forked from dragonfly/dragonfly
-
Notifications
You must be signed in to change notification settings - Fork 0
/
dragonfly-script.py
159 lines (142 loc) · 6.99 KB
/
dragonfly-script.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
#!/usr/bin/env python
"""
Command line tool for Dragonfly.
-- kandasamy@cs.cmu.edu
-- kvysyara@andrew.cmu.edu
Usage:
python dragonfly-script.py --config <config file in .json or .pb format> --options <options file>
See README in main repository for examples.
"""
# pylint: disable=relative-import
# pylint: disable=invalid-name
# pylint: disable=import-error
from __future__ import print_function
import os
# import imp
from importlib import import_module
import sys
# Local
from dragonfly import maximise_function, minimise_function, \
maximise_multifidelity_function, \
minimise_multifidelity_function, \
multiobjective_maximise_functions, \
multiobjective_minimise_functions
from dragonfly.exd.cp_domain_utils import load_config_file
from dragonfly.exd.exd_utils import get_unique_list_of_option_args
from dragonfly.utils.option_handler import get_option_specs, load_options
# Get options
from dragonfly.opt.ga_optimiser import ga_opt_args
from dragonfly.opt.gp_bandit import get_all_euc_gp_bandit_args, \
get_all_cp_gp_bandit_args, get_all_mf_euc_gp_bandit_args, \
get_all_mf_cp_gp_bandit_args
from dragonfly.opt.random_optimiser import euclidean_random_optimiser_args, \
mf_euclidean_random_optimiser_args, \
cp_random_optimiser_args, mf_cp_random_optimiser_args
from dragonfly.opt.multiobjective_gp_bandit import get_all_euc_moo_gp_bandit_args, \
get_all_cp_moo_gp_bandit_args
from dragonfly.opt.random_multiobjective_optimiser import \
euclidean_random_multiobjective_optimiser_args, \
cp_random_multiobjective_optimiser_args
dragonfly_args = [ \
get_option_specs('config', False, None, 'Path to the json or pb config file. '),
get_option_specs('options', False, None, 'Path to the options file. '),
get_option_specs('max_or_min', False, 'max', 'Whether to maximise or minimise. '),
get_option_specs('max_capital', False, -1.0,
'Maximum capital (available budget) to be used in the experiment. '),
get_option_specs('capital_type', False, 'return_value',
'Maximum capital (available budget) to be used in the experiment. '),
get_option_specs('is_multi_objective', False, 0,
'If True, will treat it as a multiobjective optimisation problem. '),
get_option_specs('opt_method', False, 'bo',
('Optimisation method. Default is bo. This should be one of bo, ga, ea, direct, ' +
' pdoo, or rand, but not all methods apply to all problems.')),
get_option_specs('report_progress', False, 'default',
('How to report progress. Should be one of "default" (prints to stdout), ' +
'"silent" (no reporting), or a filename (writes to file).')),
]
def get_command_line_args():
""" Returns all arguments for the command line. """
ret = dragonfly_args + \
ga_opt_args + \
euclidean_random_optimiser_args + cp_random_optimiser_args + \
mf_euclidean_random_optimiser_args + mf_cp_random_optimiser_args + \
get_all_euc_gp_bandit_args() + get_all_cp_gp_bandit_args() + \
get_all_mf_euc_gp_bandit_args() + get_all_mf_cp_gp_bandit_args() + \
euclidean_random_multiobjective_optimiser_args + \
cp_random_multiobjective_optimiser_args + \
get_all_euc_moo_gp_bandit_args() + get_all_cp_moo_gp_bandit_args()
return get_unique_list_of_option_args(ret)
def main():
""" Main function. """
options = load_options(get_command_line_args(), cmd_line=True)
# Load domain and objective
config = load_config_file(options.config)
if hasattr(config, 'fidel_space'):
is_mf = True
else:
is_mf = False
# Load module
expt_dir = os.path.dirname(os.path.abspath(os.path.realpath(options.config)))
if not os.path.exists(expt_dir):
raise ValueError("Experiment directory does not exist.")
sys.path.append(expt_dir)
obj_module = import_module(config.name, expt_dir)
sys.path.remove(expt_dir)
# Set capital
if options.max_capital < 0:
raise ValueError('max_capital (time or number of evaluations) must be positive.')
# Call optimiser
_print_prefix = 'Maximising' if options.max_or_min == 'max' else 'Minimising'
call_to_optimise = {
'single': {'max': maximise_function, 'min': minimise_function},
'single_mf': {'max': maximise_multifidelity_function,
'min': minimise_multifidelity_function},
'multi': {'max': multiobjective_maximise_functions,
'min': multiobjective_minimise_functions},
}
if not options.is_multi_objective:
if is_mf:
print('%s multi-fidelity function on\n Fidelity-Space: %s.\n Domain: %s.\n'%(
_print_prefix, config.fidel_space, config.domain))
opt_val, opt_pt, history = call_to_optimise['single_mf'][options.max_or_min](
obj_module.objective, fidel_space=None, domain=None,
fidel_to_opt=config.fidel_to_opt, fidel_cost_func=obj_module.cost,
max_capital=options.max_capital, capital_type=options.capital_type,
opt_method=options.opt_method, config=config, options=options,
reporter=options.report_progress)
else:
print('%s function on Domain: %s.\n'%(_print_prefix, config.domain))
opt_val, opt_pt, history = call_to_optimise['single'][options.max_or_min](
obj_module.objective, domain=None, max_capital=options.max_capital,
capital_type=options.capital_type, opt_method=options.opt_method,
config=config, options=options, reporter=options.report_progress)
# curr_opt_vals, curr_true_opt_vals, query_true_vals, query_vals
print(history.query_true_vals)
print('Optimum Value in %d evals: %0.4f'%(len(history.curr_opt_points), opt_val))
print('Optimum Point: %s.'%(opt_pt))
else:
if is_mf:
raise ValueError('Multi-objective multi-fidelity optimisation has not been ' +
'implemented yet.')
else:
# Check format of function caller
if hasattr(obj_module, 'objectives'):
objectives_to_pass = obj_module.objectives
num_objectives = len(objectives_to_pass)
else:
num_objectives = obj_module.num_objectives
objectives_to_pass = (obj_module.compute_objectives, obj_module.num_objectives)
print('%s %d multiobjective functions on Domain: %s.\n'%(_print_prefix,
num_objectives, config.domain))
print(objectives_to_pass)
pareto_values, pareto_points, history = \
call_to_optimise['multi'][options.max_or_min](objectives_to_pass,
domain=None, max_capital=options.max_capital, capital_type=options.capital_type,
opt_method=options.opt_method, config=config, options=options,
reporter=options.report_progress)
num_pareto_points = len(pareto_points)
print('Found %d Pareto Points: %s.'%(num_pareto_points, pareto_points))
print('Corresponding Pareto Values: %s.'%(pareto_values))
if __name__ == '__main__':
sys.path.insert(0, os.getcwd())
main()