Pushing the docs to dev/ for branch: master, commit 61ce18fcb748ec76082ec1a83dc026d64f842a51

Author: sklearn-ci

dev/_downloads/3409d9766d352cc9f9b169d4a799a87a/auto_examples_python.zip

@@ -94,6 +94,7 @@
 
 plt.xticks(())
 plt.yticks(())
-plt.title('Selected GMM: full model, 2 components')
+plt.title(f'Selected GMM: {best_gmm.covariance_type} model, '
+          f'{best_gmm.n_components} components')
 plt.subplots_adjust(hspace=.35, bottom=.02)
 plt.show()

dev/_downloads/61adcadacdb5cfe445011b0b0d065d44/plot_gmm_selection.py

@@ -26,7 +26,7 @@
       },
       "outputs": [],
       "source": [
-        "import numpy as np\nimport itertools\n\nfrom scipy import linalg\nimport matplotlib.pyplot as plt\nimport matplotlib as mpl\n\nfrom sklearn import mixture\n\nprint(__doc__)\n\n# Number of samples per component\nn_samples = 500\n\n# Generate random sample, two components\nnp.random.seed(0)\nC = np.array([[0., -0.1], [1.7, .4]])\nX = np.r_[np.dot(np.random.randn(n_samples, 2), C),\n          .7 * np.random.randn(n_samples, 2) + np.array([-6, 3])]\n\nlowest_bic = np.infty\nbic = []\nn_components_range = range(1, 7)\ncv_types = ['spherical', 'tied', 'diag', 'full']\nfor cv_type in cv_types:\n    for n_components in n_components_range:\n        # Fit a Gaussian mixture with EM\n        gmm = mixture.GaussianMixture(n_components=n_components,\n                                      covariance_type=cv_type)\n        gmm.fit(X)\n        bic.append(gmm.bic(X))\n        if bic[-1] < lowest_bic:\n            lowest_bic = bic[-1]\n            best_gmm = gmm\n\nbic = np.array(bic)\ncolor_iter = itertools.cycle(['navy', 'turquoise', 'cornflowerblue',\n                              'darkorange'])\nclf = best_gmm\nbars = []\n\n# Plot the BIC scores\nplt.figure(figsize=(8, 6))\nspl = plt.subplot(2, 1, 1)\nfor i, (cv_type, color) in enumerate(zip(cv_types, color_iter)):\n    xpos = np.array(n_components_range) + .2 * (i - 2)\n    bars.append(plt.bar(xpos, bic[i * len(n_components_range):\n                                  (i + 1) * len(n_components_range)],\n                        width=.2, color=color))\nplt.xticks(n_components_range)\nplt.ylim([bic.min() * 1.01 - .01 * bic.max(), bic.max()])\nplt.title('BIC score per model')\nxpos = np.mod(bic.argmin(), len(n_components_range)) + .65 +\\\n    .2 * np.floor(bic.argmin() / len(n_components_range))\nplt.text(xpos, bic.min() * 0.97 + .03 * bic.max(), '*', fontsize=14)\nspl.set_xlabel('Number of components')\nspl.legend([b[0] for b in bars], cv_types)\n\n# Plot the winner\nsplot = plt.subplot(2, 1, 2)\nY_ = clf.predict(X)\nfor i, (mean, cov, color) in enumerate(zip(clf.means_, clf.covariances_,\n                                           color_iter)):\n    v, w = linalg.eigh(cov)\n    if not np.any(Y_ == i):\n        continue\n    plt.scatter(X[Y_ == i, 0], X[Y_ == i, 1], .8, color=color)\n\n    # Plot an ellipse to show the Gaussian component\n    angle = np.arctan2(w[0][1], w[0][0])\n    angle = 180. * angle / np.pi  # convert to degrees\n    v = 2. * np.sqrt(2.) * np.sqrt(v)\n    ell = mpl.patches.Ellipse(mean, v[0], v[1], 180. + angle, color=color)\n    ell.set_clip_box(splot.bbox)\n    ell.set_alpha(.5)\n    splot.add_artist(ell)\n\nplt.xticks(())\nplt.yticks(())\nplt.title('Selected GMM: full model, 2 components')\nplt.subplots_adjust(hspace=.35, bottom=.02)\nplt.show()"
