def test_compute_models_parallel_sklearn():
passed_params = {'n_components', 'learning_method', 'evaluate_every', 'max_iter', 'n_jobs'}
varying_params = [dict(n_components=k) for k in range(2, 5)]
const_params = dict(learning_method='batch', evaluate_every=1, max_iter=3, n_jobs=1)
models = tm_sklearn.compute_models_parallel(EVALUATION_TEST_DTM, varying_params, const_params)
assert len(models) == len(varying_params)
for param_set, model in models:
assert set(param_set.keys()) == passed_params
assert isinstance(model, LatentDirichletAllocation)
assert isinstance(model.components_, np.ndarray)
def test_compute_models_parallel_sklearn_multiple_docs():
# 1 doc, no varying params
const_params = dict(n_components=3, learning_method='batch', evaluate_every=1, max_iter=3, n_jobs=1)
models = tm_sklearn.compute_models_parallel(EVALUATION_TEST_DTM, constant_parameters=const_params)
assert len(models) == 1
assert type(models) is list
assert len(models[0]) == 2
param1, model1 = models[0]
assert param1 == const_params
assert isinstance(model1, LatentDirichletAllocation)
assert isinstance(model1.components_, np.ndarray)
# 1 *named* doc, some varying params
passed_params = {'n_components', 'learning_method', 'evaluate_every', 'max_iter', 'n_jobs'}
const_params = dict(learning_method='batch', evaluate_every=1, max_iter=3, n_jobs=1)
varying_params = [dict(n_components=k) for k in range(2, 5)]
docs = {'test1': EVALUATION_TEST_DTM}
models = tm_sklearn.compute_models_parallel(docs, varying_params,
constant_parameters=const_params)
assert len(models) == len(docs)
assert isinstance(models, dict)
assert set(models.keys()) == {'test1'}
for d, m in models.items():
assert d == 'test1'
assert len(m) == len(varying_params)
for param_set, model in m:
assert set(param_set.keys()) == passed_params
assert isinstance(model, LatentDirichletAllocation)
assert isinstance(model.components_, np.ndarray)
# n docs, no varying params
const_params = dict(n_components=3, learning_method='batch', evaluate_every=1, max_iter=3, n_jobs=1)
models = tm_sklearn.compute_models_parallel(EVALUATION_TEST_DTM_MULTI, constant_parameters=const_params)
assert len(models) == len(EVALUATION_TEST_DTM_MULTI)
assert isinstance(models, dict)
assert set(models.keys()) == set(EVALUATION_TEST_DTM_MULTI.keys())
for d, m in models.items():
assert len(m) == 1
for param_set, model in m:
assert set(param_set.keys()) == set(const_params.keys())
assert isinstance(model, LatentDirichletAllocation)
assert isinstance(model.components_, np.ndarray)
# n docs, some varying params
passed_params = {'n_components', 'learning_method', 'evaluate_every', 'max_iter', 'n_jobs'}
const_params = dict(learning_method='batch', evaluate_every=1, max_iter=3, n_jobs=1)
varying_params = [dict(n_components=k) for k in range(2, 5)]
models = tm_sklearn.compute_models_parallel(EVALUATION_TEST_DTM_MULTI, varying_params,
constant_parameters=const_params)
assert len(models) == len(EVALUATION_TEST_DTM_MULTI)
assert isinstance(models, dict)
assert set(models.keys()) == set(EVALUATION_TEST_DTM_MULTI.keys())
for d, m in models.items():
assert len(m) == len(varying_params)
for param_set, model in m:
assert set(param_set.keys()) == passed_params
assert isinstance(model, LatentDirichletAllocation)
assert isinstance(model.components_, np.ndarray)