from sklearn_benchmarks.reporting.hp_match import HpMatchReporting
from sklearn_benchmarks.utils import default_run_dir, default_report_config
from pathlib import Path
import pandas as pd
pd.set_option('display.max_colwidth', None)
pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', None)

scikit-learn-intelex (Intel® oneAPI) vs. scikit-learn

run_dir = default_run_dir()
report_config = default_report_config("scikit_learn_intelex_vs_scikit_learn")

# Parameters
report_config = {
    "benchmarking_estimator_parameters": {
        "sklearn_KNeighborsClassifier_brute_force": [
            "n_neighbors",
            "algorithm",
            "n_jobs",
        ],
        "sklearnex_KNeighborsClassifier_brute_force": [],
        "sklearn_KNeighborsClassifier_kd_tree": ["n_neighbors", "algorithm", "n_jobs"],
        "sklearnex_KNeighborsClassifier_kd_tree": [],
        "sklearn_KMeans_tall": [
            "init",
            "algorithm",
            "n_clusters",
            "max_iter",
            "n_init",
            "tol",
        ],
        "sklearnex_KMeans_tall": [],
        "sklearn_KMeans_short": [
            "init",
            "algorithm",
            "n_clusters",
            "max_iter",
            "n_init",
            "tol",
        ],
        "sklearnex_KMeans_short": [],
        "sklearn_LogisticRegression": [
            "penalty",
            "dual",
            "tol",
            "C",
            "fit_intercept",
            "intercept_scaling",
            "class_weight",
            "random_state",
            "solver",
            "max_iter",
            "multi_class",
            "verbose",
            "warm_start",
            "n_jobs",
            "l1_ratio",
        ],
        "sklearnex_LogisticRegression": [],
        "sklearn_LinearRegression": [],
        "sklearnex_LinearRegression": [],
        "sklearn_Ridge": ["alpha"],
        "sklearnex_Ridge": [],
        "sklearn_TSNE": [],
        "sklearnex_TSNE": [],
        "sklearn_PCA": ["n_components"],
        "sklearnex_PCA": [],
        "sklearn_HistGradientBoostingClassifier_best": [
            "learning_rate",
            "n_iter_no_change",
            "max_leaf_nodes",
            "max_bins",
            "min_samples_leaf",
            "max_iter",
        ],
        "sklearn_HistGradientBoostingClassifier": [
            "learning_rate",
            "n_iter_no_change",
            "max_leaf_nodes",
            "max_bins",
            "min_samples_leaf",
            "max_iter",
            "early_stopping",
        ],
        "lightgbm": [
            "learning_rate",
            "num_leaves",
            "max_bin",
            "min_data_in_leaf",
            "bagging_fraction",
            "feature_fraction_bynode",
            "n_estimators",
        ],
        "xgboost": [
            "validate_parameters",
            "tree_method",
            "use_label_encoder",
            "learning_rate",
            "max_leaves",
            "max_bin",
            "min_child_weight",
            "subsample",
            "colsample_bynode",
            "n_estimators",
        ],
        "catboost_lossguide": [
            "allow_writing_files",
            "grow_policy",
            "bootstrap_type",
            "learning_rate",
            "max_leaves",
            "border_count",
            "min_data_in_leaf",
            "subsample",
            "rsm",
            "max_depth",
            "iterations",
        ],
        "catboost_symmetric": [
            "allow_writing_files",
            "grow_policy",
            "bootstrap_type",
            "learning_rate",
            "border_count",
            "min_data_in_leaf",
            "subsample",
            "rsm",
            "max_depth",
            "iterations",
        ],
    },
    "parameters": {"n_cols": 2},
    "log_scale": True,
    "other_library": "sklearnex",
    "estimators": [
        {
            "title": "KNeighborsClassifier (brute force)",
            "name": "KNeighborsClassifier_brute_force",
            "split_bars_by_parameters": ["n_jobs"],
        },
        {
            "title": "KNeighborsClassifier (KD tree)",
            "name": "KNeighborsClassifier_kd_tree",
            "split_bars_by_parameters": ["n_jobs"],
        },
        {"title": "KMeans (tall)", "name": "KMeans_tall"},
        {"title": "KMeans (short)", "name": "KMeans_short"},
        {"title": "Logistic Regression", "name": "LogisticRegression"},
        {"title": "LinearRegression", "name": "LinearRegression"},
        {"title": "Ridge", "name": "Ridge"},
        {"title": "TSNE", "name": "TSNE"},
        {"title": "PCA", "name": "PCA"},
    ],
}
run_dir = "./results/local/20220316T125341/"

run_dir = Path(run_dir)

reporting = HpMatchReporting(**report_config, run_dir=run_dir)
reporting.make_report()

