knn_clsf.py

# ===============================================================================
# Copyright 2020-2021 Intel Corporation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================

import argparse

import bench
import numpy as np


def main():
    from sklearn.neighbors import KNeighborsClassifier

    # Load generated data
    X_train, X_test, y_train, y_test = bench.load_data(params)
    params.n_classes = len(np.unique(y_train))

    # Create classification object
    knn_clsf = KNeighborsClassifier(n_neighbors=params.n_neighbors,
                                    weights=params.weights,
                                    algorithm=params.method,
                                    metric=params.metric,
                                    n_jobs=params.n_jobs)

    # Measure time and accuracy on fitting
    train_time, _ = bench.measure_function_time(
        knn_clsf.fit, X_train, y_train, params=params)
    if params.task == 'classification':
        y_pred = knn_clsf.predict(X_train)
        y_proba = knn_clsf.predict_proba(X_train)
        train_acc = bench.accuracy_score(y_train, y_pred)
        train_log_loss = bench.log_loss(y_train, y_proba)
        train_roc_auc = bench.roc_auc_score(y_train, y_proba)

    # Measure time and accuracy on prediction
    if params.task == 'classification':
        predict_time, yp = bench.measure_function_time(knn_clsf.predict, X_test,
                                                       params=params)
        y_proba = knn_clsf.predict_proba(X_test)
        test_acc = bench.accuracy_score(y_test, yp)
        test_log_loss = bench.log_loss(y_test, y_proba)
        test_roc_auc = bench.roc_auc_score(y_test, y_proba)
    else:
        predict_time, _ = bench.measure_function_time(knn_clsf.kneighbors, X_test,
                                                      params=params)

    if params.task == 'classification':
        bench.print_output(
            library='sklearn',
            algorithm=knn_clsf._fit_method + '_knn_clsf',
            stages=['training', 'prediction'],
            params=params,
            functions=['knn_clsf.fit', 'knn_clsf.predict'],
            times=[train_time, predict_time],
            metric_type=['accuracy', 'log_loss', 'roc_auc'],
            metrics=[
                [train_acc, test_acc],
                [train_log_loss, test_log_loss],
                [train_roc_auc, test_roc_auc],
            ],
            data=[X_train, X_test],
            alg_instance=knn_clsf,
        )
    else:
        bench.print_output(
            library='sklearn',
            algorithm=knn_clsf._fit_method + '_knn_search',
            stages=['training', 'search'],
            params=params,
            functions=['knn_clsf.fit', 'knn_clsf.kneighbors'],
            times=[train_time, predict_time],
            metric_type=None,
            metrics=[],
            data=[X_train, X_test],
            alg_instance=knn_clsf,
        )


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description='scikit-learn kNN classifier benchmark')

    parser.add_argument('--task', default='classification', type=str,
                        choices=('search', 'classification'),
                        help='kNN task: search or classification')
    parser.add_argument('--n-neighbors', default=5, type=int,
                        help='Number of neighbors to use')
    parser.add_argument('--weights', type=str, default='uniform',
                        help='Weight function used in prediction')
    parser.add_argument('--method', type=str, default='brute',
                        choices=('brute', 'kd_tree', 'ball_tree', 'auto'),
                        help='Algorithm used to compute the nearest neighbors')
    parser.add_argument('--metric', type=str, default='euclidean',
                        help='Distance metric to use')
    params = bench.parse_args(parser)
    bench.run_with_context(params, main)