def test_metric_func(self):
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
tree_method='hist')
data, validation_data = get_data()
def pyrmsle(y_true, y_pred):
y_pred[y_pred < -1] = -1 + 1e-6
elements = np.power(np.log1p(y_true) - np.log1p(y_pred), 2)
return float(np.sqrt(np.sum(elements) / len(y_true)))
with pytest.raises(ValueError) as exeinfo:
auto_xgb_reg.fit(data=data,
epochs=1,
validation_data=validation_data,
metric=pyrmsle,
search_space=get_xgb_search_space(),
n_sampling=4)
assert "metric_mode" in str(exeinfo)
auto_xgb_reg.fit(data=data,
epochs=1,
validation_data=validation_data,
metric=pyrmsle,
metric_mode="min",
search_space=get_xgb_search_space(),
n_sampling=4)
def test_fit(self):
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
tree_method='hist')
data, validation_data = get_data()
auto_xgb_reg.fit(data=data,
validation_data=validation_data,
search_space=create_XGB_recipe(),
n_sampling=4,
epochs=1,
metric="mse")
best_model = auto_xgb_reg.get_best_model()
assert 5 <= best_model.model.n_estimators <= 10
assert 2 <= best_model.model.max_depth <= 5
def test_fit(self):
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
tree_method='hist')
data, validation_data = get_data()
auto_xgb_reg.fit(data=data,
validation_data=validation_data,
search_space=get_xgb_search_space(),
n_sampling=4,
epochs=1,
metric="mae")
best_model = auto_xgb_reg.get_best_model()
assert 5 <= best_model.n_estimators <= 10
assert 2 <= best_model.max_depth <= 5
best_config = auto_xgb_reg.get_best_config()
assert all(k in best_config.keys()
for k in get_xgb_search_space().keys())
def test_metric(self):
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
tree_method='hist')
data, validation_data = get_data()
with pytest.raises(ValueError) as exeinfo:
auto_xgb_reg.fit(data=data,
epochs=1,
validation_data=validation_data,
metric="logloss",
search_space=get_xgb_search_space(),
n_sampling=4)
assert "metric_mode" in str(exeinfo)
auto_xgb_reg.fit(data=data,
epochs=1,
validation_data=validation_data,
metric="logloss",
metric_mode="min",
search_space=get_xgb_search_space(),
n_sampling=4)
def test_data_creator(self):
train_data_creator, val_data_creator = get_data_creators()
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
tree_method='hist')
model_search_space = get_xgb_search_space()
# todo: change to hp.choice_n
search_space = {
"features":
hp.sample_from(
lambda spec: np.random.choice(["f1", "f2", "f3"], size=2))
}
search_space.update(model_search_space)
auto_xgb_reg.fit(data=train_data_creator,
epochs=1,
validation_data=val_data_creator,
metric="logloss",
metric_mode="min",
search_space=search_space,
n_sampling=2)
best_config = auto_xgb_reg.get_best_config()
assert all(k in best_config.keys() for k in search_space.keys())
assert len(best_config["features"]) == 2
max_depth=list(max_depth_range),
lr=lr,
min_child_weight=min_child_weight)
search_alg = None
search_alg_params = None
scheduler = None
scheduler_params = None
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
**config)
auto_xgb_reg.fit(data=(X_train, y_train),
validation_data=(X_val, y_val),
metric="rmse",
n_sampling=recipe.num_samples,
search_space=recipe.search_space(),
search_alg=search_alg,
search_alg_params=None,
scheduler=scheduler,
scheduler_params=scheduler_params)
print("Training completed.")
best_model = auto_xgb_reg.get_best_model()
y_hat = best_model.predict(X_val)
rmse = best_model.evaluate(X_val, y_val, metrics=["rmse"])
print("Evaluate: the square root of mean square error is", rmse)
ray_ctx.stop()
sc.stop()
n_estimators=list(n_estimators_range),
max_depth=list(max_depth_range),
lr=lr,
min_child_weight=min_child_weight)
search_alg = None
search_alg_params = None
scheduler = None
scheduler_params = None
auto_xgb_reg = AutoXGBRegressor(cpus_per_trial=2,
name="auto_xgb_regressor",
**config)
auto_xgb_reg.fit(data=(X_train, y_train),
validation_data=(X_val, y_val),
recipe=recipe,
metric="rmse",
search_alg=search_alg,
search_alg_params=None,
scheduler=scheduler,
scheduler_params=scheduler_params)
print("Training completed.")
best_model = auto_xgb_reg.get_best_model()
y_hat = best_model.predict(X_val)
rmse = best_model.evaluate(X_val, y_val, metrics=["rmse"])
print("Evaluate: the square root of mean square error is", rmse)
ray_ctx.stop()
sc.stop()