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 = {
    "other_library": "sklearnex",
    "parameters": {"n_cols": 2},
    "log_scale": True,
    "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",
        ],
    },
    "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/20220316T125727/"

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	1.423	0.046	0.056	0.0	-1	1	0.060	0.003	23.516	23.553	Download	Download
3	fit	100000	100000	100	1.323	0.014	0.060	0.0	-1	5	0.052	0.000	25.675	25.676	Download	Download
6	fit	100000	100000	100	1.415	0.057	0.057	0.0	1	100	0.051	0.000	27.523	27.523	Download	Download
9	fit	100000	100000	100	1.398	0.038	0.057	0.0	-1	100	0.056	0.000	24.788	24.788	Download	Download
12	fit	100000	100000	100	1.427	0.039	0.056	0.0	1	5	0.052	0.000	27.332	27.333	Download	Download
15	fit	100000	100000	100	1.416	0.042	0.056	0.0	1	1	0.057	0.000	24.960	24.960	Download	Download
18	fit	100000	100000	2	0.055	0.001	0.029	0.0	-1	1	0.010	0.000	5.277	5.278	Download	Download
21	fit	100000	100000	2	0.055	0.000	0.029	0.0	-1	5	0.011	0.000	5.159	5.159	Download	Download
24	fit	100000	100000	2	0.055	0.000	0.029	0.0	1	100	0.011	0.000	5.217	5.218	Download	Download
27	fit	100000	100000	2	0.055	0.000	0.029	0.0	-1	100	0.010	0.000	5.333	5.334	Download	Download
30	fit	100000	100000	2	0.055	0.000	0.029	0.0	1	5	0.010	0.000	5.236	5.237	Download	Download
33	fit	100000	100000	2	0.055	0.001	0.029	0.0	1	1	0.010	0.000	5.299	5.301	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	3.300	0.133	0.000	0.003	-1	1	0.211	0.002	15.629	15.630	Download	Download
2	predict	100000	1	100	0.026	0.003	0.000	0.026	-1	1	0.010	0.002	2.565	2.597	Download	Download
4	predict	100000	1000	100	4.061	0.080	0.000	0.004	-1	5	0.218	0.006	18.609	18.616	Download	Download
5	predict	100000	1	100	0.028	0.004	0.000	0.028	-1	5	0.009	0.000	3.035	3.036	Download	Download
7	predict	100000	1000	100	2.230	0.004	0.000	0.002	1	100	0.267	0.006	8.355	8.357	Download	Download
8	predict	100000	1	100	0.023	0.000	0.000	0.023	1	100	0.010	0.000	2.391	2.393	Download	Download
10	predict	100000	1000	100	4.241	0.071	0.000	0.004	-1	100	0.276	0.009	15.375	15.384	Download	Download
11	predict	100000	1	100	0.029	0.003	0.000	0.029	-1	100	0.010	0.000	2.978	2.980	Download	Download
13	predict	100000	1000	100	2.242	0.005	0.000	0.002	1	5	0.225	0.009	9.945	9.953	Download	Download
14	predict	100000	1	100	0.024	0.000	0.000	0.024	1	5	0.010	0.001	2.397	2.400	Download	Download
16	predict	100000	1000	100	1.390	0.013	0.001	0.001	1	1	0.224	0.013	6.203	6.213	Download	Download
17	predict	100000	1	100	0.026	0.007	0.000	0.026	1	1	0.010	0.000	2.662	2.663	Download	Download
19	predict	100000	1000	2	2.916	0.090	0.000	0.003	-1	1	0.045	0.002	64.635	64.688	Download	Download
20	predict	100000	1	2	0.007	0.002	0.000	0.007	-1	1	0.001	0.000	5.862	5.942	Download	Download
22	predict	100000	1000	2	4.063	0.098	0.000	0.004	-1	5	0.039	0.003	104.285	104.533	Download	Download
23	predict	100000	1	2	0.007	0.002	0.000	0.007	-1	5	0.001	0.000	6.270	6.318	Download	Download