We assume here there is a perfect match between the hyperparameters of both librairies. For a given set of parameters and a given dataset, we compute the speed-up time scikit-learn / time sklearnex. For instance, a speed-up of 2 means that sklearnex is twice as fast as scikit-learn for a given set of parameters and a given dataset.

KNeighborsClassifier (brute force) ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: algorithm=brute.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	100000	100000	100	1.161	0.030	0.069	0.0	-1	1	0.049	0.001	23.633	23.635	Download	Download
3	fit	100000	100000	100	1.165	0.047	0.069	0.0	-1	5	0.044	0.001	26.588	26.590	Download	Download
6	fit	100000	100000	100	1.164	0.023	0.069	0.0	1	100	0.044	0.001	26.396	26.398	Download	Download
9	fit	100000	100000	100	1.120	0.037	0.071	0.0	-1	100	0.049	0.000	23.011	23.011	Download	Download
12	fit	100000	100000	100	1.170	0.025	0.068	0.0	1	5	0.044	0.001	26.814	26.817	Download	Download
15	fit	100000	100000	100	1.182	0.024	0.068	0.0	1	1	0.049	0.000	24.135	24.135	Download	Download
18	fit	100000	100000	2	0.045	0.001	0.036	0.0	-1	1	0.008	0.000	5.346	5.348	Download	Download
21	fit	100000	100000	2	0.045	0.002	0.035	0.0	-1	5	0.008	0.000	5.353	5.355	Download	Download
24	fit	100000	100000	2	0.045	0.001	0.035	0.0	1	100	0.008	0.000	5.437	5.440	Download	Download
27	fit	100000	100000	2	0.045	0.002	0.036	0.0	-1	100	0.008	0.000	5.402	5.402	Download	Download
30	fit	100000	100000	2	0.045	0.002	0.035	0.0	1	5	0.008	0.000	5.383	5.385	Download	Download
33	fit	100000	100000	2	0.044	0.002	0.036	0.0	1	1	0.008	0.000	5.292	5.293	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	100000	1000	100	2.259	0.121	0.000	0.002	-1	1	0.169	0.003	13.355	13.358	Download	Download
2	predict	100000	1	100	0.020	0.002	0.000	0.020	-1	1	0.008	0.000	2.556	2.558	Download	Download
4	predict	100000	1000	100	2.814	0.066	0.000	0.003	-1	5	0.170	0.002	16.580	16.580	Download	Download
5	predict	100000	1	100	0.022	0.002	0.000	0.022	-1	5	0.008	0.000	2.827	2.828	Download	Download
7	predict	100000	1000	100	1.846	0.003	0.000	0.002	1	100	0.205	0.001	9.020	9.020	Download	Download
8	predict	100000	1	100	0.018	0.000	0.000	0.018	1	100	0.008	0.000	2.360	2.361	Download	Download
10	predict	100000	1000	100	2.606	0.048	0.000	0.003	-1	100	0.205	0.002	12.720	12.721	Download	Download
11	predict	100000	1	100	0.021	0.002	0.000	0.021	-1	100	0.008	0.000	2.694	2.696	Download	Download
13	predict	100000	1000	100	1.849	0.016	0.000	0.002	1	5	0.168	0.001	11.024	11.024	Download	Download
14	predict	100000	1	100	0.019	0.000	0.000	0.019	1	5	0.008	0.000	2.406	2.406	Download	Download
16	predict	100000	1000	100	1.170	0.017	0.001	0.001	1	1	0.166	0.001	7.049	7.049	Download	Download
17	predict	100000	1	100	0.017	0.000	0.000	0.017	1	1	0.008	0.000	2.212	2.213	Download	Download
19	predict	100000	1000	2	1.683	0.024	0.000	0.002	-1	1	0.026	0.000	65.311	65.315	Download	Download
20	predict	100000	1	2	0.005	0.002	0.000	0.005	-1	1	0.001	0.000	5.789	5.818	Download	Download
22	predict	100000	1000	2	2.423	0.022	0.000	0.002	-1	5	0.027	0.000	89.728	89.730	Download	Download
23	predict	100000	1	2	0.006	0.002	0.000	0.006	-1	5	0.001	0.000	7.983	8.105	Download	Download
25	predict	100000	1000	2	1.809	0.008	0.000	0.002	1	100	0.060	0.000	30.080	30.080	Download	Download
26	predict	100000	1	2	0.003	0.000	0.000	0.003	1	100	0.001	0.000	3.047	3.054	Download	Download
28	predict	100000	1000	2	2.460	0.021	0.000	0.002	-1	100	0.060	0.000	40.773	40.774	Download	Download
29	predict	100000	1	2	0.007	0.002	0.000	0.007	-1	100	0.001	0.000	8.006	8.149	Download	Download
31	predict	100000	1000	2	1.790	0.005	0.000	0.002	1	5	0.027	0.000	66.836	66.837	Download	Download
32	predict	100000	1	2	0.003	0.000	0.000	0.003	1	5	0.001	0.000	3.416	3.423	Download	Download
34	predict	100000	1000	2	1.034	0.005	0.000	0.001	1	1	0.026	0.000	39.874	39.876	Download	Download
35	predict	100000	1	2	0.002	0.000	0.000	0.002	1	1	0.001	0.000	2.201	2.224	Download	Download

