class SimpleModelTest(opus_unittest.OpusTestCase):
def setUp(self):
self.data = {
'id':
arange(10) + 1,
'attribute':
array([3000, 2800, 1000, 550, 600, 1000, 2000, 500, 100, 1000])
}
storage = StorageFactory().get_storage('dict_storage')
storage.write_table(table_name='dataset', table_data=self.data)
self.dataset = Dataset(in_storage=storage,
in_table_name='dataset',
id_name=['id'])
def test_simple_model(self):
m = SimpleModel()
m.run(self.dataset,
'sqrt(dataset.attribute)',
outcome_attribute='sqrtattr')
self.assertEqual(
ma.allclose(self.dataset.get_attribute('sqrtattr'),
sqrt(self.data['attribute'])), True)
self.assertEqual(
'sqrtattr' in self.dataset.get_primary_attribute_names(), True)
def test_simple_model_without_outcome_attribute(self):
m = SimpleModel()
m.run(self.dataset, 'lattr = ln(dataset.attribute)')
self.assertEqual(
ma.allclose(self.dataset.get_attribute('lattr'),
log(self.data['attribute'])), True)
self.assertEqual('lattr' in self.dataset.get_primary_attribute_names(),
True)
class SimpleModelTest(opus_unittest.OpusTestCase):
def setUp(self):
self.data = {
'id': arange(10)+1,
'attribute': array([3000,2800,1000,550,600,1000,2000,500,100,1000]),
'sqrt_outcome': zeros(10)
}
storage = StorageFactory().get_storage('dict_storage')
storage.write_table(table_name = 'dataset', table_data = self.data)
self.dataset = Dataset(in_storage=storage, in_table_name='dataset', id_name=['id'])
def test_simple_model(self):
m = SimpleModel()
m.run(self.dataset, 'sqrt(dataset.attribute)', outcome_attribute='sqrtattr')
self.assertEqual(ma.allclose(self.dataset.get_attribute('sqrtattr'), sqrt(self.data['attribute'])), True)
self.assertEqual('sqrtattr' in self.dataset.get_primary_attribute_names(), True)
def test_simple_model_with_filter(self):
m = SimpleModel()
m.run(self.dataset, 'sqrt(dataset.attribute)', outcome_attribute='sqrt_outcome', dataset_filter='dataset.attribute>1000')
expected = array([1, 1, 0, 0, 0, 0, 1, 0, 0, 0]) * sqrt(self.data['attribute'])
self.assertEqual(ma.allclose(self.dataset.get_attribute('sqrt_outcome'),
expected), True)
self.assertEqual('sqrt_outcome' in self.dataset.get_primary_attribute_names(), True)
def MASKEDtest_simple_model_with_random_filter(self):
m = SimpleModel()
m.run(self.dataset, 'sqrt(dataset.attribute)',
outcome_attribute='sqrt_outcome',
dataset_filter='(dataset.attribute>=1000) & (random_like(dataset.attribute)<=0.5)',
)
con_filter = self.dataset['attribute']>=1000
results = self.dataset['sqrt_outcome'][con_filter]
expected = sqrt(self.data['attribute'])[con_filter]
#test half of the elements passing filter are being sqrt
self.assertEqual((results==expected).sum(), expected.size/2)
self.assertEqual((results!=expected).sum(), expected.size/2)
def test_simple_model_without_outcome_attribute(self):
m = SimpleModel()
m.run(self.dataset, 'lattr = ln(dataset.attribute)')
self.assertEqual(ma.allclose(self.dataset.get_attribute('lattr'), log(self.data['attribute'])), True)
self.assertEqual('lattr' in self.dataset.get_primary_attribute_names(), True)
def test_simple_model_with_outcome_values(self):
m = SimpleModel()
m.run(self.dataset, outcome_attribute='iniattr', outcome_values=zeros(10)-1)
self.assertEqual(ma.allclose(self.dataset.get_attribute('iniattr'), array(10*[-1])), True)
self.assertEqual('iniattr' in self.dataset.get_primary_attribute_names(), True)
# run with filter
m.run(self.dataset, outcome_attribute='iniattr', outcome_values=arange(10)+1, dataset_filter='dataset.attribute>1000')
expected = array([1, 2, -1, -1, -1, -1, 7, -1, -1, -1])
self.assertEqual(ma.allclose(self.dataset.get_attribute('iniattr'), expected), True)
class DeleteAttributeModelTest(opus_unittest.OpusTestCase):
def setUp(self):
self.data = {
'id': arange(10)+1,
'attribute': array([3000,2800,1000,550,600,1000,2000,500,100,1000]),
}
storage = StorageFactory().get_storage('dict_storage')
storage.write_table(table_name = 'dataset', table_data = self.data)
self.dataset = Dataset(in_storage=storage, in_table_name='dataset', id_name=['id'])