25	predict	100000	1000	2	2.181	0.003	0.000	0.002	1	100	0.084	0.003	25.979	25.992	Download	Download
26	predict	100000	1	2	0.003	0.000	0.000	0.003	1	100	0.002	0.000	2.031	2.074	Download	Download
28	predict	100000	1000	2	4.005	0.102	0.000	0.004	-1	100	0.084	0.004	47.462	47.517	Download	Download
29	predict	100000	1	2	0.007	0.002	0.000	0.007	-1	100	0.001	0.000	5.509	5.594	Download	Download
31	predict	100000	1000	2	2.178	0.010	0.000	0.002	1	5	0.038	0.002	57.150	57.210	Download	Download
32	predict	100000	1	2	0.003	0.000	0.000	0.003	1	5	0.001	0.000	3.054	3.100	Download	Download
34	predict	100000	1000	2	1.232	0.008	0.000	0.001	1	1	0.036	0.003	33.861	33.948	Download	Download
35	predict	100000	1	2	0.002	0.000	0.000	0.002	1	1	0.001	0.000	2.269	2.292	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	4.174	0.033	0.019	0.0	-1	1	1.030	0.024	4.051	4.052	Download	Download
3	fit	1000000	1000000	10	4.110	0.028	0.019	0.0	-1	5	1.018	0.042	4.039	4.043	Download	Download
6	fit	1000000	1000000	10	4.086	0.045	0.020	0.0	1	100	0.970	0.019	4.211	4.212	Download	Download
9	fit	1000000	1000000	10	4.008	0.031	0.020	0.0	-1	100	1.003	0.028	3.995	3.997	Download	Download
12	fit	1000000	1000000	10	4.070	0.016	0.020	0.0	1	5	1.017	0.037	4.001	4.003	Download	Download
15	fit	1000000	1000000	10	4.113	0.042	0.019	0.0	1	1	0.970	0.013	4.239	4.239	Download	Download
18	fit	1000	1000	2	0.001	0.000	0.025	0.0	-1	1	0.001	0.000	0.503	0.508	Download	Download
21	fit	1000	1000	2	0.001	0.000	0.022	0.0	-1	5	0.001	0.000	0.530	0.556	Download	Download
24	fit	1000	1000	2	0.001	0.000	0.024	0.0	1	100	0.002	0.000	0.432	0.449	Download	Download
27	fit	1000	1000	2	0.001	0.000	0.025	0.0	-1	100	0.002	0.002	0.345	0.486	Download	Download
30	fit	1000	1000	2	0.001	0.000	0.021	0.0	1	5	0.001	0.000	0.578	0.593	Download	Download
33	fit	1000	1000	2	0.001	0.000	0.025	0.0	1	1	0.001	0.000	0.455	0.457	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	0.528	0.016	0.000	0.001	-1	1	0.132	0.006	4.004	4.008	Download	Download
2	predict	1000000	1	10	0.004	0.001	0.000	0.004	-1	1	0.000	0.000	10.363	10.891	Download	Download
4	predict	1000000	1000	10	0.974	0.016	0.000	0.001	-1	5	0.225	0.017	4.333	4.346	Download	Download
5	predict	1000000	1	10	0.004	0.001	0.000	0.004	-1	5	0.000	0.000	10.519	11.073	Download	Download
7	predict	1000000	1000	10	5.494	0.036	0.000	0.005	1	100	0.641	0.021	8.576	8.580	Download	Download
8	predict	1000000	1	10	0.004	0.001	0.000	0.004	1	100	0.001	0.000	4.732	4.906	Download	Download
10	predict	1000000	1000	10	3.204	0.079	0.000	0.003	-1	100	0.691	0.022	4.637	4.640	Download	Download
11	predict	1000000	1	10	0.007	0.001	0.000	0.007	-1	100	0.001	0.000	8.777	9.104	Download	Download
13	predict	1000000	1000	10	1.660	0.009	0.000	0.002	1	5	0.228	0.008	7.283	7.287	Download	Download
14	predict	1000000	1	10	0.002	0.000	0.000	0.002	1	5	0.001	0.000	3.036	3.202	Download	Download
16	predict	1000000	1000	10	0.914	0.052	0.000	0.001	1	1	0.132	0.003	6.944	6.947	Download	Download
17	predict	1000000	1	10	0.001	0.000	0.000	0.001	1	1	0.000	0.000	3.264	3.493	Download	Download
19	predict	1000	1000	2	0.029	0.002	0.001	0.000	-1	1	0.001	0.000	34.501	36.027	Download	Download
20	predict	1000	1	2	0.004	0.001	0.000	0.004	-1	1	0.000	0.000	21.788	25.534	Download	Download
22	predict	1000	1000	2	0.033	0.003	0.000	0.000	-1	5	0.001	0.000	26.388	26.778	Download	Download
23	predict	1000	1	2	0.003	0.000	0.000	0.003	-1	5	0.000	0.000	9.423	9.893	Download	Download
25	predict	1000	1000	2	0.042	0.000	0.000	0.000	1	100	0.006	0.001	6.647	6.706	Download	Download
26	predict	1000	1	2	0.001	0.000	0.000	0.001	1	100	0.000	0.000	3.789	4.350	Download	Download
28	predict	1000	1000	2	0.044	0.002	0.000	0.000	-1	100	0.006	0.001	7.094	7.146	Download	Download
29	predict	1000	1	2	0.003	0.000	0.000	0.003	-1	100	0.000	0.000	10.978	11.982	Download	Download
31	predict	1000	1000	2	0.026	0.001	0.001	0.000	1	5	0.001	0.000	25.095	25.537	Download	Download
32	predict	1000	1	2	0.001	0.000	0.000	0.001	1	5	0.000	0.000	3.661	4.219	Download	Download
34	predict	1000	1000	2	0.024	0.000	0.001	0.000	1	1	0.001	0.000	27.630	29.226	Download	Download
35	predict	1000	1	2	0.001	0.000	0.000	0.001	1	1	0.000	0.000	3.612	4.145	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	0.633	0.013	0.025	0.0	30	random	0.487	0.035	1.299	1.303	Download	Download
3	fit	1000000	1000000	2	0.674	0.019	0.024	0.0	30	k-means++	0.536	0.036	1.257	1.259	Download	Download
6	fit	1000000	1000000	100	5.653	0.248	0.142	0.0	30	random	3.041	0.058	1.859	1.859	Download	Download
9	fit	1000000	1000000	100	5.886	0.112	0.136	0.0	30	k-means++	3.219	0.059	1.828	1.829	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	0.000	0.0	0.057	0.0	30	random	0.0	0.0	1.004	1.106	Download	Download
2	predict	1000000	1	2	0.000	0.0	0.000	0.0	30	random	0.0	0.0	1.063	1.208	Download	Download
4	predict	1000000	1000	2	0.000	0.0	0.055	0.0	30	k-means++	0.0	0.0	0.850	0.932	Download	Download
5	predict	1000000	1	2	0.000	0.0	0.000	0.0	30	k-means++	0.0	0.0	1.041	1.341	Download	Download
7	predict	1000000	1000	100	0.001	0.0	1.436	0.0	30	random	0.0	0.0	1.380	1.471	Download	Download
8	predict	1000000	1	100	0.000	0.0	0.003	0.0	30	random	0.0	0.0	1.134	1.269	Download	Download
10	predict	1000000	1000	100	0.001	0.0	1.164	0.0	30	k-means++	0.0	0.0	1.682	1.775	Download	Download
11	predict	1000000	1	100	0.000	0.0	0.003	0.0	30	k-means++	0.0	0.0	1.113	1.266	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	0.002200	0.000789	0.36363	0.000002	20	full	random	20	300	1	1.000000e-16	0.222701	0.001633	0.000213	0.297469	1.346842	1.358212
10	sklearn_KMeans_short	sklearn	0.076616	predict	10000	1000	100	0.001714	0.000191	0.46675	0.000002	20	full	k-means++	20	300	1	1.000000e-16	0.408030	0.001583	0.000228	0.331414	1.082483	1.093605

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	0.105	0.005	0.002	0.0	20	random	0.040	0.001	2.608	2.609	Download	Download
3	fit	10000	10000	2	0.259	0.005	0.001	0.0	20	k-means++	0.107	0.002	2.424	2.424	Download	Download
6	fit	10000	10000	100	0.254	0.006	0.031	0.0	20	random	0.177	0.007	1.436	1.437	Download	Download