Profiling traces can be visualized using Perfetto UI.

KNeighborsClassifier (KD tree) ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: algorithm=kd_tree.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000000	1000000	10	2.729	0.074	0.029	0.0	-1	1	0.692	0.013	3.942	3.943	Download	Download
3	fit	1000000	1000000	10	2.766	0.047	0.029	0.0	-1	5	0.687	0.014	4.027	4.028	Download	Download
6	fit	1000000	1000000	10	2.627	0.050	0.030	0.0	1	100	0.686	0.013	3.826	3.827	Download	Download
9	fit	1000000	1000000	10	2.691	0.062	0.030	0.0	-1	100	0.698	0.021	3.858	3.860	Download	Download
12	fit	1000000	1000000	10	2.656	0.051	0.030	0.0	1	5	0.679	0.005	3.909	3.910	Download	Download
15	fit	1000000	1000000	10	2.751	0.084	0.029	0.0	1	1	0.657	0.008	4.185	4.185	Download	Download
18	fit	1000	1000	2	0.001	0.000	0.029	0.0	-1	1	0.001	0.000	0.562	0.567	Download	Download
21	fit	1000	1000	2	0.001	0.000	0.029	0.0	-1	5	0.001	0.000	0.592	0.599	Download	Download
24	fit	1000	1000	2	0.001	0.000	0.030	0.0	1	100	0.001	0.000	0.572	0.578	Download	Download
27	fit	1000	1000	2	0.001	0.000	0.030	0.0	-1	100	0.001	0.000	0.576	0.584	Download	Download
30	fit	1000	1000	2	0.001	0.000	0.030	0.0	1	5	0.001	0.000	0.577	0.583	Download	Download
33	fit	1000	1000	2	0.001	0.000	0.029	0.0	1	1	0.001	0.000	0.574	0.580	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	1000000	1000	10	0.398	0.005	0.000	0.000	-1	1	0.097	0.003	4.120	4.121	Download	Download
2	predict	1000000	1	10	0.003	0.000	0.000	0.003	-1	1	0.000	0.000	9.516	10.218	Download	Download
4	predict	1000000	1000	10	0.751	0.014	0.000	0.001	-1	5	0.177	0.013	4.238	4.249	Download	Download
5	predict	1000000	1	10	0.003	0.000	0.000	0.003	-1	5	0.000	0.000	9.497	10.176	Download	Download
7	predict	1000000	1000	10	4.195	0.042	0.000	0.004	1	100	0.524	0.016	8.009	8.013	Download	Download
8	predict	1000000	1	10	0.002	0.001	0.000	0.002	1	100	0.001	0.000	4.334	4.577	Download	Download
10	predict	1000000	1000	10	2.450	0.043	0.000	0.002	-1	100	0.521	0.009	4.707	4.708	Download	Download
11	predict	1000000	1	10	0.004	0.001	0.000	0.004	-1	100	0.001	0.000	7.122	7.473	Download	Download
13	predict	1000000	1000	10	1.235	0.018	0.000	0.001	1	5	0.167	0.002	7.400	7.400	Download	Download
14	predict	1000000	1	10	0.001	0.000	0.000	0.001	1	5	0.000	0.000	3.787	4.076	Download	Download
16	predict	1000000	1000	10	0.678	0.009	0.000	0.001	1	1	0.096	0.002	7.055	7.056	Download	Download
17	predict	1000000	1	10	0.001	0.000	0.000	0.001	1	1	0.000	0.000	3.485	3.732	Download	Download
19	predict	1000	1000	2	0.023	0.000	0.001	0.000	-1	1	0.001	0.000	43.785	44.988	Download	Download
20	predict	1000	1	2	0.002	0.000	0.000	0.002	-1	1	0.000	0.000	14.839	17.360	Download	Download
22	predict	1000	1000	2	0.025	0.000	0.001	0.000	-1	5	0.001	0.000	29.977	30.407	Download	Download
23	predict	1000	1	2	0.002	0.000	0.000	0.002	-1	5	0.000	0.000	15.124	17.251	Download	Download
25	predict	1000	1000	2	0.035	0.000	0.000	0.000	1	100	0.005	0.000	7.527	7.541	Download	Download
26	predict	1000	1	2	0.001	0.000	0.000	0.001	1	100	0.000	0.000	3.564	4.096	Download	Download
28	predict	1000	1000	2	0.036	0.000	0.000	0.000	-1	100	0.005	0.000	7.741	7.749	Download	Download
29	predict	1000	1	2	0.002	0.000	0.000	0.002	-1	100	0.000	0.000	13.482	15.441	Download	Download
31	predict	1000	1000	2	0.022	0.000	0.001	0.000	1	5	0.001	0.000	26.606	26.935	Download	Download
32	predict	1000	1	2	0.001	0.000	0.000	0.001	1	5	0.000	0.000	3.641	4.303	Download	Download
34	predict	1000	1000	2	0.020	0.000	0.001	0.000	1	1	0.001	0.000	37.709	39.074	Download	Download
35	predict	1000	1	2	0.001	0.000	0.000	0.001	1	1	0.000	0.000	3.663	4.322	Download	Download

