Pushing the docs to dev/ for branch: master, commit f65cdf2d5e5f2b753393190598f9c9f63f34dbe1

Author: sklearn-ci

dev/_downloads/3409d9766d352cc9f9b169d4a799a87a/auto_examples_python.zip

@@ -14,7 +14,7 @@
 coefficients are exactly 0. When regularization gets progressively looser,
 coefficients can get non-zero values one after the other.
 
-Here we choose the SAGA solver because it can efficiently optimize for the
+Here we choose the liblinear solver because it can efficiently optimize for the
 Logistic Regression loss with a non-smooth, sparsity inducing l1 penalty.
 
 Also note that we set a low value for the tolerance to make sure that the model
@@ -55,9 +55,10 @@
 
 print("Computing regularization path ...")
 start = time()
-clf = linear_model.LogisticRegression(penalty='l1', solver='saga',
+clf = linear_model.LogisticRegression(penalty='l1', solver='liblinear',
                                       tol=1e-6, max_iter=int(1e6),
-                                      warm_start=True)
+                                      warm_start=True,
+                                      intercept_scaling=10000.)
 coefs_ = []
 for c in cs:
     clf.set_params(C=c)

dev/_downloads/998b4a46568dd85981260617d1ad84f5/plot_logistic_path.py

@@ -15,7 +15,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "\n# Regularization path of L1- Logistic Regression\n\n\n\nTrain l1-penalized logistic regression models on a binary classification\nproblem derived from the Iris dataset.\n\nThe models are ordered from strongest regularized to least regularized. The 4\ncoefficients of the models are collected and plotted as a \"regularization\npath\": on the left-hand side of the figure (strong regularizers), all the\ncoefficients are exactly 0. When regularization gets progressively looser,\ncoefficients can get non-zero values one after the other.\n\nHere we choose the SAGA solver because it can efficiently optimize for the\nLogistic Regression loss with a non-smooth, sparsity inducing l1 penalty.\n\nAlso note that we set a low value for the tolerance to make sure that the model\nhas converged before collecting the coefficients.\n\nWe also use warm_start=True which means that the coefficients of the models are\nreused to initialize the next model fit to speed-up the computation of the\nfull-path.\n"
+        "\n# Regularization path of L1- Logistic Regression\n\n\n\nTrain l1-penalized logistic regression models on a binary classification\nproblem derived from the Iris dataset.\n\nThe models are ordered from strongest regularized to least regularized. The 4\ncoefficients of the models are collected and plotted as a \"regularization\npath\": on the left-hand side of the figure (strong regularizers), all the\ncoefficients are exactly 0. When regularization gets progressively looser,\ncoefficients can get non-zero values one after the other.\n\nHere we choose the liblinear solver because it can efficiently optimize for the\nLogistic Regression loss with a non-smooth, sparsity inducing l1 penalty.\n\nAlso note that we set a low value for the tolerance to make sure that the model\nhas converged before collecting the coefficients.\n\nWe also use warm_start=True which means that the coefficients of the models are\nreused to initialize the next model fit to speed-up the computation of the\nfull-path.\n"
       ]
     },
     {
@@ -26,7 +26,7 @@
       },
       "outputs": [],
       "source": [
-        "print(__doc__)\n\n# Author: Alexandre Gramfort <[email protected]>\n# License: BSD 3 clause\n\nfrom time import time\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nfrom sklearn import linear_model\nfrom sklearn import datasets\nfrom sklearn.svm import l1_min_c\n\niris = datasets.load_iris()\nX = iris.data\ny = iris.target\n\nX = X[y != 2]\ny = y[y != 2]\n\nX /= X.max()  # Normalize X to speed-up convergence\n\n# #############################################################################\n# Demo path functions\n\ncs = l1_min_c(X, y, loss='log') * np.logspace(0, 7, 16)\n\n\nprint(\"Computing regularization path ...\")\nstart = time()\nclf = linear_model.LogisticRegression(penalty='l1', solver='saga',\n                                      tol=1e-6, max_iter=int(1e6),\n                                      warm_start=True)\ncoefs_ = []\nfor c in cs:\n    clf.set_params(C=c)\n    clf.fit(X, y)\n    coefs_.append(clf.coef_.ravel().copy())\nprint(\"This took %0.3fs\" % (time() - start))\n\ncoefs_ = np.array(coefs_)\nplt.plot(np.log10(cs), coefs_, marker='o')\nymin, ymax = plt.ylim()\nplt.xlabel('log(C)')\nplt.ylabel('Coefficients')\nplt.title('Logistic Regression Path')\nplt.axis('tight')\nplt.show()"
+        "print(__doc__)\n\n# Author: Alexandre Gramfort <[email protected]>\n# License: BSD 3 clause\n\nfrom time import time\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nfrom sklearn import linear_model\nfrom sklearn import datasets\nfrom sklearn.svm import l1_min_c\n\niris = datasets.load_iris()\nX = iris.data\ny = iris.target\n\nX = X[y != 2]\ny = y[y != 2]\n\nX /= X.max()  # Normalize X to speed-up convergence\n\n# #############################################################################\n# Demo path functions\n\ncs = l1_min_c(X, y, loss='log') * np.logspace(0, 7, 16)\n\n\nprint(\"Computing regularization path ...\")\nstart = time()\nclf = linear_model.LogisticRegression(penalty='l1', solver='liblinear',\n                                      tol=1e-6, max_iter=int(1e6),\n                                      warm_start=True,\n                                      intercept_scaling=10000.)\ncoefs_ = []\nfor c in cs:\n    clf.set_params(C=c)\n    clf.fit(X, y)\n    coefs_.append(clf.coef_.ravel().copy())\nprint(\"This took %0.3fs\" % (time() - start))\n\ncoefs_ = np.array(coefs_)\nplt.plot(np.log10(cs), coefs_, marker='o')\nymin, ymax = plt.ylim()\nplt.xlabel('log(C)')\nplt.ylabel('Coefficients')\nplt.title('Logistic Regression Path')\nplt.axis('tight')\nplt.show()"
       ]
     }
   ],