def run_cross_validation(settings, targets, classifiers, pipelines):
print 'Cross-validation task'
print 'Targets', ', '.join(targets)
print 'Pipelines:\n ', '\n '.join([p.get_name() for p in pipelines])
print 'Classifiers', ', '.join([c[1] for c in classifiers])
run_prepare_data_for_cross_validation(settings, targets, pipelines)
# run on pool first, then show results after
pool = Pool(settings.N_jobs)
for i, pipeline in enumerate(pipelines):
for j, (classifier, classifier_name) in enumerate(classifiers):
for k, target in enumerate(targets):
progress_str = 'P=%d/%d C=%d/%d T=%d/%d' % (i+1, len(pipelines), j+1, len(classifiers), k+1, len(targets))
cross_validation_score(settings, target, pipeline, classifier, classifier_name,
strategy=cross_validation_strategy, pool=pool, progress_str=progress_str, return_data=False, quiet=True)
pool.close()
pool.join()
summaries = []
best = {}
for p_num, pipeline in enumerate(pipelines):
for c_num, (classifier, classifier_name) in enumerate(classifiers):
mean_scores = []
median_scores = []
datas = []
for target in targets:
print 'Running %s pipeline %s classifier %s' % (target, pipeline.get_name(), classifier_name)
data = cross_validation_score(settings, target, pipeline, classifier, classifier_name,
strategy=cross_validation_strategy, quiet=True)
datas.append(data)
if data.mean_score != data.median_score:
print '%.3f (mean)' % data.mean_score, data.mean_scores
print '%.3f (median)' % data.median_score, data.median_scores
else:
print '%.3f' % data.mean_score
mean_scores.append(data.mean_score)
median_scores.append(data.median_score)
best_score = best.get(target, [0, None, None, None])[0]
cur_score = max(data.mean_score, data.median_score)
if cur_score > best_score:
best[target] = [cur_score, pipeline, classifier, classifier_name]
name = 'p=%d c=%d %s mean %s' % (p_num, c_num, classifier_name, pipeline.get_name())
summary = get_score_summary(name, mean_scores)
summaries.append((summary, np.mean(mean_scores)))
print summary
name = 'p=%d c=%d %s median %s' % (p_num, c_num, classifier_name, pipeline.get_name())
summary = get_score_summary(name, median_scores)
summaries.append((summary, np.mean(median_scores)))
print summary
print_results(summaries)
print '\nbest'
for target in targets:
pipeline = best[target][1]
classifier_name = best[target][3]
print target, best[target][0], classifier_name, pipeline.get_names()
def run_cross_validation(settings, targets, pipelines, mask_range, split_ratios, classifiers):
pool = Pool(settings.N_jobs)
for i, pipeline in enumerate(pipelines):
for j, (classifier, classifier_name) in enumerate(classifiers):
for k, target in enumerate(targets):
pool.apply_async(cross_validation_score, [settings, target, pipeline, classifier, classifier_name], {'quiet': True})
for split_num, split_ratio in enumerate(split_ratios):
masks = generate_feature_masks(settings, target, pipeline, np.max(mask_range), split_ratio, random_state=0, quiet=True)
for mask_num, mask in enumerate(masks):
progress_str = 'P=%d/%d C=%d/%d T=%d/%d S=%d/%d M=%d/%d' % (i+1, len(pipelines), j+1, len(classifiers), k+1, len(targets), split_num+1, len(split_ratios), mask_num+1, len(masks))
cross_validation_score(settings, target, pipeline, classifier, classifier_name, feature_mask=mask, quiet=True, return_data=False, pool=pool, progress_str=progress_str)
pool.close()
pool.join()
print 'Finished cross validation mp'
summaries = []
for p_num, pipeline in enumerate(pipelines):
for classifier, classifier_name in classifiers:
scores_full = []
scores_masked = [[[] for y in mask_range] for x in split_ratios]
for i, target in enumerate(targets):
run_prepare_data_for_cross_validation(settings, [target], [pipeline], quiet=True)
data = cross_validation_score(settings, target, pipeline, classifier, classifier_name, pool=None, quiet=True)
scores_full.append(data.mean_score)
for split_index, split_ratio in enumerate(split_ratios):
masks = generate_feature_masks(settings, target, pipeline, np.max(mask_range), split_ratio, random_state=0, quiet=True)
