def compare_ground_truth():
data = read_data_sc2('../data/sequences-TZ-45.txt')[:5000]
data = reduce_k_length(10, data)
target = '1'
enable_i = True
# if we want to average
nb_launched = 5
pool = Pool(processes=3, maxtasksperchild=1)
iterations_limit = 50
iteration_step = 1000
data_final = {'WRAcc': [], 'iterations': [], 'Algorithm': []}
# found with exaustive search
ground_truth = 0.008893952000000009
for i in range(10):
print('Iteration: {}'.format(i))
for i in range(nb_launched):
result_ucb_opti = pool.apply_async(
seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'iterations_limit': iterations_limit
})
data_add_generic(
data_final,
WRAcc=max(0, average_results(result_ucb_opti.get())) /
ground_truth,
iterations=iterations_limit,
Algorithm='SeqScout')
iterations_limit += iteration_step
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
ax = sns.lineplot(data=df, x='iterations', y='WRAcc', hue='Algorithm')
ax.set(xlabel='iterations', ylabel='WRAcc')
plt.savefig('./ground_truth/gt.png')
df.to_pickle('./ground_truth/result')
if SHOW:
plt.show()
def quality_over_size():
number_dataset = 6
data_origin, target, enable_i = datasets[number_dataset]
pool = Pool(processes=3, maxtasksperchild=1)
# if we want to average
nb_launched = 5
size = 15
size_step = 4
data_final = {'WRAcc': [], 'size': [], 'Algorithm': []}
for i in range(10):
print('Iteration: {}'.format(i))
data = reduce_k_length(size, data_origin)
for i in range(nb_launched):
results_misere = pool.apply_async(misere, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_beam = pool.apply_async(beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
result_ucb_opti = pool.apply_async(seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
results_mcts = pool.apply_async(launch_mcts, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_misere = results_misere.get()
results_beam = results_beam.get()
result_ucb_opti = result_ucb_opti.get()
results_mcts = results_mcts.get()
if len(results_beam) < TOP_K:
print("Too few beam: {}".format(len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few SeqScout: {}".format(len(result_ucb_opti)))
if len(results_misere) < TOP_K:
print("Too few misere: {}".format(len(results_misere)))
if len(results_mcts) < TOP_K:
print("Too few MCTSExtent: {}".format(len(results_mcts)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
size=size,
Algorithm='misere')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_beam)),
size=size,
Algorithm='beam')
data_add_generic(data_final,
WRAcc=max(0, average_results(result_ucb_opti)),
size=size,
Algorithm='SeqScout')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_mcts)),
size=size,
Algorithm='MCTSExtent')
size += size_step
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
ax = sns.lineplot(data=df, x='size', y='WRAcc', hue='Algorithm')
ax.set(xlabel='Length max', ylabel='WRAcc')
# ax.set(xlabel='Time(s)', ylabel='Average WRAcc top-10 patterns')
plt.savefig('./space_size/over_size.png')
df.to_pickle('./space_size/result')
if SHOW:
plt.show()
def quality_over_theta():
number_dataset = 1
data, target, enable_i = datasets[number_dataset]
pool = Pool(processes=5, maxtasksperchild=1)
# if we want to average
nb_launched = 5
theta = 0.1
data_final = {'WRAcc': [], 'theta': [], 'Algorithm': []}
for i in range(10):
print('Iteration: {}'.format(i))
for i in range(nb_launched):
results_misere = pool.apply_async(misere, (data, target), {
'time_budget': TIME_BUDGET_XP,
'theta': theta
})
results_beam = pool.apply_async(
beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'theta': theta
})
result_ucb_opti = pool.apply_async(
seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'theta': theta
})
results_mcts = pool.apply_async(launch_mcts, (data, target), {
'time_budget': TIME_BUDGET_XP,
'theta': theta
})
results_misere = results_misere.get()
results_beam = results_beam.get()
result_ucb_opti = result_ucb_opti.get()
results_mcts = results_mcts.get()
if len(results_beam) < TOP_K:
print("Too few beam: {}".format(len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few ucb: {}".format(len(result_ucb_opti)))
if len(results_misere) < TOP_K:
print("Too few misere: {}".format(len(results_misere)))
if len(results_mcts) < TOP_K:
print("Too few mctsextent : {}".format(len(results_mcts)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
theta=theta,
Algorithm='misere')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_beam)),
theta=theta,
Algorithm='beam')
data_add_generic(data_final,
WRAcc=max(0, average_results(result_ucb_opti)),
theta=theta,
Algorithm='SeqScout')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_mcts)),
theta=theta,
Algorithm='MCTSExtent')
theta += 0.1
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
ax = sns.lineplot(data=df, x='theta', y='WRAcc', hue='Algorithm')
plt.savefig('./theta/over_theta.png')
df.to_pickle('./theta/result')
if SHOW:
plt.show()
def barplot_dataset_iterations():
pool = Pool(processes=5, maxtasksperchild=1)
xp_repeat = 5
data_final = {'WRAcc': [], 'dataset': [], 'Algorithm': []}
for i, (data, target, enable_i) in enumerate(datasets):
print("Dataset {}".format(datasets_names[i]))
for j in range(xp_repeat):
results_misere = pool.apply_async(misere, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_beam = pool.apply_async(beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
result_ucb_opti = pool.apply_async(seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
results_mcts = pool.apply_async(launch_mcts, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_misere = results_misere.get()
results_beam = results_beam.get()
result_ucb_opti = result_ucb_opti.get()
results_mcts = results_mcts.get()
if len(results_misere) < TOP_K:
print("Too few example on misere on dataset {}: {} results".
