Tabular Regression

The following example shows how to fit a sample regression model with AutoPyTorch

import os
import tempfile as tmp
import warnings

import sklearn.datasets
import sklearn.model_selection

os.environ['JOBLIB_TEMP_FOLDER'] = tmp.gettempdir()
os.environ['OMP_NUM_THREADS'] = '1'
os.environ['OPENBLAS_NUM_THREADS'] = '1'
os.environ['MKL_NUM_THREADS'] = '1'

warnings.simplefilter(action='ignore', category=UserWarning)
warnings.simplefilter(action='ignore', category=FutureWarning)

from autoPyTorch.api.tabular_regression import TabularRegressionTask

Data Loading

X, y = sklearn.datasets.fetch_openml(name='boston', return_X_y=True, as_frame=True)
X_train, X_test, y_train, y_test = sklearn.model_selection.train_test_split(
    X,
    y,
    random_state=1,
)

Build and fit a regressor

api = TabularRegressionTask()

Search for an ensemble of machine learning algorithms

api.search(
    X_train=X_train,
    y_train=y_train,
    X_test=X_test.copy(),
    y_test=y_test.copy(),
    optimize_metric='r2',
    total_walltime_limit=300,
    func_eval_time_limit_secs=50,
    dataset_name="Boston"
)

<autoPyTorch.api.tabular_regression.TabularRegressionTask object at 0x7f6ea60a78b0>

Print the final ensemble performance before refit

y_pred = api.predict(X_test)
score = api.score(y_pred, y_test)
print(score)

# Print statistics from search
print(api.sprint_statistics())

{'r2': 0.9407884171054208}
autoPyTorch results:
        Dataset name: Boston
        Optimisation Metric: r2
        Best validation score: 0.8670098636440993
        Number of target algorithm runs: 23
        Number of successful target algorithm runs: 22
        Number of crashed target algorithm runs: 0
        Number of target algorithms that exceeded the time limit: 1
        Number of target algorithms that exceeded the memory limit: 0

Refit the models on the full dataset.

api.refit(
    X_train=X_train,
    y_train=y_train,
    X_test=X_test,
    y_test=y_test,
    dataset_name="Boston",
    total_walltime_limit=500,
    run_time_limit_secs=50
    # you can change the resampling strategy to
    # for example, CrossValTypes.k_fold_cross_validation
    # to fit k fold models and have a voting classifier
    # resampling_strategy=CrossValTypes.k_fold_cross_validation
)

<autoPyTorch.api.tabular_regression.TabularRegressionTask object at 0x7f6ea60a78b0>

Print the final ensemble performance after refit

y_pred = api.predict(X_test)
score = api.score(y_pred, y_test)
print(score)

# Print the final ensemble built by AutoPyTorch
print(api.show_models())

{'r2': 0.9378277102071694}
|    | Preprocessing                                                                                    | Estimator                                                       |   Weight |
|---:|:-------------------------------------------------------------------------------------------------|:----------------------------------------------------------------|---------:|
|  0 | None                                                                                             | CBLearner                                                       |     0.44 |
|  1 | SimpleImputer,Variance Threshold,NoCoalescer,OneHotEncoder,StandardScaler,NoFeaturePreprocessing | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential |     0.42 |
|  2 | SimpleImputer,Variance Threshold,NoCoalescer,OneHotEncoder,StandardScaler,NoFeaturePreprocessing | no embedding,ShapedMLPBackbone,FullyConnectedHead,nn.Sequential |     0.1  |
|  3 | None                                                                                             | LGBMLearner                                                     |     0.04 |