def test_digits_sqrt_sieve_batch():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3], random_state=0)
model.partial_fit(X_digits)
assert_array_equal(model.ranking, digits_sieve_ranking)
assert_array_almost_equal(model.gains, digits_sieve_gains, 4)
assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_modular_object():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer=ModularGreedy(random_state=0))
model.fit(X_digits_cupy)
assert_array_equal(model.ranking, digits_modular_ranking)
assert_array_almost_equal(model.gains, digits_modular_gains, 4)
def test_digits_two_stage_sparse():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='two-stage')
model.fit(X_digits_sparse)
assert_array_equal(model.ranking, digits_ranking)
assert_array_almost_equal(model.gains, digits_gains, 4)
def test_digits_lazy():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='lazy')
model.fit(X_digits_cupy)
assert_array_equal(model.ranking, digits_ranking)
assert_array_almost_equal(model.gains, digits_gains, 4)
def test_digits_approximate_object():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer=ApproximateLazyGreedy())
model.fit(X_digits_cupy)
assert_array_equal(model.ranking, digits_approx_ranking)
assert_array_almost_equal(model.gains, digits_approx_gains, 4)
def test_digits_precomputed_two_stage(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='precomputed', optimizer='two-stage') model.fit(X_digits_cosine_cupy) assert_array_equal(model.ranking, digits_cosine_ranking) assert_array_almost_equal(model.gains, digits_cosine_gains, 4)
def test_digits_cosine_modular_sparse(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='precomputed', optimizer='modular', random_state=0) model.fit(X_digits_cosine_sparse) assert_array_equal(model.ranking, digits_cosine_modular_ranking) assert_array_almost_equal(model.gains, digits_cosine_modular_gains, 4)
def test_digits_stochastic():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='stochastic',
random_state=0)
model.fit(X_digits_cupy)
assert_array_equal(model.ranking, digits_stochastic_ranking)
assert_array_almost_equal(model.gains, digits_stochastic_gains, 4)
def test_digits_two_stage_init():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='two-stage',
initial_subset=digits_ranking[:5])
model.fit(X_digits_cupy)
assert_array_equal(model.ranking[:-5], digits_ranking[5:])
assert_array_almost_equal(model.gains[:-5], digits_gains[5:], 4)
def test_digits_modular_sparse():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='modular',
random_state=0)
model.fit(X_digits_sparse)
assert_array_equal(model.ranking, digits_modular_ranking)
assert_array_almost_equal(model.gains, digits_modular_gains, 4)
def test_digits_sample_object():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer=SampleGreedy(random_state=0))
model.fit(X_digits)
assert_array_equal(model.ranking, digits_sample_ranking)
assert_array_almost_equal(model.gains, digits_sample_gains, 4)
assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_sqrt_modular_object(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='cosine', optimizer=ModularGreedy(random_state=0)) model.fit(X_digits) assert_array_equal(model.ranking, digits_cosine_modular_ranking) assert_array_almost_equal(model.gains, digits_cosine_modular_gains, 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_precomputed_two_stage_init(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='precomputed', optimizer='two-stage', initial_subset=digits_cosine_ranking[:5]) model.fit(X_digits_cosine_cupy) assert_array_equal(model.ranking[:-5], digits_cosine_ranking[5:]) assert_array_almost_equal(model.gains[:-5], digits_cosine_gains[5:], 4)
def test_digits_euclidean_lazy(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='euclidean', optimizer='lazy') model.fit(X_digits) assert_array_equal(model.ranking, digits_euclidean_ranking) assert_array_almost_equal(model.gains, digits_euclidean_gains, 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_corr_two_stage(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='corr', optimizer='two-stage') model.fit(X_digits) assert_array_equal(model.ranking, digits_corr_ranking) assert_array_almost_equal(model.gains, digits_corr_gains, 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_cosine_greedi_ll_sparse():
model1 = FacilityLocationSelection(100)
model2 = GraphCutSelection(100)
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
metric='precomputed', optimizer='greedi', optimizer_kwds={
'optimizer1': 'lazy', 'optimizer2': 'lazy'}, random_state=0)