for mask_index, num_masks in enumerate(mask_range):
predictions = []
y_cvs = None
for mask in masks[0:num_masks]:
data = cross_validation_score(settings, target, pipeline, classifier, classifier_name, feature_mask=mask, pool=None, quiet=True)
predictions.append(data.mean_predictions)
if y_cvs is None:
y_cvs = data.y_cvs
else:
for y_cv_1, y_cv_2 in zip(y_cvs, data.y_cvs):
assert np.alltrue(y_cv_1 == y_cv_2)
predictions = np.mean(predictions, axis=0)
scores = [roc_auc_score(y_cv, p) for p, y_cv in zip(predictions, y_cvs)]
score = np.mean(scores)
scores_masked[split_index][mask_index].append(score)
summary = get_score_summary('%s p=%d full' % (classifier_name, p_num), scores_full, np.mean(scores_full), targets)
summaries.append((summary, np.mean(scores_full)))
for split_index, split_ratio in enumerate(split_ratios):
for mask_index, num_masks in enumerate(mask_range):
scores = scores_masked[split_index][mask_index]
summary = get_score_summary('%s p=%d split_ratio=%s masks=%d' % (classifier_name, p_num, split_ratio, num_masks), scores, np.mean(scores), targets)
summaries.append((summary, np.mean(scores)))
print summary
print_results(summaries)
def run_cross_validation(settings, targets, pipelines, mask_range, split_ratios, classifiers):
pool = Pool(settings.N_jobs)
for i, pipeline in enumerate(pipelines):
for j, (classifier, classifier_name) in enumerate(classifiers):
for k, target in enumerate(targets):
pool.apply_async(cross_validation_score, [settings, target, pipeline, classifier, classifier_name], {'quiet': True})
for split_num, split_ratio in enumerate(split_ratios):
masks = generate_feature_masks(settings, target, pipeline, np.max(mask_range), split_ratio, random_state=0, quiet=True)
for mask_num, mask in enumerate(masks):
progress_str = 'P=%d/%d C=%d/%d T=%d/%d S=%d/%d M=%d/%d' % (i+1, len(pipelines), j+1, len(classifiers), k+1, len(targets), split_num+1, len(split_ratios), mask_num+1, len(masks))
cross_validation_score(settings, target, pipeline, classifier, classifier_name, feature_mask=mask, quiet=True, return_data=False, pool=pool, progress_str=progress_str)
pool.close()
pool.join()
print('Finished cross validation mp')
summaries = []
for p_num, pipeline in enumerate(pipelines):
for classifier, classifier_name in classifiers:
scores_full = []
scores_masked = [[[] for y in mask_range] for x in split_ratios]
for i, target in enumerate(targets):
run_prepare_data_for_cross_validation(settings, [target], [pipeline], quiet=True)
data = cross_validation_score(settings, target, pipeline, classifier, classifier_name, pool=None, quiet=True)
scores_full.append(data.mean_score)
for split_index, split_ratio in enumerate(split_ratios):
masks = generate_feature_masks(settings, target, pipeline, np.max(mask_range), split_ratio, random_state=0, quiet=True)
for mask_index, num_masks in enumerate(mask_range):
predictions = []
y_cvs = None
for mask in masks[0:num_masks]:
data = cross_validation_score(settings, target, pipeline, classifier, classifier_name, feature_mask=mask, pool=None, quiet=True)
predictions.append(data.mean_predictions)
if y_cvs is None:
y_cvs = data.y_cvs
else:
for y_cv_1, y_cv_2 in zip(y_cvs, data.y_cvs):
assert np.alltrue(y_cv_1 == y_cv_2)
predictions = np.mean(predictions, axis=0)
scores = [roc_auc_score(y_cv, p) for p, y_cv in zip(predictions, y_cvs)]
score = np.mean(scores)
scores_masked[split_index][mask_index].append(score)
summary = get_score_summary('%s p=%d full' % (classifier_name, p_num), scores_full)
summaries.append((summary, np.mean(scores_full)))
for split_index, split_ratio in enumerate(split_ratios):
for mask_index, num_masks in enumerate(mask_range):
scores = scores_masked[split_index][mask_index]
summary = get_score_summary('%s p=%d split_ratio=%s masks=%d' % (classifier_name, p_num, split_ratio, num_masks), scores)
summaries.append((summary, np.mean(scores)))
print(summary)
print_results(summaries)
def main():
settings = load_settings()
targets = [