format(datasets_names[i], len(results_misere)))
if len(results_beam) < TOP_K:
print(
"Too few example on beam_search on dataset {}: {} results".
format(datasets_names[i], len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few example on seqscout on dataset {}: {} results".
format(datasets_names[i], len(result_ucb_opti)))
if len(results_mcts) < TOP_K:
print(
"Too few example on mctsextend on dataset {}: {} results".
format(datasets_names[i], len(results_mcts)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
dataset=datasets_names[i],
Algorithm='misere')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_beam)),
dataset=datasets_names[i],
Algorithm='beam')
data_add_generic(data_final,
WRAcc=max(0, average_results(result_ucb_opti)),
dataset=datasets_names[i],
Algorithm='SeqScout')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_mcts)),
dataset=datasets_names[i],
Algorithm='MCTSExtent')
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
ax = sns.barplot(x='dataset', y='WRAcc', hue='Algorithm', data=df)
plt.savefig('./wracc_datasets/iterations_boxplot.png')
df.to_pickle('./wracc_datasets/result')
if SHOW:
plt.show()
def other_measures():
pool = Pool(processes=5, maxtasksperchild=1)
xp_repeat = 5
nb_iterations = 10000
for i, (data, target, enable_i) in enumerate(datasets):
print("Dataset {}".format(datasets_names[i]))
for measure in ['Informedness', 'F1']:
mean_misere = 0
mean_beam = 0
mean_seqscout = 0
mean_mcts = 0
for j in range(xp_repeat):
results_misere = pool.apply_async(
misere, (data, target), {
'time_budget': TIME_BUDGET_XP,
'quality_measure': measure,
'iterations_limit': nb_iterations
})
results_beam = pool.apply_async(
beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'quality_measure': measure,
'iterations_limit': nb_iterations
})
result_ucb_opti = pool.apply_async(
seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'quality_measure': measure,
'iterations_limit': nb_iterations
})
results_mcts = pool.apply_async(
launch_mcts, (data, target), {
'time_budget': TIME_BUDGET_XP,
'iterations_limit': nb_iterations,
'quality_measure': measure
})
results_misere = results_misere.get()
results_beam = results_beam.get()
result_ucb_opti = result_ucb_opti.get()
results_mcts = results_mcts.get()
if len(results_misere) < TOP_K:
print(
"Too few example on misere on dataset {}: {} results".
format(datasets_names[i], len(results_misere)))
if len(results_beam) < TOP_K:
print(
"Too few example on beam_search on dataset {}: {} results"
.format(datasets_names[i], len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few example on ucb on dataset {}: {} results".
