def test_data_normalization(self):
self.t = thresher.Thresher(labels=(0, 1))
actual_classes = [0, 0, 1, 1]
compute_result = self.t.optimize_threshold(self.scores, actual_classes)
print(f'[ThresherVerySmallTest] Result found: {compute_result}')
self.assertTrue(
0.3 <= compute_result < 0.4,
msg="Checking proper result for the ThresherVerySmallTest")
def test_data_case_parallel(self):
self.t = thresher.Thresher(algorithm_params={'n_jobs': 3})
actual_classes = [-1, -1, -1, -1, -1, -1, -1, 1, 1]
compute_result = self.t.optimize_threshold(self.scores, actual_classes)
print(f'[ThresherVerySmallTest] Result found: {compute_result}')
self.assertTrue(
0.3 <= compute_result < 0.4,
msg="Checking proper result for the ThresherVerySmallTest")
def setUp(self):
# Preparing data for unit test
self.t = thresher.Thresher(verbose=False, progress_bar=False)
self.alt_t = thresher.Thresher(algorithm='linear',
verbose=False,
progress_bar=False)
self.alt_t2 = thresher.Thresher(algorithm='sim',
verbose=False,
progress_bar=False)
self.alt_t3 = thresher.Thresher(algorithm='grid')
self.alt_t4 = thresher.Thresher(algorithm='sgrid',
algorithm_params={
'no_of_decimal_places': 2,
'stoch_ratio': 0.10
})
self.alt_t5 = thresher.Thresher(algorithm='sgrid',
algorithm_params={
'no_of_decimal_places': 3,
'stoch_ratio': 0.06,
'reshuffle': True
})
print('Preparing data for ThresherMediumTest...')
medium_data = get_sample_data(path='./')
self.scores = list(medium_data['pred'].values)
self.actual_classes = list(medium_data['actual'].values)
self.left_allowed, self.right_allowed = 0.40, 0.65
def setUp(self):
# Preparing data for unit test
self.t = thresher.Thresher(progress_bar=False)
self.scores = [0.1, 0.15, 0.2, 0.22, 0.27, 0.29, 0.3, 0.4, 0.7]
import thresher
from thresher.tests.sample_data import get_sample_data
t = thresher.Thresher(progress_bar=True)
print('Currently supported algorithms:')
print(t.get_supported_algorithms())
case_small_scores = [0.1, 0.3, 0.4, 0.7]
case_small_labels = [-1, -1, 1, 1]
print(
f'Optimization result: {t.optimize_threshold(case_small_scores, case_small_labels)}'
)
t = thresher.Thresher(progress_bar=True, verbose=True)
medium_data = get_sample_data()
case_medium_scores = list(medium_data['pred'].values)
case_medium_labels = list(medium_data['actual'].values)
print(
f'Optimization result: {t.optimize_threshold(case_medium_scores, case_medium_labels)}'
)
t = thresher.Thresher(algorithm='gen', progress_bar=True, verbose=True)
medium_data = get_sample_data()
case_medium_scores = list(medium_data['pred'].values)
case_medium_labels = list(medium_data['actual'].values)
import thresher
from thresher.tests.sample_data import get_sample_data
t = thresher.Thresher(verbose=True, algorithm_params={'n_jobs': 3})
print('Currently supported algorithms:')
print(t.get_supported_algorithms())
case_small_scores = [0.1, 0.15, 0.2, 0.22, 0.27, 0.29, 0.3, 0.4, 0.7]
case_small_labels = [-1, -1, -1, -1, -1, -1, -1, 1, 1]
print(
f'Optimization result: {t.optimize_threshold(case_small_scores, case_small_labels)}'
)
print('Done')