def test_get_model_parameter(self):
self.assertTrue(expr=len(ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=list(CLUSTER_ALGORITHMS.keys())
).get_model_parameter().keys()
) > 0
)
def test_train_non_negative_matrix_factorization(self):
_model_generator: ClusteringGenerator = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['nmf'],
sep=',',
train_data_path=DATA_FILE_PATH,
)
_model = _model_generator.generate_model()
_model.train()
self.assertTrue(expr=len(_model.cluster_label) > 0 and _model.nmi > 0.0)
def test_train_latent_semantic_indexing(self):
_model_generator: ClusteringGenerator = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['lsi'],
sep=',',
train_data_path=DATA_FILE_PATH,
)
_model = _model_generator.generate_model()
_model.train()
self.assertTrue(expr=len(_model.cluster_label) > 0 and _model.nmi > 0.0)
def test_train_gibbs_sampling_dirichlet_multinomial_model(self):
_model_generator: ClusteringGenerator = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['gsdmm'],
sep=',',
train_data_path=DATA_FILE_PATH,
)
_model = _model_generator.generate_model()
_model.train()
self.assertTrue(expr=len(_model.cluster_label) > 0 and _model.nmi > 0.0)
def test_generate_latent_semantic_indexing_model(self):
_net_gen: object = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['lsi'],
tokenize=False,
random=True,
sep=',',
cloud=None,
train_data_path=DATA_FILE_PATH
).generate_model()
self.assertTrue(expr=isinstance(_net_gen.model, LatentSemanticIndexing))
def test_generate_latent_dirichlet_allocation_model(self):
_net_gen: object = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['lda'],
tokenize=False,
random=True,
sep=',',
cloud=None,
train_data_path=DATA_FILE_PATH
).generate_model()
self.assertTrue(expr=isinstance(_net_gen.model, LatentDirichletAllocation))
def test_generate_gibbs_sampling_dirichlet_multinomial_model(self):
_net_gen: object = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['gsdmm'],
tokenize=True,
random=True,
sep=',',
cloud=None,
train_data_path=DATA_FILE_PATH
).generate_model()
self.assertTrue(expr=isinstance(_net_gen.model, GibbsSamplingDirichletMultinomialModeling))
def test_generate_params(self):
_model_generator: ClusteringGenerator = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=list(CLUSTER_ALGORITHMS.keys()),
sep=',',
train_data_path=DATA_FILE_PATH,
)
_model = _model_generator.generate_model()
_mutated_param: dict = copy.deepcopy(_model.model_param_mutated)
_model_generator.generate_params(param_rate=0.1, force_param=None)
self.assertTrue(expr=len(_mutated_param.keys()) < len(_model_generator.model_param_mutated.keys()))
def test_generate_non_negative_matrix_factorization_model(self):
_net_gen: object = ClusteringGenerator(predictor='tweet',
model_name=None,
cluster_params=None,
models=['nmf'],
tokenize=False,
random=True,
sep=',',
cloud=None,
train_data_path=DATA_FILE_PATH
).generate_model()
self.assertTrue(expr=isinstance(_net_gen.model, NonNegativeMatrixFactorization))