format(datasets_names[i], len(result_ucb_opti)))
if len(results_mcts) < TOP_K:
print(
"Too few example on mctsextend on dataset {}: {} results"
.format(datasets_names[i], len(results_mcts)))
mean_misere += average_results(results_misere)
mean_beam += average_results(results_beam)
mean_seqscout += average_results(result_ucb_opti)
mean_mcts += average_results(results_mcts)
mean_misere = mean_misere / xp_repeat
mean_beam = mean_beam / xp_repeat
mean_seqscout = mean_seqscout / xp_repeat
mean_mcts = mean_mcts / xp_repeat
print(
'For datasets {}, measure {}, algorithm misere the means score is: {}'
.format(datasets_names[i], measure, mean_misere))
print(
'For datasets {}, measure {}, algorithm beam_search the means score is: {}'
.format(datasets_names[i], measure, mean_beam))
print(
'For datasets {}, measure {}, algorithm ucb the means score is: {}'
.format(datasets_names[i], measure, mean_seqscout))
print(
'For datasets {}, measure {}, algorithm mctsextend the means score is: {}'
.format(datasets_names[i], measure, mean_mcts))
def show_quality_over_iterations_ucb(number_dataset):
data, target, enable_i = datasets[number_dataset]
# if we want to average
nb_launched = 5
pool = Pool(processes=3, maxtasksperchild=1)
iterations_limit = 50
iterations_step = 1000
data_final = {'WRAcc': [], 'iterations': [], 'Algorithm': []}
for i in range(12):
print('Iteration: {}'.format(i))
for i in range(nb_launched):
results_misere = pool.apply_async(
misere, (data, target), {
'time_budget': TIME_BUDGET_XP,
'iterations_limit': iterations_limit
})
results_beam = pool.apply_async(
beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'iterations_limit': iterations_limit
})
result_ucb_opti = pool.apply_async(
seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP,
'iterations_limit': iterations_limit
})
results_mcts = pool.apply_async(
launch_mcts, (data, target), {
'time_budget': TIME_BUDGET_XP,
'iterations_limit': iterations_limit
})
data_add_generic(data_final,
WRAcc=max(0,
average_results(results_misere.get())),
iterations=iterations_limit,
Algorithm='misere')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_beam.get())),
iterations=iterations_limit,
Algorithm='beam')
data_add_generic(data_final,
WRAcc=max(0,
average_results(result_ucb_opti.get())),
iterations=iterations_limit,
Algorithm='SeqScout')
data_add_generic(data_final,
WRAcc=max(0, average_results(results_mcts.get())),
iterations=iterations_limit,
Algorithm='MCTSExtent')
iterations_limit += iterations_step
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
ax = sns.lineplot(data=df, x='iterations', y='WRAcc', hue='Algorithm')
plt.savefig('./iterations_ucb/over_iterations{}.png'.format(
datasets_names[number_dataset]))
df.to_pickle('./iterations_ucb/result_{}'.format(
datasets_names[number_dataset]))
if SHOW:
plt.show()
def barplot_dataset_improvement_iterations():
pool = Pool(processes=5, maxtasksperchild=1)
xp_repeat = 5
seqscout_vs_beam = {'Improvement': [], 'dataset': []}
seqscout_vs_misere = {'Improvement': [], 'dataset': []}
mcts_vs_misere = {'Improvement': [], 'dataset': []}
mcts_vs_beam = {'Improvement': [], 'dataset': []}
mcts_vs_seqscout = {'Improvement': [], 'dataset': []}
for i, (data, target, enable_i) in enumerate(datasets):
print("Dataset {}".format(datasets_names[i]))
for j in range(xp_repeat):
results_misere = pool.apply_async(misere, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_beam = pool.apply_async(beam_search, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
result_ucb_opti = pool.apply_async(seq_scout, (data, target), {
'enable_i': enable_i,
'time_budget': TIME_BUDGET_XP
})
results_mcts = pool.apply_async(launch_mcts, (data, target),
{'time_budget': TIME_BUDGET_XP})
results_misere = results_misere.get()
results_beam = results_beam.get()
result_ucb_opti = result_ucb_opti.get()
results_mcts = results_mcts.get()
if len(results_misere) < TOP_K:
print("Too few example on misere on dataset {}: {} results".
format(datasets_names[i], len(results_misere)))
if len(results_beam) < TOP_K:
print(
"Too few example on beam_search on dataset {}: {} results".
format(datasets_names[i], len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few example on seqscout on dataset {}: {} results".
format(datasets_names[i], len(result_ucb_opti)))
if len(results_mcts) < TOP_K:
print(
"Too few example on mctsextend on dataset {}: {} results".