+        "import numpy as np\nimport itertools\n\nfrom scipy import linalg\nimport matplotlib.pyplot as plt\nimport matplotlib as mpl\n\nfrom sklearn import mixture\n\nprint(__doc__)\n\n# Number of samples per component\nn_samples = 500\n\n# Generate random sample, two components\nnp.random.seed(0)\nC = np.array([[0., -0.1], [1.7, .4]])\nX = np.r_[np.dot(np.random.randn(n_samples, 2), C),\n          .7 * np.random.randn(n_samples, 2) + np.array([-6, 3])]\n\nlowest_bic = np.infty\nbic = []\nn_components_range = range(1, 7)\ncv_types = ['spherical', 'tied', 'diag', 'full']\nfor cv_type in cv_types:\n    for n_components in n_components_range:\n        # Fit a Gaussian mixture with EM\n        gmm = mixture.GaussianMixture(n_components=n_components,\n                                      covariance_type=cv_type)\n        gmm.fit(X)\n        bic.append(gmm.bic(X))\n        if bic[-1] < lowest_bic:\n            lowest_bic = bic[-1]\n            best_gmm = gmm\n\nbic = np.array(bic)\ncolor_iter = itertools.cycle(['navy', 'turquoise', 'cornflowerblue',\n                              'darkorange'])\nclf = best_gmm\nbars = []\n\n# Plot the BIC scores\nplt.figure(figsize=(8, 6))\nspl = plt.subplot(2, 1, 1)\nfor i, (cv_type, color) in enumerate(zip(cv_types, color_iter)):\n    xpos = np.array(n_components_range) + .2 * (i - 2)\n    bars.append(plt.bar(xpos, bic[i * len(n_components_range):\n                                  (i + 1) * len(n_components_range)],\n                        width=.2, color=color))\nplt.xticks(n_components_range)\nplt.ylim([bic.min() * 1.01 - .01 * bic.max(), bic.max()])\nplt.title('BIC score per model')\nxpos = np.mod(bic.argmin(), len(n_components_range)) + .65 +\\\n    .2 * np.floor(bic.argmin() / len(n_components_range))\nplt.text(xpos, bic.min() * 0.97 + .03 * bic.max(), '*', fontsize=14)\nspl.set_xlabel('Number of components')\nspl.legend([b[0] for b in bars], cv_types)\n\n# Plot the winner\nsplot = plt.subplot(2, 1, 2)\nY_ = clf.predict(X)\nfor i, (mean, cov, color) in enumerate(zip(clf.means_, clf.covariances_,\n                                           color_iter)):\n    v, w = linalg.eigh(cov)\n    if not np.any(Y_ == i):\n        continue\n    plt.scatter(X[Y_ == i, 0], X[Y_ == i, 1], .8, color=color)\n\n    # Plot an ellipse to show the Gaussian component\n    angle = np.arctan2(w[0][1], w[0][0])\n    angle = 180. * angle / np.pi  # convert to degrees\n    v = 2. * np.sqrt(2.) * np.sqrt(v)\n    ell = mpl.patches.Ellipse(mean, v[0], v[1], 180. + angle, color=color)\n    ell.set_clip_box(splot.bbox)\n    ell.set_alpha(.5)\n    splot.add_artist(ell)\n\nplt.xticks(())\nplt.yticks(())\nplt.title(f'Selected GMM: {best_gmm.covariance_type} model, '\n          f'{best_gmm.n_components} components')\nplt.subplots_adjust(hspace=.35, bottom=.02)\nplt.show()"
       ]
     }
   ],