Profiling traces can be visualized using Perfetto UI.

KMeans (tall) ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: algorithm=full, n_clusters=3, max_iter=30, n_init=1, tol=1e-16.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	init	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000000	1000000	2	0.481	0.003	0.033	0.0	30	random	0.367	0.028	1.310	1.314	Download	Download
3	fit	1000000	1000000	2	0.558	0.005	0.029	0.0	30	k-means++	0.428	0.030	1.305	1.308	Download	Download
6	fit	1000000	1000000	100	4.647	0.178	0.172	0.0	30	random	2.635	0.053	1.764	1.764	Download	Download
9	fit	1000000	1000000	100	4.774	0.075	0.168	0.0	30	k-means++	2.751	0.056	1.735	1.735	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	init	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	1000000	1000	2	0.0	0.0	0.066	0.0	30	random	0.000	0.0	1.180	1.322	Download	Download
2	predict	1000000	1	2	0.0	0.0	0.000	0.0	30	random	0.000	0.0	1.294	1.490	Download	Download
4	predict	1000000	1000	2	0.0	0.0	0.064	0.0	30	k-means++	0.000	0.0	1.112	1.220	Download	Download
5	predict	1000000	1	2	0.0	0.0	0.000	0.0	30	k-means++	0.000	0.0	1.334	1.575	Download	Download
7	predict	1000000	1000	100	0.0	0.0	1.902	0.0	30	random	0.001	0.0	0.776	0.989	Download	Download
8	predict	1000000	1	100	0.0	0.0	0.004	0.0	30	random	0.000	0.0	1.002	1.155	Download	Download
10	predict	1000000	1000	100	0.0	0.0	1.871	0.0	30	k-means++	0.000	0.0	1.352	1.466	Download	Download
11	predict	1000000	1	100	0.0	0.0	0.004	0.0	30	k-means++	0.000	0.0	1.250	1.448	Download	Download

Profiling traces can be visualized using Perfetto UI.