format(datasets_names[i], len(results_mcts)))
data_add_generic(seqscout_vs_misere,
Improvement=compute_improvement(
average_results(result_ucb_opti),
average_results(results_misere)),
dataset=datasets_names[i])
data_add_generic(seqscout_vs_beam,
Improvement=compute_improvement(
average_results(result_ucb_opti),
average_results(results_beam)),
dataset=datasets_names[i])
data_add_generic(mcts_vs_misere,
Improvement=compute_improvement(
average_results(results_mcts),
average_results(results_misere)),
dataset=datasets_names[i])
data_add_generic(mcts_vs_beam,
Improvement=compute_improvement(
average_results(results_mcts),
average_results(results_beam)),
dataset=datasets_names[i])
data_add_generic(mcts_vs_seqscout,
Improvement=compute_improvement(
average_results(results_mcts),
average_results(result_ucb_opti)),
dataset=datasets_names[i])
sns.set(rc={'figure.figsize': (8, 6.5)})
plt.clf()
df = pd.DataFrame(data=seqscout_vs_misere)
ax = sns.barplot(x='dataset', y='Improvement', data=df)
ax.set(yscale='symlog')
annotate_bars(ax)
plt.savefig('./improvement/seqscout_vs_misere.png')
df.to_pickle('./improvement/seqscout_vs_misere')
plt.clf()
df = pd.DataFrame(data=seqscout_vs_beam)
bx = sns.barplot(x='dataset', y='Improvement', data=df)
bx.set(yscale='symlog')
annotate_bars(bx)
plt.savefig('./improvement/seqscout_vs_beam.png')
df.to_pickle('./improvement/seqscout_vs_beam')
plt.clf()
df = pd.DataFrame(data=mcts_vs_misere)
cx = sns.barplot(x='dataset', y='Improvement', data=df)
cx.set(yscale='symlog')
annotate_bars(cx)
plt.savefig('./improvement/mcts_vs_misere.png')
df.to_pickle('./improvement/mcts_vs_misere')
plt.clf()
df = pd.DataFrame(data=mcts_vs_beam)
dx = sns.barplot(x='dataset', y='Improvement', data=df)
dx.set(yscale='symlog')
annotate_bars(dx)
plt.savefig('./improvement/mcts_vs_beam.png')
df.to_pickle('./improvement/mcts_vs_beam')
plt.clf()
df = pd.DataFrame(data=mcts_vs_seqscout)
ex = sns.barplot(x='dataset', y='Improvement', data=df)
ex.set(yscale='symlog')
annotate_bars(ex)
plt.savefig('./improvement/mcts_vs_seqscout.png')
df.to_pickle('./improvement/mcts_bs_seqscout')
plt.clf()
if SHOW:
plt.show()
def barplot_increase_local_optima(it_number):
pool = Pool(processes=5, maxtasksperchild=1)
xp_repeat = 5
data_final = {
'WRAcc': [],
'WRAcc_opti': [],
'dataset': [],
'Algorithm': []
}
for i, (data, target, enable_i) in enumerate(datasets):
print("Dataset {}".format(datasets_names[i]))
for j in range(xp_repeat):
results_misere = pool.apply_async(
optimizer,
(data, target, 'misere', TIME_BUDGET_XP, it_number, enable_i))
results_beam = pool.apply_async(
optimizer, (data, target, 'BeamSearch', TIME_BUDGET_XP,
it_number, enable_i))
results_ucb_opti = pool.apply_async(
optimizer, (data, target, 'SeqScout', TIME_BUDGET_XP,
it_number, enable_i))
results_mcts = pool.apply_async(
optimizer, (data, target, 'MCTSExtent', TIME_BUDGET_XP,
it_number, enable_i))
results_misere, results_misere_opti = results_misere.get()
results_beam, results_beam_opti = results_beam.get()
result_ucb_opti, results_ucb_opti = results_ucb_opti.get()
results_mcts, results_mcts_opti = results_mcts.get()
if len(results_misere) < TOP_K:
print("Too few example on misere on dataset {}: {} results".
format(datasets_names[i], len(results_misere)))
if len(results_beam) < TOP_K:
print(
"Too few example on beam_search on dataset {}: {} results".
format(datasets_names[i], len(results_beam)))
if len(result_ucb_opti) < TOP_K:
print("Too few example on seqscout on dataset {}: {} results".
format(datasets_names[i], len(result_ucb_opti)))
if len(results_mcts) < TOP_K:
print(
"Too few example on mctsextend on dataset {}: {} results".
format(datasets_names[i], len(results_mcts)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
dataset=datasets_names[i],
Algorithm='misere',
WRAcc_opti=max(
0, average_results(results_misere_opti)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
dataset=datasets_names[i],
Algorithm='beam',
WRAcc_opti=max(
0, average_results(results_beam_opti)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
dataset=datasets_names[i],
Algorithm='SeqScout',
WRAcc_opti=max(0,
average_results(results_ucb_opti)))
data_add_generic(data_final,
WRAcc=max(0, average_results(results_misere)),
dataset=datasets_names[i],
Algorithm='MCTSExtent',
WRAcc_opti=max(
0, average_results(results_mcts_opti)))
df = pd.DataFrame(data=data_final)
sns.set(rc={'figure.figsize': (8, 6.5)})
df.to_pickle('./local_opti_increase/result_{}'.format(it_number))
plt.clf()
sns.set_color_codes("pastel")
ax = sns.barplot(x='dataset', y='WRAcc_opti', hue='Algorithm', data=df)
sns.set_color_codes("muted")
ax = sns.barplot(x='dataset', y='WRAcc', hue='Algorithm', data=df)
plt.savefig('./local_opti_increase/barplot_{}.png'.format(it_number))
if SHOW:
plt.show()