gradient_free_optimizers.optimizers.local_opt.downhill_simplex.DownhillSimplexOptimizer.iterate() - Code Metrics - Inspection of "v1.0.0" - SimonBlanke/Gradient-Free-Optimizers - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( 626f23...be3b1e )

by Simon

created 2021-12-01 10:57 UTC

DownhillSimplexOptimizer.iterate() B

↳ Parent: gradient_free_optimizers.optimizers.local_opt.downhill_simplex

Complexity

Conditions

Size

Total Lines	48
Code Lines	32

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	6
eloc	32
nop	1
dl	0
loc	48
rs	8.1786
c	0
b	0
f	0

# Author: Simon Blanke
# Email: [email protected]
# License: MIT License


import random
import numpy as np

from ..base_optimizer import BaseOptimizer
from ...search import Search


def sort_list_idx(list_):
    list_np = np.array(list_)
    idx_sorted = list(list_np.argsort()[::-1])
    return idx_sorted


def centeroid(array_list):
    centeroid = []
    for idx in range(array_list[0].shape[0]):
        center_dim_pos = []
        for array in array_list:
            center_dim_pos.append(array[idx])

        center_dim_mean = np.array(center_dim_pos).mean()
        centeroid.append(center_dim_mean)

    return centeroid


class DownhillSimplexOptimizer(BaseOptimizer, Search):
    name = "Downhill Simplex Optimizer"

    def __init__(
        self,
        *args,
        alpha=1,
        gamma=2,
        beta=0.5,
        sigma=0.5,
        **kwargs,
    ):
        super().__init__(*args, **kwargs)

        self.alpha = alpha
        self.gamma = gamma
        self.beta = beta
        self.sigma = sigma

        self.n_simp_positions = len(self.conv.search_space) + 1
        self.simp_positions = []

        self.simplex_step = 0

    def finish_initialization(self):
        idx_sorted = sort_list_idx(self.scores_valid)
        self.simplex_pos = [self.positions_valid[idx] for idx in idx_sorted]
        self.simplex_scores = [self.scores_valid[idx] for idx in idx_sorted]

        n_inits = len(self.positions_valid)
        if n_inits < self.n_simp_positions:
            print("\n Error: Not enough initial positions to form simplex")
            raise ValueError("Increase number of initial positions")

        self.simplex_step = 1

        self.i_x_0 = 0
        self.i_x_N_1 = -2
        self.i_x_N = -1

    @BaseOptimizer.track_nth_iter
    def iterate(self):
        simplex_stale = all(
            [np.array_equal(self.simplex_pos[0], array) for array in self.simplex_pos]
        )

        if simplex_stale:
            idx_sorted = sort_list_idx(self.scores_valid)
            self.simplex_pos = [self.positions_valid[idx] for idx in idx_sorted]
            self.simplex_scores = [self.scores_valid[idx] for idx in idx_sorted]

            self.simplex_step = 1

        if self.simplex_step == 1:
            idx_sorted = sort_list_idx(self.simplex_scores)
            self.simplex_pos = [self.simplex_pos[idx] for idx in idx_sorted]
            self.simplex_scores = [self.simplex_scores[idx] for idx in idx_sorted]

            self.center_array = centeroid(self.simplex_pos[:-1])

            r_pos = self.center_array + self.alpha * (
                self.center_array - self.simplex_pos[-1]
            )
            self.r_pos = self.conv2pos(r_pos)
            return self.r_pos

        elif self.simplex_step == 2:
            e_pos = self.center_array + self.gamma * (
                self.center_array - self.simplex_pos[-1]
            )
            self.e_pos = self.conv2pos(e_pos)
            self.simplex_step = 1

            return self.e_pos

        elif self.simplex_step == 3:
            # iter Contraction
            c_pos = self.h_pos + self.beta * (self.center_array - self.h_pos)
            c_pos = self.conv2pos(c_pos)

            return c_pos

        elif self.simplex_step == 4:
            # iter Shrink
            pos = self.simplex_pos[self.compress_idx]
            pos = pos + self.sigma * (self.simplex_pos[0] - pos)

