def plot_experiments(*exp_names,
log=False,
transform_map=lambda x: x,
ylim=(0, 1),
stat_type=["objective"],
color=None,
fill=True):
plt.figure(figsize=(12, 8))
if log:
plt.yscale("log")
if ylim is not None:
plt.ylim(ylim)
for s in stat_type:
rename_dict = {}
for name in exp_names:
appendix = (".train" if s == "fitness" else ".test")
rename_dict[name] = name + appendix
utils.plot_experiments(OUT_DIR,
exp_names,
transform_map=transform_map,
stat_type=s,
rename_dict=rename_dict,
color=color,
fill=fill)
plt.show()
def plot_experiments(*exp_names, transform_map=lambda x: x - 150000, ylim=[10000, 150000]):
plt.figure(figsize=(12,8))
plt.yscale("log")
# settings specific for tsp_std.in input
if ylim is not None:
plt.ylim(ylim)
utils.plot_experiments(OUT_DIR, exp_names, transform_map=transform_map)
plt.show()
# use and edit this file to make all the plots you need - it is generally easier
# than plotting directly after the run of the algorithm
import utils
import matplotlib.pyplot as plt
plt.figure(figsize=(12, 8))
utils.plot_experiments('rules', ['default'])
plt.show()
# plt.savefig(f_name + 'fitness.png')
# plt.cla()
fit_name = 'f10'
stat_type = "fitness"
plt.yscale('log')
# plt.xscale('log')
# utils.plot_experiments(
# OUT_DIR, ['default' + '.' + fit_name, 'adapt_variance' + '.' + fit_name, 'adapt_fitness_probability.' + fit_name], stat_type=stat_type)
# plt.savefig(fit_name + stat_type + ".png")
experiments = ['diff_evolution' + '.' + name for name in fit_names]
# utils.plot_experiments(
# OUT_DIR, experiments, stat_type=stat_type)
# plt.savefig("diff_evolution_" + stat_type + ".png")
# utils.plot_experiments(
# OUT_DIR, ['diff_evolution' + '.' + "f01"], stat_type=stat_type)
# plt.savefig("diff_evolution_" + stat_type + ".png")
# for name in fit_names:
# plt.xscale('log')
# utils.plot_experiments(
# OUT_DIR, ['diff_evolution' + '.' + name], stat_type=stat_type)
# plt.savefig("diff_evolution_" + name + stat_type + ".png")
# plt.close()
for name in fit_names:
utils.plot_experiments(OUT_DIR, ['diff_evolution' + '.' + name],
stat_type=stat_type)
plt.savefig("diff_evolution_" + stat_type + ".png")
# use and edit this file to make all the plots you need - it is generally easier
# than plotting directly after the run of the algorithm
import utils
import matplotlib.pyplot as plt
plt.figure(figsize=(12,8))
utils.plot_experiments('continuous', ['default.f01', 'tuned.f01'],
rename_dict={'default.f01': 'Sphere (default)',
'tuned.f01': 'Sphere (tuned)'})
plt.show()
def plot_all_experiments():
with open(OUT_DIR + "/" + EXP_NAMES_LIST, "r") as f:
utils.plot_experiments(OUT_DIR,
list(set(f.read().strip().split("\n"))))
plt.show()
# log.write_files()
# print an overview of the best individuals from each run
for i, bi in enumerate(best_inds):
print(
f'Run {i}: difference = {fit(bi).objective}, bin weights = {bin_weights(weights, bi)}'
)
# write summary logs for the whole experiment
utils.summarize_experiment(OUT_DIR, EXP_ID)
# read the summary log and plot the experiment
evals, lower, mean, upper = utils.get_plot_data(OUT_DIR, EXP_ID)
plt.figure(figsize=(12, 8))
utils.plot_experiment(evals,
lower,
mean,
upper,
legend_name='Default settings')
plt.legend()
plt.show()
# you can also plot mutiple experiments at the same time using
# utils.plot_experiments, e.g. if you have two experiments 'default' and
# 'tuned' both in the 'partition' directory, you can call
utils.plot_experiments('partition', ['default', 'tuned'],
rename_dict={'default': 'Default setting'})
# the rename_dict can be used to make reasonable entries in the legend -
# experiments that are not in the dict use their id (in this case, the
# legend entries would be 'Default settings' and 'tuned')
import utils
utils.plot_experiments(
'.', [
'partition/k10_p100_g500_x0.8_m0.2_mf0.1_r10(default)',
'final_bad/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned)',
'final_worse/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned)',
'final_bad_but_better/k10-p1024-g20480-s10-x0.0-m0.1-mf0.002-ms0.15-dc0.0-r10(tuned)',
'final_better_9/k10-p500-g10000-s10-x0.0-m0.002-mf0.05-ms0.33-dc0.1-r10(tuned)',
'final_better_11/k10-p800-g20000-s10-x0.0-m0.001-mf0.02-ms0.15-dc0.05-r10(tuned)'
],
rename_dict={
'final_better_11/k10-p800-g20000-s10-x0.0-m0.001-mf0.02-ms0.15-dc0.05-r10(tuned)':
'final_better_11/k10-p800-g20000-s10-x0.0-m0.001-mf0.02-ms0.15-dc0.05-r10(tuned - best found)',
'final_better_9/k10-p500-g10000-s10-x0.0-m0.002-mf0.05-ms0.33-dc0.1-r10(tuned)':
'final_better_9/k10-p500-g10000-s10-x0.0-m0.002-mf0.05-ms0.33-dc0.1-r10(tuned - enhanced fit)',
'final_bad/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned)':
'final_bad/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned - no special mutations)',
'final_worse/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned)':
'final_worse/k10-p800-g10000-s10-x0.0-m0.00125-mf0.01-ms0.15-dc0.0-r10(tuned - no special mut & init)',
'final_bad_but_better/k10-p1024-g20480-s10-x0.0-m0.1-mf0.002-ms0.15-dc0.0-r10(tuned)':
'final_bad_but_better/k10-p1024-g20480-s10-x0.0-m0.1-mf0.002-ms0.15-dc0.0-r10(tuned - init + low params - no special mut)',
})
import utils
for graph in ['f01', 'f02', 'f06', 'f08', 'f10']:
utils.plot_experiments('continuous', [
'.'.join(['default', graph]),
'.'.join(['diff-threesome', graph]),
'.'.join(['diff-threesome-small-F', graph]),
'.'.join(['diff-threesome-small-CF', graph]),
'.'.join(['diff-basic-adaptive', graph]),
'.'.join(['diff-threesome-adaptive', graph]),
'.'.join(['diff-g******g-adaptive', graph]),
])