Eddie Bergman: Update example to use predefined_split properly (#1340)

Author: Github Actions

development/_downloads/bb7d59d9ddc2ff29f0d6eb99747a3347/example_resampling.ipynb

@@ -141,7 +141,7 @@
       },
       "outputs": [],
       "source": [
-        "resampling_strategy = sklearn.model_selection.PredefinedSplit(\n    test_fold=np.where(X_train[:, 0] < np.mean(X_train[:, 0]))[0]\n)\n\nautoml = autosklearn.classification.AutoSklearnClassifier(\n    time_left_for_this_task=120,\n    per_run_time_limit=30,\n    tmp_folder='/tmp/autosklearn_resampling_example_tmp',\n    disable_evaluator_output=False,\n    resampling_strategy=resampling_strategy,\n)\nautoml.fit(X_train, y_train, dataset_name='breast_cancer')"
+        "selected_indices = (X_train[:, 0] < np.mean(X_train[:, 0])).astype(int)\nresampling_strategy = sklearn.model_selection.PredefinedSplit(\n    test_fold=selected_indices\n)\n\nautoml = autosklearn.classification.AutoSklearnClassifier(\n    time_left_for_this_task=120,\n    per_run_time_limit=30,\n    tmp_folder='/tmp/autosklearn_resampling_example_tmp',\n    disable_evaluator_output=False,\n    resampling_strategy=resampling_strategy,\n)\nautoml.fit(X_train, y_train, dataset_name='breast_cancer')\n\nprint(automl.sprint_statistics())"
       ]
     },
     {

development/_downloads/bc82bea3a5dd7bdba60b65220891d9e5/examples_python.zip

@@ -98,8 +98,9 @@
 # data by the first feature. In practice, one would use a splitting according
 # to the use case at hand.
 
+selected_indices = (X_train[:, 0] < np.mean(X_train[:, 0])).astype(int)
 resampling_strategy = sklearn.model_selection.PredefinedSplit(
-    test_fold=np.where(X_train[:, 0] < np.mean(X_train[:, 0]))[0]
+    test_fold=selected_indices
 )
 
 automl = autosklearn.classification.AutoSklearnClassifier(
@@ -111,6 +112,8 @@
 )
 automl.fit(X_train, y_train, dataset_name='breast_cancer')
 
+print(automl.sprint_statistics())
+
 ############################################################################
 # For custom resampling strategies (i.e. resampling strategies that are not
 # defined as strings by Auto-sklearn) it is necessary to perform a refit: