|
1 | 1 | { |
2 | 2 | "INCLUDE": ["../common/sklearn.json"], |
3 | 3 | "PARAMETERS_SETS": { |
4 | | - "common parameters": { |
5 | | - "data": { |
6 | | - "split_kwargs": { |
7 | | - "train_size": 8000, |
8 | | - "test_size": 2000, |
9 | | - "shuffle": true, |
| 4 | + "throughput settings": { |
| 5 | + "bench": { |
| 6 | + "num_instances": 7, |
| 7 | + "cores_per_instance": 8, |
| 8 | + "measurement_duration": 30 |
| 9 | + } |
| 10 | + }, |
| 11 | + "linear regression": { |
| 12 | + "algorithm": { |
| 13 | + "estimator": "LinearRegression", |
| 14 | + "estimator_params": { |
| 15 | + "fit_intercept": true, |
| 16 | + "copy_X": true |
| 17 | + } |
| 18 | + } |
| 19 | + }, |
| 20 | + "random forest regressor": { |
| 21 | + "algorithm": { |
| 22 | + "estimator": "RandomForestRegressor", |
| 23 | + "estimator_params": { |
| 24 | + "n_estimators": 100, |
| 25 | + "max_depth": 12, |
10 | 26 | "random_state": 42 |
11 | 27 | } |
| 28 | + } |
| 29 | + }, |
| 30 | + "knn classifier kdtree": { |
| 31 | + "algorithm": { |
| 32 | + "estimator": "KNeighborsClassifier", |
| 33 | + "estimator_params": { |
| 34 | + "n_neighbors": 10, |
| 35 | + "weights": "uniform", |
| 36 | + "algorithm": "kd_tree", |
| 37 | + "metric": "minkowski", |
| 38 | + "p": 2 |
| 39 | + } |
12 | 40 | }, |
13 | | - "algorithm": { "device": "default" } |
| 41 | + "data": { |
| 42 | + "preprocessing_kwargs": { "normalize": "standard" } |
| 43 | + } |
14 | 44 | }, |
15 | | - "throughput settings": { |
16 | | - "bench": { |
17 | | - "num_instances": 4, |
18 | | - "cores_per_instance": 4, |
19 | | - "measurement_duration": 30 |
| 45 | + "kmeans": { |
| 46 | + "algorithm": { |
| 47 | + "estimator": "KMeans", |
| 48 | + "estimator_params": { |
| 49 | + "n_clusters": 10, |
| 50 | + "n_init": 1, |
| 51 | + "max_iter": 100, |
| 52 | + "tol": 1e-4, |
| 53 | + "random_state": 42, |
| 54 | + "algorithm": "lloyd" |
| 55 | + }, |
| 56 | + "estimator_methods": { |
| 57 | + "inference": "predict" |
| 58 | + } |
| 59 | + }, |
| 60 | + "data": { |
| 61 | + "preprocessing_kwargs": { "normalize": "standard" } |
20 | 62 | } |
21 | 63 | }, |
22 | | - "datasets": { |
| 64 | + "linear regression datasets": { |
23 | 65 | "data": [ |
24 | 66 | { |
25 | | - "source": "make_classification", |
| 67 | + "source": "make_regression", |
26 | 68 | "generation_kwargs": { |
27 | | - "n_classes": 2, |
28 | | - "n_samples": 10000, |
29 | | - "n_features": 64, |
30 | | - "n_informative": 32 |
| 69 | + "n_samples": 500000, |
| 70 | + "n_features": 50, |
| 71 | + "n_informative": 10, |
| 72 | + "noise": 20.0, |
| 73 | + "random_state": 42 |
31 | 74 | } |
| 75 | + }, |
| 76 | + { |
| 77 | + "source": "make_regression", |
| 78 | + "generation_kwargs": { |
| 79 | + "n_samples": 100000, |
| 80 | + "n_features": 2000, |
| 81 | + "n_informative": 10, |
| 82 | + "noise": 60.0, |
| 83 | + "random_state": 42 |
| 84 | + } |
| 85 | + }, |
| 86 | + { |
| 87 | + "dataset": "year_prediction_msd", |
| 88 | + "split_kwargs": { "train_size": 0.8, "test_size": 0.2 } |
| 89 | + }, |
| 90 | + { |
| 91 | + "dataset": "hepmass", |
| 92 | + "split_kwargs": { "train_size": 0.8, "test_size": 0.2 } |
32 | 93 | } |
33 | 94 | ] |
34 | 95 | }, |
35 | | - "algorithms": [ |
36 | | - { |
37 | | - "algorithm": { |
38 | | - "estimator": "RandomForestClassifier", |
39 | | - "estimator_params": { "n_estimators": 50 } |
| 96 | + "random forest datasets": { |
| 97 | + "data": [ |
| 98 | + { |
| 99 | + "source": "make_regression", |
| 100 | + "generation_kwargs": { |
| 101 | + "n_samples": 100000, |
| 102 | + "n_features": 20, |
| 103 | + "n_informative": 10, |
| 104 | + "noise": 10.0, |
| 105 | + "random_state": 42 |
| 106 | + } |
| 107 | + }, |
| 108 | + { |
| 109 | + "source": "make_regression", |
| 110 | + "generation_kwargs": { |
| 111 | + "n_samples": 50000, |
