def run_big():
training_data = generate_training_data("./data/GUM_tsv/training/*.tsv")
num_features = training_data[0][0][0][0][1].shape[0]
random1 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
random2 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
model = Model(training_data)
model.train()
testing_data = generate_test_data("./data/GUM_tsv/testing/*.tsv")
evaluator = Evaluator()
evaluator.print_translation = False
for data in testing_data:
evaluator.add_case(TestCase(*data))
evaluator.evaluate(model)
def run_small():
bob_is_bob = generate_test_data("./data/tests/bob_is_bob.tsv")[0]
bob_is_erik = generate_test_data("./data/tests/bob_is_erik.tsv")[0]
bob_is_not_erik = generate_test_data("./data/tests/bob_is_not_erik.tsv")[0]
example = generate_test_data("./data/example/*.tsv")[0]
num_features = bob_is_bob[0][0][0][1].shape[0]
random1 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
random2 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
bob_is_bob_training = generate_training_data(
"./data/tests/bob_is_bob.tsv")[0]
bob_is_erik_training = generate_training_data(
"./data/tests/bob_is_erik.tsv")[0]
bob_is_not_erik_training = generate_training_data(
"./data/tests/bob_is_not_erik.tsv")[0]
example_training = generate_training_data("./data/example/*.tsv")[0]
training_data = [
bob_is_bob_training, bob_is_erik_training, bob_is_not_erik_training,
example_training
]
model = Model(training_data)
model.train()
model_dt = ModelDT(training_data)
model_dt.train()
evaluator = Evaluator()
evaluator.add_case(TestCase(*bob_is_bob))
evaluator.add_case(TestCase(*bob_is_erik))
evaluator.add_case(TestCase(*bob_is_not_erik))
(x1, y1) = random1
# Pass an empty set and a name manually, as the random data has no name or mapping.
evaluator.add_case(TestCase(x1, y1, {}, "random1"))
(x2, y2) = random2
evaluator.add_case(TestCase(x2, y2, {}, "random2"))
evaluator.add_case(TestCase(*example))
print("Normal model")
evaluator.evaluate(model)
print("Decision tree model")
evaluator.evaluate(model_dt)
from evaluator import Evaluator, TestCase
num_features = 100
training1 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
training2 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
random1 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
random2 = random_data(node_count=10,
cluster_count=2,
features=num_features,
max_edges=30)
training_data = [training1, training2]
model = Model(training_data)
model.train()
evaluator = Evaluator()
evaluator.add_case(TestCase("Training 1", *training1))
evaluator.add_case(TestCase("Training 2", *training2))
evaluator.add_case(TestCase("Random 1", *random1))
evaluator.add_case(TestCase("Random 2", *random2))
evaluator.evaluate(model)