KMeans (short) ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: algorithm=full, n_clusters=300, max_iter=20, n_init=1, tol=1e-16.

Mismatches between validation scores

WARNING! Mismatch between validation scores for 2 predictions (25.0%).
The observed differences can be found in the diff_adjusted_rand_scores column.
The chosen difference threshold is 0.01 for adjusted_rand_score.
The biggest difference is 0.077 for adjusted_rand_score.
See details in the dataframe below.

	estimator	library	diff_adjusted_rand_scores	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	algorithm	init	max_iter	n_clusters	n_init	tol	adjusted_rand_score_sklearn	mean_duration_sklearnex	std_duration_sklearnex	adjusted_rand_score_sklearnex	speedup	std_speedup
7	sklearn_KMeans_short	sklearn	0.074768	predict	10000	1000	100	0.001333	0.000239	0.599971	0.000001	20	full	random	20	300	1	1.000000e-16	0.222701	0.001153	0.000145	0.297469	1.156055	1.165137
10	sklearn_KMeans_short	sklearn	0.076616	predict	10000	1000	100	0.001371	0.000179	0.583326	0.000001	20	full	k-means++	20	300	1	1.000000e-16	0.408030	0.001138	0.000151	0.331414	1.204783	1.215317

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	init	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	10000	10000	2	0.081	0.005	0.002	0.0	20	random	0.034	0.002	2.383	2.387	Download	Download
3	fit	10000	10000	2	0.204	0.011	0.001	0.0	20	k-means++	0.087	0.002	2.360	2.361	Download	Download
6	fit	10000	10000	100	0.200	0.015	0.040	0.0	20	random	0.126	0.001	1.579	1.579	Download	Download
9	fit	10000	10000	100	0.568	0.021	0.014	0.0	20	k-means++	0.325	0.019	1.750	1.753	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	init	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	10000	1000	2	0.001	0.0	0.023	0.0	20	random	0.001	0.0	0.987	0.993	Download	Download
2	predict	10000	1	2	0.000	0.0	0.000	0.0	20	random	0.000	0.0	1.248	1.450	Download	Download
4	predict	10000	1000	2	0.001	0.0	0.025	0.0	20	k-means++	0.001	0.0	0.867	0.875	Download	Download
5	predict	10000	1	2	0.000	0.0	0.000	0.0	20	k-means++	0.000	0.0	1.196	1.412	Download	Download
7	predict	10000	1000	100	0.001	0.0	0.600	0.0	20	random	0.001	0.0	1.156	1.165	Download	Download
8	predict	10000	1	100	0.000	0.0	0.003	0.0	20	random	0.000	0.0	1.300	1.441	Download	Download
10	predict	10000	1000	100	0.001	0.0	0.583	0.0	20	k-means++	0.001	0.0	1.205	1.215	Download	Download
11	predict	10000	1	100	0.000	0.0	0.003	0.0	20	k-means++	0.000	0.0	0.968	1.045	Download	Download

Profiling traces can be visualized using Perfetto UI.

Logistic Regression ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: penalty=l2, dual=False, tol=0.0001, C=1.0, fit_intercept=True, intercept_scaling=1.0, class_weight=nan, random_state=nan, solver=lbfgs, max_iter=100, multi_class=auto, verbose=0, warm_start=False, n_jobs=nan, l1_ratio=nan.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000000	1000000	100	10.091	0.055	0.079	0.000	[20]	1.741	0.049	5.796	5.799	Download	Download
3	fit	1000	1000	10000	0.711	0.050	0.112	0.001	[27]	0.691	0.039	1.030	1.031	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	1000000	1000	100	0.000	0.0	2.935	0.0	[20]	0.000	0.0	0.707	0.751	Download	Download
2	predict	1000000	1	100	0.000	0.0	0.012	0.0	[20]	0.000	0.0	0.292	0.330	Download	Download
4	predict	1000	100	10000	0.002	0.0	5.298	0.0	[27]	0.003	0.0	0.536	0.539	Download	Download
5	predict	1000	1	10000	0.000	0.0	0.934	0.0	[27]	0.001	0.0	0.121	0.124	Download	Download

Profiling traces can be visualized using Perfetto UI.

