def setUp(self) -> None:
super(TestScaler, self).setUp()
boston = load_boston()
np.random.seed(10)
ix = np.random.permutation(boston.data.shape[0])
L = int(len(ix) * 0.8)
train_ix = ix[:L]
test_ix = ix[L:]
X_train = pd.DataFrame(boston.data[train_ix, :],
columns=boston.feature_names)
y_train = boston.target[train_ix]
X_test = pd.DataFrame(boston.data[test_ix, :],
columns=boston.feature_names)
y_test = boston.target[test_ix]
pipe = AutoFlowRegressor(
# consider_ordinal_as_cat=True,
resource_manager=self.mock_resource_manager)
pipe.fit(
X_train=X_train,
y_train=y_train,
X_test=X_test,
y_test=y_test,
is_not_realy_run=True,
)
self.X_train = pipe.data_manager.X_train
self.X_train.index = train_ix
self.y_train = pipe.data_manager.y_train
self.X_test = pipe.data_manager.X_test
self.X_test.index = test_ix
self.y_test = pipe.data_manager.y_test
self.index = deepcopy(train_ix)
def test_single_regressor_with_X_test(self):
X, y = load_boston(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y)
pipe = AutoFlowRegressor(DAG_workflow={"num->target": ["elasticnet"]},
initial_runs=1,
run_limit=1,
debug=True,
resource_manager=self.mock_resource_manager)
pipe.fit(X_train, y_train, X_test, y_test)
# score = accuracy_score(y_test, y_pred)
score = pipe.score(X_test, y_test)
print(score)
self.assertGreater(score, 0)
pipe.resource_manager.init_trial_table()
trial = pipe.resource_manager.TrialModel
records = trial.select().where(
trial.experiment_id == pipe.experiment_id)
for record in records:
self.assertTrue(record is not None)
self.assertTrue(
isinstance(record.test_all_score, dict)
and bool(record.test_all_score)
# and record.test_all_score["r2"] > 0
)
# print(record.test_all_score["r2"])
pipe.resource_manager.close_trial_table()
def test_single_regressor(self):
X, y = load_boston(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y)
pipe = AutoFlowRegressor(DAG_workflow={"num->target": ["elasticnet"]},
initial_runs=1,
run_limit=1,
debug=True,
resource_manager=self.mock_resource_manager)
pipe.fit(X_train, y_train)
# score = accuracy_score(y_test, y_pred)
score = pipe.score(X_test, y_test)
print(score)
self.assertGreater(score, 0)
def test_ensemble_regressors(self):
X, y = load_boston(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y)
pipe = AutoFlowRegressor(
DAG_workflow={
"num->scaled": ["scale.standardize"],
"scaled->target": ["elasticnet"]
},
initial_runs=2,
run_limit=2,
n_jobs=2,
resource_manager=self.mock_resource_manager,
debug=True,
)
pipe.fit(X_train,
y_train,
splitter=ShuffleSplit(n_splits=2,
test_size=0.2,
random_state=42))
score = pipe.score(X_test, y_test)
print(score)
assert score > 0.5
for splitter in [
# LeaveOneOut(),
ShuffleSplit(n_splits=20, test_size=0.3, random_state=42),
KFold()
]:
pipe.fit(X_train, y_train, splitter=splitter)
score = pipe.score(X_test, y_test)
assert score > 0.5
print("splitter:", splitter)
print("test r2:", score)
import os
import joblib
import numpy as np
import pandas as pd
from sklearn.model_selection import KFold
from autoflow import AutoFlowRegressor
train_df = pd.read_csv("../data/train_regression.csv")
train_df.replace("NA", np.nan, inplace=True)
test_df = pd.read_csv("../data/test_regression.csv")
test_df.replace("NA", np.nan, inplace=True)
trained_pipeline = AutoFlowRegressor(initial_runs=5,
run_limit=10,
n_jobs=1,
included_regressors=["lightgbm"],
per_run_time_limit=60)
column_descriptions = {
"id": "Id",
"target": "SalePrice",
}
if not os.path.exists("autoflow_regression.bz2"):
trained_pipeline.fit(X_train=train_df,
X_test=test_df,
column_descriptions=column_descriptions,
splitter=KFold(n_splits=3,
shuffle=True,
random_state=42),
fit_ensemble_params=False)
# if you want to see the workflow AutoFlow is searching, you can use `draw_workflow_space` to visualize