            return self.conv2pos(pos)

    def evaluate(self, score_new):
        self.score_new = score_new

        if self.simplex_step != 0:
            self.prev_pos = self.positions_valid[-1]

        if self.simplex_step == 1:
            # self.r_pos = self.prev_pos
            self.r_score = score_new

            if self.r_score > self.simplex_scores[0]:
                self.simplex_step = 2

            elif self.r_score > self.simplex_scores[-2]:
                # if r is better than x N-1
                self.simplex_pos[-1] = self.r_pos
                self.simplex_scores[-1] = self.r_score
                self.simplex_step = 1

            if self.simplex_scores[-1] > self.r_score:
                self.h_pos = self.simplex_pos[-1]
                self.h_score = self.simplex_scores[-1]
            else:
                self.h_pos = self.r_pos
                self.h_score = self.r_score

            self.simplex_step = 3

        elif self.simplex_step == 2:
            self.e_score = score_new

            if self.e_score > self.r_score:
                self.simplex_scores[-1] = self.e_pos
            elif self.r_score > self.e_score:
                self.simplex_scores[-1] = self.r_pos
            else:
                self.simplex_scores[-1] = random.choice([self.e_pos, self.r_pos])[0]

        elif self.simplex_step == 3:
            # eval Contraction
            self.c_pos = self.prev_pos
            self.c_score = score_new

            if self.c_score > self.simplex_scores[-1]:
                self.simplex_scores[-1] = self.c_score
                self.simplex_pos[-1] = self.c_pos

                self.simplex_step = 1

            else:
                # start Shrink
                self.simplex_step = 4
                self.compress_idx = 0

        elif self.simplex_step == 4:
            # eval Shrink
            self.simplex_scores[self.compress_idx] = score_new
            self.simplex_pos[self.compress_idx] = self.prev_pos

            self.compress_idx += 1

            if self.compress_idx == self.n_simp_positions:
                self.simplex_step = 1


