def test_csv_merge(self): X_train, y_train, X_test, y_test = get_dev_split(raw) X_test_ext = get_extended_X(X_test, raw) # Check join with tube.csv. taid = 'TA-00034' df = X_test_ext[X_test['tube_assembly_id'] == taid] ref_diams = raw['tube'][raw['tube']['tube_assembly_id'] == taid] nose.tools.assert_equal(len(df['diameter'].unique()), 1) nose.tools.assert_equal(len(ref_diams), 1) nose.tools.assert_equal( df['diameter'].unique()[0], ref_diams['diameter'].values[0]) # Check join with specs.csv. taid = 'TA-00207' ref_specs = ['SP-0063', 'SP-0070', 'SP-0080'] df = X_test_ext[X_test['tube_assembly_id'] == taid] assert len(df['specs']) >= 1 for val in df['specs']: nose.tools.assert_equal(val, ref_specs) # Check join with bill_of_materials.csv. taid = 'TA-00249' ref_components = [ ('C-1536', 2.0), ('C-1642', 1.0), ('C-1649', 1.0), ] df = X_test_ext[X_test['tube_assembly_id'] == taid] assert len(df['components']) >= 1 for val in df['components']: nose.tools.assert_equal(val, ref_components)
def test_csv_merge(self): X_train, y_train, X_test, y_test = get_dev_split(raw) X_test_ext = get_extended_X(X_test, raw) # Check join with tube.csv. taid = 'TA-00034' df = X_test_ext[X_test['tube_assembly_id'] == taid] ref_diams = raw['tube'][raw['tube']['tube_assembly_id'] == taid] nose.tools.assert_equal(len(df['diameter'].unique()), 1) nose.tools.assert_equal(len(ref_diams), 1) nose.tools.assert_equal(df['diameter'].unique()[0], ref_diams['diameter'].values[0]) # Check join with specs.csv. taid = 'TA-00207' ref_specs = ['SP-0063', 'SP-0070', 'SP-0080'] df = X_test_ext[X_test['tube_assembly_id'] == taid] assert len(df['specs']) >= 1 for val in df['specs']: nose.tools.assert_equal(val, ref_specs) # Check join with bill_of_materials.csv. taid = 'TA-00249' ref_components = [ ('C-1536', 2.0), ('C-1642', 1.0), ('C-1649', 1.0), ] df = X_test_ext[X_test['tube_assembly_id'] == taid] assert len(df['components']) >= 1 for val in df['components']: nose.tools.assert_equal(val, ref_components)
def test_custom_features(self): X_train, y_train, X_test, y_test = get_dev_split(raw) X_train_ext = get_extended_X(X_train, raw) featurizer = CustomFeaturizer() featurizer.fit(X_train_ext) X_train_feats = featurizer.transform(X_train_ext) # Check adj_quantity feature. taid = 'TA-01916' df = X_train_feats[X_train['tube_assembly_id'] == taid] nose.tools.assert_equal(len(df), 2) nose.tools.assert_true(np.all(df.min_order_quantity == [1, 4])) nose.tools.assert_true(np.all(df.quantity == [1, 1])) nose.tools.assert_true(np.all(df.adj_quantity == [1, 4]))