree = 0
|
|
|
|
# original method, using pre-built fittest_dict
|
|
# file.write('\n The leading Trees and their associated expressions are:')
|
|
# for n in sorted(self.fittest_dict):
|
|
# file.write('\n\t ' + str(n) + ' : ' + str(self.fittest_dict[n]))
|
|
|
|
# revised method, re-evaluating all Trees from stored fitness score
|
|
for tree_id in range(1, len(self.population_b)):
|
|
|
|
fitness = float(self.population_b[tree_id][12][1])
|
|
|
|
if self.kernel == 'c': # display best fit Trees for the CLASSIFY kernel
|
|
if fitness >= fitness_best: # find the Tree with Maximum fitness score
|
|
fitness_best = fitness; fittest_tree = tree_id # set best fitness Tree
|
|
|
|
elif self.kernel == 'r': # display best fit Trees for the REGRESSION kernel
|
|
if fitness_best == 0: fitness_best = fitness # set the baseline first time through
|
|
if fitness <= fitness_best: # find the Tree with Minimum fitness score
|
|
fitness_best = fitness; fittest_tree = tree_id # set best fitness Tree
|
|
|
|
elif self.kernel == 'm': # display best fit Trees for the MATCH kernel
|
|
if fitness == self.data_train_rows: # find the Tree with a perfect match for all data rows
|
|
fitness_best = fitness; fittest_tree = tree_id # set best fitness Tree
|
|
|
|
# elif self.kernel == '[other]': # display best fit Trees for the [other] kernel
|
|
# if fitness [>=, <=] fitness_best: # find the Tree with [Maximum or Minimum] fitness score
|
|
# fitness_best = fitness; fittest_tree = tree_id # set best fitness Tree
|
|
|
|
# print 'fitness_best:', fitness_best, 'fittest_tree:', fittest_tree
|
|
|
|
|
|
# test the most fit Tree and write to the .txt log
|
|
self.fx_eval_poly(self.population_b[int(fittest_tree)]) # generate the raw and sympified equation for the given Tree using SymPy
|
|
expr = str(self.algo_sym) # get simplified expression and process it by TF - tested 2017 02/02
|
|
result = self.fx_fitness_eval(expr, self.data_test, get_labels=True)
|
|
|
|
file.write('\n\n Tree ' + str(fittest_tree) + ' is the most fit, with expression:')
|
|
file.write('\n\n ' + str(self.algo_sym))
|
|
|
|
if self.kernel == 'c':
|
|
file.write('\n\n Classification fitness score: {}'.format(result['fitness']))
|
|
file.write('\n\n Precision-Recall report:\n {}'.format(skm.classification_report(result['solution'], result['labels'][0])))
|
|
file.write('\n Confusion matrix:\n {}'.format(skm.confusion_matrix(result['solution'], result['labels'][0])))
|
|
|
|
elif self.kernel == 'r':
|
|
MSE, fitness = skm.mean_squared_error(result['result'], result['solution']), result['fitness']
|
|
file.write('\n\n Regression fitness score: {}'.format(fitness))
|
|
file.write('\n Mean Squared Error: {}'.format(MSE))
|
|
|
|
elif self.kernel == 'm':
|
|
file.write('\n\n Matching fitness score: {}'.format(result['fitness']))
|
|
|
|
# elif self.kernel == '[other]':
|
|
# file.write( ... )
|
|
|
|
else: file.write('\n\n There were no evolved solutions generated in this run... your species has gone extinct!')
|
|
|
|
file.write('\n\n')
|
|
file.close()
|
|
|
|
return
|
|
|
|
|