1			# Author: Simon Blanke
2			# Email: [email protected]
3			# License: MIT License
4
5
6			import random
7			import numpy as np
8
9			from ..base_optimizer import BaseOptimizer
10			from ...search import Search
11
12
13			def sort_list_idx(list_):
14			list_np = np.array(list_)
15			idx_sorted = list(list_np.argsort()[::-1])
16			return idx_sorted
17
18
19			def centeroid(array_list):
20			centeroid = []
21			for idx in range(array_list[0].shape[0]):
22			center_dim_pos = []
23			for array in array_list:
24			center_dim_pos.append(array[idx])
25
26			center_dim_mean = np.array(center_dim_pos).mean()
27			centeroid.append(center_dim_mean)
28
29			return centeroid
30
31
32			class DownhillSimplexOptimizer(BaseOptimizer, Search):
33			name = "Downhill Simplex Optimizer"
34
35			def __init__(
36			self,
37			*args,
38			alpha=1,
39			gamma=2,
40			beta=0.5,
41			sigma=0.5,
42			**kwargs,
43			):
44			super().__init__(args, *kwargs)
45
46			self.alpha = alpha
47			self.gamma = gamma
48			self.beta = beta
49			self.sigma = sigma
50
51			self.n_simp_positions = len(self.conv.search_space) + 1
52			self.simp_positions = []
53
54			self.simplex_step = 0
55
56			def finish_initialization(self):
57			idx_sorted = sort_list_idx(self.scores_valid)
58			self.simplex_pos = [self.positions_valid[idx] for idx in idx_sorted]
59			self.simplex_scores = [self.scores_valid[idx] for idx in idx_sorted]
60
61			n_inits = len(self.positions_valid)
62			if n_inits < self.n_simp_positions:
63			print("\n Error: Not enough initial positions to form simplex")
64			raise ValueError("Increase number of initial positions")
65
66			self.simplex_step = 1
67
68			self.i_x_0 = 0
69			self.i_x_N_1 = -2
70			self.i_x_N = -1
71
72			@BaseOptimizer.track_nth_iter
73			def iterate(self):
74			simplex_stale = all(
75			[np.array_equal(self.simplex_pos[0], array) for array in self.simplex_pos]
76			)
77
78			if simplex_stale:
79			idx_sorted = sort_list_idx(self.scores_valid)
80			self.simplex_pos = [self.positions_valid[idx] for idx in idx_sorted]
81			self.simplex_scores = [self.scores_valid[idx] for idx in idx_sorted]
82
83			self.simplex_step = 1
84
85			if self.simplex_step == 1:
86			idx_sorted = sort_list_idx(self.simplex_scores)
87			self.simplex_pos = [self.simplex_pos[idx] for idx in idx_sorted]
88			self.simplex_scores = [self.simplex_scores[idx] for idx in idx_sorted]
89
90			self.center_array = centeroid(self.simplex_pos[:-1])
91
92			r_pos = self.center_array + self.alpha * (
93			self.center_array - self.simplex_pos[-1]
94			)
95			self.r_pos = self.conv2pos(r_pos)
96			return self.r_pos
97
98			elif self.simplex_step == 2:
99			e_pos = self.center_array + self.gamma * (
100			self.center_array - self.simplex_pos[-1]
101			)
102			self.e_pos = self.conv2pos(e_pos)
103			self.simplex_step = 1
104
105			return self.e_pos
106
107			elif self.simplex_step == 3:
108			# iter Contraction
109			c_pos = self.h_pos + self.beta * (self.center_array - self.h_pos)
110			c_pos = self.conv2pos(c_pos)
111
112			return c_pos
113
114			elif self.simplex_step == 4:
115			# iter Shrink
116			pos = self.simplex_pos[self.compress_idx]
117			pos = pos + self.sigma * (self.simplex_pos[0] - pos)
118
119			return self.conv2pos(pos)
120
121			def evaluate(self, score_new):
122			self.score_new = score_new
123
124			if self.simplex_step != 0:
125			self.prev_pos = self.positions_valid[-1]
126
127			if self.simplex_step == 1:
128			# self.r_pos = self.prev_pos
129			self.r_score = score_new
130
131			if self.r_score > self.simplex_scores[0]:
132			self.simplex_step = 2
133
134			elif self.r_score > self.simplex_scores[-2]:
135			# if r is better than x N-1
136			self.simplex_pos[-1] = self.r_pos
137			self.simplex_scores[-1] = self.r_score
138			self.simplex_step = 1
139
140			if self.simplex_scores[-1] > self.r_score:
141			self.h_pos = self.simplex_pos[-1]
142			self.h_score = self.simplex_scores[-1]
143			else:
144			self.h_pos = self.r_pos
145			self.h_score = self.r_score
146
147			self.simplex_step = 3
148
149			elif self.simplex_step == 2:
150			self.e_score = score_new
151
152			if self.e_score > self.r_score:
153			self.simplex_scores[-1] = self.e_pos
154			elif self.r_score > self.e_score:
155			self.simplex_scores[-1] = self.r_pos
156			else:
157			self.simplex_scores[-1] = random.choice([self.e_pos, self.r_pos])[0]
158
159			elif self.simplex_step == 3:
160			# eval Contraction
161			self.c_pos = self.prev_pos
162			self.c_score = score_new
163
164			if self.c_score > self.simplex_scores[-1]:
165			self.simplex_scores[-1] = self.c_score
166			self.simplex_pos[-1] = self.c_pos
167
168			self.simplex_step = 1
169
170			else:
171			# start Shrink
172			self.simplex_step = 4
173			self.compress_idx = 0
174
175			elif self.simplex_step == 4:
176			# eval Shrink
177			self.simplex_scores[self.compress_idx] = score_new
178			self.simplex_pos[self.compress_idx] = self.prev_pos
179
180			self.compress_idx += 1
181
182			if self.compress_idx == self.n_simp_positions:
183			self.simplex_step = 1
184

SimonBlanke / Gradient-Free-Optimizers

Push — master ( 626f23...be3b1e )

DownhillSimplexOptimizer.iterate() B

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like