'Dog_1', 'Dog_2', 'Dog_3', 'Dog_4', 'Dog_5', 'Patient_1', 'Patient_2'
]
# The genetic algorithm will be run individually on each pipeline group
pipeline_groups = [
([
Pipeline(InputSource(), Preprocess(), Windower(75), PFD()),
], 0.55),
([
Pipeline(InputSource(), Preprocess(), Windower(75), Hurst()),
], 0.55),
([
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy(
[0.25, 1, 1.75, 2.5, 3.25, 4, 5, 8.5, 12, 15.5, 19.5,
24])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([0.25, 2, 3.5, 6, 15, 24])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([0.25, 2, 3.5, 6, 15])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([0.25, 2, 3.5])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([6, 15, 24])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([2, 3.5, 6])),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
PIBSpectralEntropy([3.5, 6, 15])),
Pipeline(InputSource(), Preprocess(), Windower(75), HFD(2)),
], 0.55),
]
make_submission = len(sys.argv) >= 2 and sys.argv[1] == 'submission'
run_ga = not make_submission
# This classifier is used in the genetic algorithm
ga_classifier, ga_classifier_name = make_svm(gamma=0.0079, C=2.7)
if run_ga:
quiet = False
summaries = []
for ngen in [10]:
for pipelines, ratio in pipeline_groups:
out = []
for target in targets:
print 'Running target', target
run_prepare_data_for_cross_validation(settings, [target],
pipelines,
quiet=True)
pipeline = FeatureConcatPipeline(*pipelines)
score, best_N = process_target(settings,
target,
pipeline,
ga_classifier,
ga_classifier_name,
ratio=ratio,
ngen=ngen,
quiet=quiet)
print target, score, [
np.sum(mask) for mask in best_N[0:10]
]
out.append((target, score, pipeline, best_N))
scores = np.array([score for _, score, _, _ in out])
summary = get_score_summary(
'%s ngen=%d' % (ga_classifier_name, ngen), scores,
np.mean(scores), targets)
summaries.append((summary, np.mean(scores)))
print summary
print_results(summaries)
if make_submission:
random_pipelines = [
Pipeline(InputSource(), Preprocess(), Windower(75),
Correlation('none')),
Pipeline(InputSource(), Preprocess(), Windower(75),
FreqCorrelation(1, None, 'none')),
Pipeline(
InputSource(Preprocess(), Windower(75), FFT(), Magnitude()),
FreqBinning(winning_bins, 'mean'), Log10(), FlattenChannels()),
]
# These classifiers are used to make the final predictions
final_classifiers = [
# make_svm(gamma=0.0079, C=2.7),
make_svm(gamma=0.0068, C=2.0),
# make_svm(gamma=0.003, C=150.0),
# make_lr(C=0.04),
# make_simple_lr(),
]
targets_and_pipelines = get_submission_targets_and_masks(
settings, targets, ga_classifier, ga_classifier_name,
pipeline_groups, random_pipelines)
for classifier, classifier_name in final_classifiers:
run_make_submission(settings, targets_and_pipelines, classifier,
classifier_name)
def main():
settings = load_settings()
targets = [
'Dog_1',
'Dog_2',
'Dog_3',
'Dog_4',
'Dog_5',
'Patient_1',
'Patient_2'
]
# The genetic algorithm will be run individually on each pipeline group
pipeline_groups = [
([
Pipeline(InputSource(), Preprocess(), Windower(75), PFD()),
], 0.55),
([
Pipeline(InputSource(), Preprocess(), Windower(75), Hurst()),
], 0.55),
([
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([0.25, 1, 1.75, 2.5, 3.25, 4, 5, 8.5, 12, 15.5, 19.5, 24])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([0.25, 2, 3.5, 6, 15, 24])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([0.25, 2, 3.5, 6, 15])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([0.25, 2, 3.5])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([6, 15, 24])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([2, 3.5, 6])),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), PIBSpectralEntropy([3.5, 6, 15])),
Pipeline(InputSource(), Preprocess(), Windower(75), HFD(2)),
], 0.55),
]
make_submission = len(sys.argv) >= 2 and sys.argv[1] == 'submission'
run_ga = not make_submission
# This classifier is used in the genetic algorithm
ga_classifier, ga_classifier_name = make_svm(gamma=0.0079, C=2.7)
