def execute():
#################### Environment ####################
env = Environment(
train_dataset=get_boston_data(),
results_path="HyperparameterHunterAssets",
holdout_dataset=get_holdout_data,
target_column="DIS",
metrics=["r2_score", "median_absolute_error"],
cv_type="KFold",
cv_params=dict(n_splits=10, random_state=1),
)
#################### CVExperiment ####################
exp_0 = CVExperiment(
model_initializer=Ridge,
model_init_params=dict(),
feature_engineer=FeatureEngineer([quantile_transform]),
)
#################### Optimization ####################
# `opt_0` recognizes `exp_0`'s `feature_engineer` and its results as valid learning material
# This is because `opt_0` marks the engineer step functions omitted by `exp_0` as `optional=True`
opt_0 = DummyOptPro(iterations=10)
opt_0.forge_experiment(
model_initializer=Ridge,
model_init_params=dict(),
feature_engineer=FeatureEngineer([
Categorical([quantile_transform, log_transform], optional=True),
Categorical([standard_scale, standard_scale_BAD], optional=True),
Categorical([square_sum_feature], optional=True),
]),
)
opt_0.go()
def env_boston():
return Environment(
train_dataset=get_boston_data(),
results_path=assets_dir,
target_column="DIS",
metrics=["r2_score"],
cv_type="KFold",
cv_params=dict(n_splits=3, random_state=1),
)
def env_boston_regression():
env = Environment(
train_dataset=get_boston_data(),
results_path=assets_dir,
target_column="DIS",
metrics=["median_absolute_error"],
cv_type="KFold",
cv_params=dict(n_splits=3, random_state=1),
)
return env
def env_boston():
return Environment(
train_dataset=get_boston_data(),
results_path=assets_dir,
holdout_dataset=get_holdout_data,
target_column="DIS",
metrics=["r2_score", "median_absolute_error"],
cv_type="KFold",
cv_params=dict(n_splits=3, random_state=1),
runs=1,
verbose=1,
)
def env_boston(request):
"""Environment fixture using the Boston regression dataset. Parametrizes `holdout_dataset`, so
all tests using this fixture will be run twice: once with no `holdout_dataset`, and once with a
`holdout_dataset` constructed using SKLearn's `train_test_split`"""
return Environment(
train_dataset=get_boston_data(),
results_path=assets_dir,
target_column="DIS",
metrics=["r2_score"],
holdout_dataset=request.param,
cv_type="KFold",
cv_params=dict(n_splits=3, random_state=1),
)