model.fit(X_digits_cosine_sparse)
assert_array_equal(model.ranking[:30], digits_cosine_greedi_ranking[:30])
assert_array_almost_equal(model.gains[:30], digits_cosine_greedi_gains[:30], 4)
def test_digits_cosine_sample(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='cosine', optimizer='sample', random_state=0) model.fit(X_digits) assert_array_equal(model.ranking, digits_cosine_sample_ranking) assert_array_almost_equal(model.gains, digits_cosine_sample_gains, 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_naive():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='naive')
model.fit(X_digits_cupy)
assert_array_equal(model.ranking[:30], digits_ranking[:30])
assert_array_equal(model.ranking[-30:], digits_ranking[-30:])
assert_array_almost_equal(model.gains, digits_gains, 4)
def test_digits_cosine_lazy_init(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='cosine', optimizer='lazy', initial_subset=digits_cosine_ranking[:5]) model.fit(X_digits) assert_array_equal(model.ranking[:-5], digits_cosine_ranking[5:]) assert_array_almost_equal(model.gains[:-5], digits_cosine_gains[5:], 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_approximate():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='approximate-lazy',
random_state=0)
model.fit(X_digits)
assert_array_equal(model.ranking, digits_approx_ranking)
assert_array_almost_equal(model.gains, digits_approx_gains, 4)
assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_naive_sparse():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='naive')
model.fit(X_digits_sparse)
assert_array_equal(model.ranking, digits_ranking)
assert_array_almost_equal(model.gains, digits_gains, 4)
assert_array_almost_equal(model.subset,
X_digits_sparse[model.ranking].toarray())
def test_digits_greedi_nl_object():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer=GreeDi(optimizer1='naive',
optimizer2='lazy',
random_state=0))
model.fit(X_digits_cupy)
assert_array_equal(model.ranking[:85], digits_greedi_ranking[:85])
assert_array_almost_equal(model.gains[:85], digits_greedi_gains[:85], 4)
def test_digits_euclidean_two_stage_init(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='euclidean', optimizer='two-stage', initial_subset=digits_euclidean_ranking[:5]) model.fit(X_digits) assert_array_equal(model.ranking[:-5], digits_euclidean_ranking[5:]) assert_array_almost_equal(model.gains[:-5], digits_euclidean_gains[5:], 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_naive_init():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='naive',
initial_subset=digits_ranking[:5])
model.fit(X_digits)
assert_array_equal(model.ranking[:20], digits_ranking[5:25])
assert_array_almost_equal(model.gains[:20], digits_gains[5:25], 4)
assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_cosine_greedi_nl_object(): model1 = FacilityLocationSelection(100) model2 = GraphCutSelection(100) model = MixtureSelection(100, [model1, model2], [1.0, 0.3], metric='cosine', optimizer=GreeDi(optimizer1='naive', optimizer2='lazy', random_state=0)) model.fit(X_digits) assert_array_equal(model.ranking[:30], digits_cosine_greedi_ranking[:30]) assert_array_almost_equal(model.gains[:30], digits_cosine_greedi_gains[:30], 4) assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_cosine_greedi_ln():
model1 = FacilityLocationSelection(100)
model2 = GraphCutSelection(100)
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
metric='cosine', optimizer='greedi',
optimizer_kwds={'optimizer1': 'lazy', 'optimizer2': 'naive'},
random_state=0)
model.fit(X_digits)
assert_array_equal(model.ranking, digits_cosine_greedi_ranking)
assert_array_almost_equal(model.gains, digits_cosine_greedi_gains, 4)
assert_array_almost_equal(model.subset, X_digits[model.ranking])
def test_digits_greedi_nl_sparse():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='greedi',
optimizer_kwds={
'optimizer1': 'naive',
'optimizer2': 'lazy'
},
random_state=0)
model.fit(X_digits_sparse)
assert_array_equal(model.ranking[:85], digits_greedi_ranking[:85])
assert_array_almost_equal(model.gains[:85], digits_greedi_gains[:85], 4)
def test_digits_greedi_ln():
model1 = FeatureBasedSelection(100, 'sqrt')
model2 = FeatureBasedSelection(100, 'log')
model = MixtureSelection(100, [model1, model2], [1.0, 0.3],
optimizer='two-stage',
optimizer_kwds={
'optimizer1': 'lazy',
'optimizer2': 'naive'
},
random_state=0)
model.fit(X_digits_cupy)
assert_array_equal(model.ranking, digits_greedi_ranking)
assert_array_almost_equal(model.gains, digits_greedi_gains, 4)