if run_ga:
quiet = False
summaries = []
for ngen in [10]:
for pipelines, ratio in pipeline_groups:
out = []
for target in targets:
print 'Running target', target
run_prepare_data_for_cross_validation(settings, [target], pipelines, quiet=True)
pipeline = FeatureConcatPipeline(*pipelines)
score, best_N = process_target(settings, target, pipeline, ga_classifier, ga_classifier_name, ratio=ratio, ngen=ngen, quiet=quiet)
print target, score, [np.sum(mask) for mask in best_N[0:10]]
out.append((target, score, pipeline, best_N))
scores = np.array([score for _, score, _, _ in out])
summary = get_score_summary('%s ngen=%d' % (ga_classifier_name, ngen), scores)
summaries.append((summary, np.mean(scores)))
print summary
print_results(summaries)
if make_submission:
random_pipelines = [
Pipeline(InputSource(), Preprocess(), Windower(75), Correlation('none')),
Pipeline(InputSource(), Preprocess(), Windower(75), FreqCorrelation(1, None, 'none')),
Pipeline(InputSource(Preprocess(), Windower(75), FFT(), Magnitude()), FreqBinning(winning_bins, 'mean'), Log10(), FlattenChannels()),
]
# These classifiers are used to make the final predictions
final_classifiers = [
# make_svm(gamma=0.0079, C=2.7),
make_svm(gamma=0.0068, C=2.0),
# make_svm(gamma=0.003, C=150.0),
# make_lr(C=0.04),
# make_simple_lr(),
]
targets_and_pipelines = get_submission_targets_and_masks(settings, targets, ga_classifier, ga_classifier_name, pipeline_groups, random_pipelines)
for classifier, classifier_name in final_classifiers:
run_make_submission(settings, targets_and_pipelines, classifier, classifier_name)
def run_cross_validation(settings, targets, classifiers, pipelines):
print 'Cross-validation task'
print 'Targets', ', '.join(targets)
print 'Pipelines:\n ', '\n '.join([p.get_name() for p in pipelines])
print 'Classifiers', ', '.join([c[1] for c in classifiers])
run_prepare_data_for_cross_validation(settings, targets, pipelines)
# run on pool first, then show results after
pool = Pool(settings.N_jobs)
for i, pipeline in enumerate(pipelines):
for j, (classifier, classifier_name) in enumerate(classifiers):
for k, target in enumerate(targets):
progress_str = 'P=%d/%d C=%d/%d T=%d/%d' % (
i + 1, len(pipelines), j + 1, len(classifiers), k + 1,
len(targets))
cross_validation_score(settings,
target,
pipeline,
classifier,
classifier_name,
strategy=cross_validation_strategy,
pool=pool,
progress_str=progress_str,
return_data=False,
quiet=True)
pool.close()
pool.join()
summaries = []
best = {}
for p_num, pipeline in enumerate(pipelines):
for c_num, (classifier, classifier_name) in enumerate(classifiers):
mean_scores = []
median_scores = []
datas = []
for target in targets:
print 'Running %s pipeline %s classifier %s' % (
target, pipeline.get_name(), classifier_name)
data = cross_validation_score(
settings,
target,
pipeline,
classifier,
classifier_name,
strategy=cross_validation_strategy,
quiet=True)
datas.append(data)
if data.mean_score != data.median_score:
print '%.3f (mean)' % data.mean_score, data.mean_scores
print '%.3f (median)' % data.median_score, data.median_scores
else:
print '%.3f' % data.mean_score
mean_scores.append(data.mean_score)
median_scores.append(data.median_score)
best_score = best.get(target, [0, None, None, None])[0]
cur_score = max(data.mean_score, data.median_score)
if cur_score > best_score:
best[target] = [
cur_score, pipeline, classifier, classifier_name
]
name = 'p=%d c=%d %s mean %s' % (p_num, c_num, classifier_name,
pipeline.get_name())
summary = get_score_summary(name, mean_scores)
summaries.append((summary, np.mean(mean_scores)))
print summary
name = 'p=%d c=%d %s median %s' % (p_num, c_num, classifier_name,
pipeline.get_name())
summary = get_score_summary(name, median_scores)
summaries.append((summary, np.mean(median_scores)))
print summary
print_results(summaries)
print '\nbest'
for target in targets:
pipeline = best[target][1]
classifier_name = best[target][3]
print target, best[target][0], classifier_name, pipeline.get_names()