Pushing the docs to dev/ for branch: main, commit ec74b2a78a3365fb49b70c12dd4e305cb5ab6be0

Author: sklearn-ci

dev/_downloads/02192f99342d6d1323161978b6c80bfc/plot_ols_ridge.zip

@@ -22,9 +22,12 @@
 
 from sklearn.datasets import make_classification
 
+n_features = 15
+feat_names = [f"feature_{i}" for i in range(15)]
+
 X, y = make_classification(
     n_samples=500,
-    n_features=15,
+    n_features=n_features,
     n_informative=3,
     n_redundant=2,
     n_repeated=0,
@@ -71,7 +74,12 @@
 import matplotlib.pyplot as plt
 import pandas as pd
 
-cv_results = pd.DataFrame(rfecv.cv_results_)
+data = {
+    key: value
+    for key, value in rfecv.cv_results_.items()
+    if key in ["n_features", "mean_test_score", "std_test_score"]
+}
+cv_results = pd.DataFrame(data)
 plt.figure()
 plt.xlabel("Number of features selected")
 plt.ylabel("Mean test accuracy")
@@ -91,3 +99,17 @@
 # cross-validation technique. The test accuracy decreases above 5 selected
 # features, this is, keeping non-informative features leads to over-fitting and
 # is therefore detrimental for the statistical performance of the models.
+
+# %%
+import numpy as np
+
+for i in range(cv.n_splits):
+    mask = rfecv.cv_results_[f"split{i}_support"][
+        rfecv.n_features_
+    ]  # mask of features selected by the RFE
+    features_selected = np.ma.compressed(np.ma.masked_array(feat_names, mask=1 - mask))
+    print(f"Features selected in fold {i}: {features_selected}")
+# %%
+# In the five folds, the selected features are consistant. This is good news,
+# it means that the selection is stable accross folds, and it confirms that
+# these features are the most informative ones.

dev/_downloads/02fe21a1f5d14bd1ebbe34aa905d9bf6/plot_multilabel.zip

@@ -33,7 +33,7 @@
       },
       "outputs": [],
       "source": [
-        "from sklearn.datasets import make_classification\n\nX, y = make_classification(\n    n_samples=500,\n    n_features=15,\n    n_informative=3,\n    n_redundant=2,\n    n_repeated=0,\n    n_classes=8,\n    n_clusters_per_class=1,\n    class_sep=0.8,\n    random_state=0,\n)"
+        "from sklearn.datasets import make_classification\n\nn_features = 15\nfeat_names = [f\"feature_{i}\" for i in range(15)]\n\nX, y = make_classification(\n    n_samples=500,\n    n_features=n_features,\n    n_informative=3,\n    n_redundant=2,\n    n_repeated=0,\n    n_classes=8,\n    n_clusters_per_class=1,\n    class_sep=0.8,\n    random_state=0,\n)"
       ]
     },
     {
@@ -69,7 +69,7 @@
       },
       "outputs": [],
       "source": [
-        "import matplotlib.pyplot as plt\nimport pandas as pd\n\ncv_results = pd.DataFrame(rfecv.cv_results_)\nplt.figure()\nplt.xlabel(\"Number of features selected\")\nplt.ylabel(\"Mean test accuracy\")\nplt.errorbar(\n    x=cv_results[\"n_features\"],\n    y=cv_results[\"mean_test_score\"],\n    yerr=cv_results[\"std_test_score\"],\n)\nplt.title(\"Recursive Feature Elimination \\nwith correlated features\")\nplt.show()"
+        "import matplotlib.pyplot as plt\nimport pandas as pd\n\ndata = {\n    key: value\n    for key, value in rfecv.cv_results_.items()\n    if key in [\"n_features\", \"mean_test_score\", \"std_test_score\"]\n}\ncv_results = pd.DataFrame(data)\nplt.figure()\nplt.xlabel(\"Number of features selected\")\nplt.ylabel(\"Mean test accuracy\")\nplt.errorbar(\n    x=cv_results[\"n_features\"],\n    y=cv_results[\"mean_test_score\"],\n    yerr=cv_results[\"std_test_score\"],\n)\nplt.title(\"Recursive Feature Elimination \\nwith correlated features\")\nplt.show()"
       ]
     },
     {
@@ -78,6 +78,24 @@
       "source": [
         "From the plot above one can further notice a plateau of equivalent scores\n(similar mean value and overlapping errorbars) for 3 to 5 selected features.\nThis is the result of introducing correlated features. Indeed, the optimal\nmodel selected by the RFE can lie within this range, depending on the\ncross-validation technique. The test accuracy decreases above 5 selected\nfeatures, this is, keeping non-informative features leads to over-fitting and\nis therefore detrimental for the statistical performance of the models.\n\n"
       ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "import numpy as np\n\nfor i in range(cv.n_splits):\n    mask = rfecv.cv_results_[f\"split{i}_support\"][\n        rfecv.n_features_\n    ]  # mask of features selected by the RFE\n    features_selected = np.ma.compressed(np.ma.masked_array(feat_names, mask=1 - mask))\n    print(f\"Features selected in fold {i}: {features_selected}\")"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "In the five folds, the selected features are consistant. This is good news,\nit means that the selection is stable accross folds, and it confirms that\nthese features are the most informative ones.\n\n"
+      ]
     }
   ],
   "metadata": {