| 112 | + "n_features": 200, |
| 113 | + "n_informative": 20, |
| 114 | + "noise": 30.0, |
| 115 | + "random_state": 42 |
| 116 | + } |
| 117 | + }, |
| 118 | + { |
| 119 | + "dataset": "year_prediction_msd", |
| 120 | + "split_kwargs": { "train_size": 0.8, "test_size": 0.2 } |
| 121 | + }, |
| 122 | + { |
| 123 | + "dataset": "medical_charges_nominal", |
| 124 | + "split_kwargs": { "ignore": true } |
40 | 125 | } |
41 | | - }, |
42 | | - { |
43 | | - "algorithm": { |
44 | | - "estimator": "KMeans", |
45 | | - "estimator_params": { |
46 | | - "n_clusters": 10, |
47 | | - "init": "random", |
48 | | - "algorithm": "lloyd", |
49 | | - "max_iter": 100 |
| 126 | + ] |
| 127 | + }, |
| 128 | + "knn datasets": { |
| 129 | + "data": [ |
| 130 | + { |
| 131 | + "source": "make_classification", |
| 132 | + "generation_kwargs": { |
| 133 | + "n_classes": 5, |
| 134 | + "n_samples": 200000, |
| 135 | + "n_features": 10, |
| 136 | + "n_informative": 8, |
| 137 | + "n_redundant": 2, |
| 138 | + "n_repeated": 0, |
| 139 | + "random_state": 42 |
| 140 | + } |
| 141 | + }, |
| 142 | + { |
| 143 | + "source": "make_classification", |
| 144 | + "generation_kwargs": { |
| 145 | + "n_classes": 5, |
| 146 | + "n_samples": 50000, |
| 147 | + "n_features": 50, |
| 148 | + "n_informative": 20, |
| 149 | + "n_redundant": 10, |
| 150 | + "n_repeated": 0, |
| 151 | + "random_state": 42 |
50 | 152 | } |
| 153 | + }, |
| 154 | + { |
| 155 | + "dataset": "letters", |
| 156 | + "split_kwargs": { "ignore": true } |
| 157 | + }, |
| 158 | + { |
| 159 | + "dataset": "codrnanorm", |
| 160 | + "split_kwargs": { "ignore": true } |
51 | 161 | } |
52 | | - }, |
53 | | - { |
54 | | - "algorithm": { |
55 | | - "estimator": "LinearRegression" |
| 162 | + ] |
| 163 | + }, |
| 164 | + "kmeans datasets": { |
| 165 | + "data": [ |
| 166 | + { |
| 167 | + "source": "make_blobs", |
| 168 | + "generation_kwargs": { |
| 169 | + "centers": 10, |
| 170 | + "cluster_std": 4.0, |
| 171 | + "n_samples": 500000, |
| 172 | + "n_features": 20, |
| 173 | + "random_state": 42 |
| 174 | + } |
| 175 | + }, |
| 176 | + { |
| 177 | + "source": "make_blobs", |
| 178 | + "generation_kwargs": { |
| 179 | + "centers": 10, |
| 180 | + "cluster_std": 8.0, |
| 181 | + "n_samples": 100000, |
| 182 | + "n_features": 200, |
| 183 | + "random_state": 42 |
| 184 | + } |
| 185 | + }, |
| 186 | + { |
| 187 | + "dataset": "creditcard", |
| 188 | + "split_kwargs": { "ignore": true } |
| 189 | + }, |
| 190 | + { |
| 191 | + "dataset": "hepmass", |
| 192 | + "split_kwargs": { "train_size": 0.8, "test_size": 0.2 } |
56 | 193 | } |
57 | | - } |
58 | | - ] |
| 194 | + ] |
| 195 | + } |
59 | 196 | }, |
60 | 197 | "TEMPLATES": { |
61 | | - "throughput_test": { |
| 198 | + "linear_regression_throughput": { |
| 199 | + "SETS": [ |
| 200 | + "sklearn-ex[cpu] implementations", |
| 201 | + "throughput settings", |
| 202 | + "linear regression", |
| 203 | + "linear regression datasets" |
| 204 | + ] |
| 205 | + }, |
| 206 | + "random_forest_throughput": { |
| 207 | + "SETS": [ |
| 208 | + "sklearn-ex[cpu] implementations", |
| 209 | + "throughput settings", |
| 210 | + "random forest regressor", |
| 211 | + "random forest datasets" |
| 212 | + ] |
| 213 | + }, |
| 214 | + "knn_throughput": { |
| 215 | + "SETS": [ |
| 216 | + "sklearn-ex[cpu] implementations", |
| 217 | + "throughput settings", |
| 218 | + "knn classifier kdtree", |
| 219 | + "knn datasets" |
| 220 | + ] |
| 221 | + }, |
| 222 | + "kmeans_throughput": { |
62 | 223 | "SETS": [ |
63 | 224 | "sklearn-ex[cpu] implementations", |
64 | | - "common parameters", |
65 | 225 | "throughput settings", |
66 | | - "datasets", |
67 | | - "algorithms" |
| 226 | + "kmeans", |
| 227 | + "kmeans datasets" |
68 | 228 | ] |
69 | 229 | } |
70 | 230 | } |
|
0 commit comments