def test_delete_attribute_model(self):
m = DeleteAttributeModel()
m.run(self.dataset, 'attr')
self.assertEqual('attribute' in self.dataset.get_primary_attribute_names(), True)
m.run(self.dataset, 'attribute')
self.assertEqual('attribute' in self.dataset.get_primary_attribute_names(), False)
class SimpleModelTest(opus_unittest.OpusTestCase):
def setUp(self):
self.data = {
'id': arange(10)+1,
'attribute': array([3000,2800,1000,550,600,1000,2000,500,100,1000])
}
storage = StorageFactory().get_storage('dict_storage')
storage.write_table(table_name = 'dataset', table_data = self.data)
self.dataset = Dataset(in_storage=storage, in_table_name='dataset', id_name=['id'])
def test_simple_model(self):
m = SimpleModel()
m.run(self.dataset, 'sqrt(dataset.attribute)', outcome_attribute='sqrtattr')
self.assertEqual(ma.allclose(self.dataset.get_attribute('sqrtattr'), sqrt(self.data['attribute'])), True)
self.assertEqual('sqrtattr' in self.dataset.get_primary_attribute_names(), True)
def test_simple_model_without_outcome_attribute(self):
m = SimpleModel()
m.run(self.dataset, 'lattr = ln(dataset.attribute)')
self.assertEqual(ma.allclose(self.dataset.get_attribute('lattr'), log(self.data['attribute'])), True)
self.assertEqual('lattr' in self.dataset.get_primary_attribute_names(), True)
class SimpleModelTest(opus_unittest.OpusTestCase):
def setUp(self):
self.data = {
'id':
arange(10) + 1,
'attribute':
array([3000, 2800, 1000, 550, 600, 1000, 2000, 500, 100, 1000]),
'sqrt_outcome':
zeros(10)
}
storage = StorageFactory().get_storage('dict_storage')
storage.write_table(table_name='dataset', table_data=self.data)
self.dataset = Dataset(in_storage=storage,
in_table_name='dataset',
id_name=['id'])
def test_simple_model(self):
m = SimpleModel()
m.run(self.dataset,
'sqrt(dataset.attribute)',
outcome_attribute='sqrtattr')
self.assertEqual(
ma.allclose(self.dataset.get_attribute('sqrtattr'),
sqrt(self.data['attribute'])), True)
self.assertEqual(
'sqrtattr' in self.dataset.get_primary_attribute_names(), True)
def test_simple_model_with_filter(self):
m = SimpleModel()
m.run(self.dataset,
'sqrt(dataset.attribute)',
outcome_attribute='sqrt_outcome',
dataset_filter='dataset.attribute>1000')
expected = array([1, 1, 0, 0, 0, 0, 1, 0, 0, 0]) * sqrt(
self.data['attribute'])
self.assertEqual(
ma.allclose(self.dataset.get_attribute('sqrt_outcome'), expected),
True)
self.assertEqual(
'sqrt_outcome' in self.dataset.get_primary_attribute_names(), True)
def MASKEDtest_simple_model_with_random_filter(self):
m = SimpleModel()
m.run(
self.dataset,
'sqrt(dataset.attribute)',
outcome_attribute='sqrt_outcome',
dataset_filter=
'(dataset.attribute>=1000) & (random_like(dataset.attribute)<=0.5)',
)
con_filter = self.dataset['attribute'] >= 1000
results = self.dataset['sqrt_outcome'][con_filter]
expected = sqrt(self.data['attribute'])[con_filter]
#test half of the elements passing filter are being sqrt
self.assertEqual((results == expected).sum(), expected.size / 2)
self.assertEqual((results != expected).sum(), expected.size / 2)
def test_simple_model_without_outcome_attribute(self):
m = SimpleModel()
m.run(self.dataset, 'lattr = ln(dataset.attribute)')
self.assertEqual(
ma.allclose(self.dataset.get_attribute('lattr'),
log(self.data['attribute'])), True)
self.assertEqual('lattr' in self.dataset.get_primary_attribute_names(),
True)
def test_simple_model_with_outcome_values(self):
m = SimpleModel()
m.run(self.dataset,
outcome_attribute='iniattr',
outcome_values=zeros(10) - 1)
self.assertEqual(
ma.allclose(self.dataset.get_attribute('iniattr'),
array(10 * [-1])), True)
self.assertEqual(
'iniattr' in self.dataset.get_primary_attribute_names(), True)
# run with filter
m.run(self.dataset,
outcome_attribute='iniattr',
outcome_values=arange(10) + 1,
dataset_filter='dataset.attribute>1000')
expected = array([1, 2, -1, -1, -1, -1, 7, -1, -1, -1])
self.assertEqual(
ma.allclose(self.dataset.get_attribute('iniattr'), expected), True)