9	fit	10000	10000	100	0.717	0.020	0.011	0.0	20	k-means++	0.427	0.009	1.677	1.677	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	0.000	0.020	0.0	20	random	0.001	0.0	0.883	0.890	Download	Download
2	predict	10000	1	2	0.000	0.000	0.000	0.0	20	random	0.000	0.0	1.209	1.376	Download	Download
4	predict	10000	1000	2	0.001	0.000	0.017	0.0	20	k-means++	0.001	0.0	1.064	1.074	Download	Download
5	predict	10000	1	2	0.000	0.000	0.000	0.0	20	k-means++	0.000	0.0	1.276	1.485	Download	Download
7	predict	10000	1000	100	0.002	0.001	0.364	0.0	20	random	0.002	0.0	1.347	1.358	Download	Download
8	predict	10000	1	100	0.000	0.000	0.002	0.0	20	random	0.000	0.0	1.178	1.269	Download	Download
10	predict	10000	1000	100	0.002	0.000	0.467	0.0	20	k-means++	0.002	0.0	1.082	1.094	Download	Download
11	predict	10000	1	100	0.000	0.000	0.003	0.0	20	k-means++	0.000	0.0	1.055	1.121	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	12.260	0.442	0.065	0.000	[20]	2.178	0.024	5.629	5.629	Download	Download
3	fit	1000	1000	10000	0.875	0.055	0.091	0.001	[27]	0.836	0.056	1.047	1.050	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	2.175	0.0	[20]	0.001	0.0	0.671	0.696	Download	Download
2	predict	1000000	1	100	0.000	0.0	0.009	0.0	[20]	0.000	0.0	0.295	0.318	Download	Download
4	predict	1000	100	10000	0.002	0.0	4.122	0.0	[27]	0.004	0.0	0.512	0.513	Download	Download
5	predict	1000	1	10000	0.000	0.0	0.731	0.0	[27]	0.001	0.0	0.122	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	2.099	0.043	0.038	0.002	True	True	deprecated	False	2.097	0.041	1.001	1.001	Download	Download
3	fit	1000000	1000000	100	8.965	0.068	0.089	0.000	True	True	deprecated	False	0.242	0.010	37.111	37.145	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	0.011	0.0	7.360	0.0	True	True	deprecated	False	0.018	0.0	0.588	0.588	Download	Download
2	predict	1000	10	10000	0.000	0.0	4.277	0.0	True	True	deprecated	False	0.000	0.0	0.465	0.488	Download	Download
4	predict	1000000	1000	100	0.000	0.0	4.241	0.0	True	True	deprecated	False	0.001	0.0	0.356	0.370	Download	Download
5	predict	1000000	10	100	0.000	0.0	0.100	0.0	True	True	deprecated	False	0.000	0.0	0.377	0.410	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	0.205	0.003	0.390	0.0	0.214	0.004	0.958	0.958	Download	Download
3	fit	1000000	1000000	100	1.256	0.024	0.637	0.0	0.291	0.014	4.308	4.314	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	0.011	0.0	7.112	0.0	0.019	0.0	0.599	0.599	Download	Download
2	predict	1000	10	10000	0.000	0.0	4.539	0.0	0.000	0.0	0.443	0.466	Download	Download
4	predict	1000000	1000	100	0.000	0.0	4.216	0.0	0.000	0.0	0.451	0.483	Download	Download
5	predict	1000000	10	100	0.000	0.0	0.099	0.0	0.000	0.0	0.366	0.393	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	4.288	0.238	0.0	0.004	1000	0.5	12.0	warn	warn	barnes_hut	euclidean	0.0	2	300	30.0	legacy	0	4.498	0.375	0.953	0.957	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	0.503	0.006	0.159	0.0	0.035	0.003	14.421	14.484	Download	Download
1	fit	10000	10000	1000	0.438	0.004	0.183	0.0	0.237	0.009	1.848	1.849	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