LinearRegression ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters:

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	copy_X	fit_intercept	normalize	positive	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000	1000	10000	1.245	0.009	0.064	0.001	True	True	deprecated	False	1.238	0.004	1.005	1.005	Download	Download
3	fit	1000000	1000000	100	7.287	0.027	0.110	0.000	True	True	deprecated	False	0.182	0.001	39.974	39.975	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	copy_X	fit_intercept	normalize	positive	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	1000	1000	10000	0.01	0.0	8.114	0.0	True	True	deprecated	False	0.016	0.0	0.605	0.605	Download	Download
2	predict	1000	10	10000	0.00	0.0	6.419	0.0	True	True	deprecated	False	0.000	0.0	0.437	0.468	Download	Download
4	predict	1000000	1000	100	0.00	0.0	6.215	0.0	True	True	deprecated	False	0.000	0.0	0.294	0.308	Download	Download
5	predict	1000000	10	100	0.00	0.0	0.119	0.0	True	True	deprecated	False	0.000	0.0	0.425	0.477	Download	Download

Profiling traces can be visualized using Perfetto UI.

Ridge ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: alpha=1e-06.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000	1000	10000	0.161	0.001	0.496	0.0	0.149	0.001	1.083	1.083	Download	Download
3	fit	1000000	1000000	100	1.079	0.008	0.741	0.0	0.217	0.001	4.976	4.976	Download	Download

predict

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
1	predict	1000	1000	10000	0.01	0.0	8.227	0.0	0.016	0.0	0.592	0.592	Download	Download
2	predict	1000	10	10000	0.00	0.0	6.235	0.0	0.000	0.0	0.428	0.452	Download	Download
4	predict	1000000	1000	100	0.00	0.0	5.830	0.0	0.000	0.0	0.462	0.499	Download	Download
5	predict	1000000	10	100	0.00	0.0	0.148	0.0	0.000	0.0	0.329	0.370	Download	Download

Profiling traces can be visualized using Perfetto UI.

TSNE ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters:

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_iter	angle	early_exaggeration	init	learning_rate	method	metric	min_grad_norm	n_components	n_iter_without_progress	perplexity	square_distances	verbose	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	1000	1000	3	3.451	0.154	0.0	0.003	1000	0.5	12.0	warn	warn	barnes_hut	euclidean	0.0	2	300	30.0	legacy	0	3.628	0.358	0.951	0.956	Download	Download

Profiling traces can be visualized using Perfetto UI.

PCA ¶

scikit-learn-intelex (2021.20220207.123530) vs. scikit-learn (1.0.2)

Speedup barplots ¶

All estimators share the following parameters: n_components=10.

Raw results ¶

fit

	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	mean_duration_sklearnex	std_duration_sklearnex	speedup	std_speedup	sklearn_profiling	sklearnex_profiling
0	fit	100000	100000	100	0.420	0.003	0.190	0.0	0.025	0.000	16.534	16.535	Download	Download
1	fit	10000	10000	1000	0.371	0.005	0.216	0.0	0.180	0.002	2.062	2.062	Download	Download

Profiling traces can be visualized using Perfetto UI.

Benchmark environment information

System

python	3.8.12 | packaged by conda-forge | (default, Jan 30 2022, 23:42:07) [GCC 9.4.0]
executable	/usr/share/miniconda/envs/sklbench/bin/python
machine	Linux-5.11.0-1028-azure-x86_64-with-glibc2.10

Dependencies

pip	22.0.4
setuptools	60.9.3
sklearn	1.0.2
numpy	1.22.3
scipy	1.8.0
Cython	None
pandas	1.4.1
matplotlib	3.5.1
joblib	1.1.0
threadpoolctl	3.1.0

Threadpool

	user_api	internal_api	prefix	filepath	version	threading_layer	architecture	num_threads
0	blas	openblas	libopenblas	/usr/share/miniconda/envs/sklbench/lib/libopenblasp-r0.3.18.so	0.3.18	pthreads	SkylakeX	2
1	openmp	openmp	libgomp	/usr/share/miniconda/envs/sklbench/lib/libgomp.so.1.0.0	None	NaN	NaN	2

Cpu_count

cpu_count	2
physical_cpu_count	2