Update notebooks 1 to 8 to latest library versions (in particular Scikit-Learn 0.20)

05_support_vector_machines.ipynb

@@ -774,6 +774,13 @@
     "y = (0.2 + 0.1 * X + 0.5 * X**2 + np.random.randn(m, 1)/10).ravel()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**Warning**: the default value of `gamma` will change from `'auto'` to `'scale'` in version 0.22 to better account for unscaled features. To preserve the same results as in the book, we explicitly set it to `'auto'`, but you should probably just use the default in your own code."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 27,
@@ -782,7 +789,7 @@
    "source": [
     "from sklearn.svm import SVR\n",
     "\n",
-    "svm_poly_reg = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1)\n",
+    "svm_poly_reg = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1, gamma=\"auto\")\n",
     "svm_poly_reg.fit(X, y)"
    ]
   },
@@ -794,8 +801,8 @@
    "source": [
     "from sklearn.svm import SVR\n",
     "\n",
-    "svm_poly_reg1 = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1)\n",
-    "svm_poly_reg2 = SVR(kernel=\"poly\", degree=2, C=0.01, epsilon=0.1)\n",
+    "svm_poly_reg1 = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1, gamma=\"auto\")\n",
+    "svm_poly_reg2 = SVR(kernel=\"poly\", degree=2, C=0.01, epsilon=0.1, gamma=\"auto\")\n",
     "svm_poly_reg1.fit(X, y)\n",
     "svm_poly_reg2.fit(X, y)"
    ]
@@ -876,7 +883,7 @@
     "ax1 = fig.add_subplot(111, projection='3d')\n",
     "plot_3D_decision_function(ax1, w=svm_clf2.coef_[0], b=svm_clf2.intercept_[0])\n",
     "\n",
-    "save_fig(\"iris_3D_plot\")\n",
+    "#save_fig(\"iris_3D_plot\")\n",
     "plt.show()"
    ]
   },
@@ -1165,7 +1172,7 @@
    "source": [
     "from sklearn.linear_model import SGDClassifier\n",
     "\n",
-    "sgd_clf = SGDClassifier(loss=\"hinge\", alpha = 0.017, max_iter = 50, random_state=42)\n",
+    "sgd_clf = SGDClassifier(loss=\"hinge\", alpha = 0.017, max_iter = 50, tol=-np.infty, random_state=42)\n",
     "sgd_clf.fit(X, y.ravel())\n",
     "\n",
     "m = len(X)\n",
@@ -1265,7 +1272,7 @@
     "lin_clf = LinearSVC(loss=\"hinge\", C=C, random_state=42)\n",
     "svm_clf = SVC(kernel=\"linear\", C=C)\n",
     "sgd_clf = SGDClassifier(loss=\"hinge\", learning_rate=\"constant\", eta0=0.001, alpha=alpha,\n",
-    "                        max_iter=100000, random_state=42)\n",
+    "                        max_iter=100000, tol=-np.infty, random_state=42)\n",
     "\n",
     "scaler = StandardScaler()\n",
     "X_scaled = scaler.fit_transform(X)\n",
@@ -1354,9 +1361,13 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from sklearn.datasets import fetch_mldata\n",
+    "try:\n",
+    "    from sklearn.datasets import fetch_openml\n",
+    "    mnist = fetch_openml('mnist_784', version=1, cache=True)\n",
+    "except ImportError:\n",
+    "    from sklearn.datasets import fetch_mldata\n",
+    "    mnist = fetch_mldata('MNIST original')\n",
     "\n",
-    "mnist = fetch_mldata(\"MNIST original\")\n",
     "X = mnist[\"data\"]\n",
     "y = mnist[\"target\"]\n",
     "\n",
@@ -1425,7 +1436,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Wow, 82% accuracy on MNIST is a really bad performance. This linear model is certainly too simple for MNIST, but perhaps we just needed to scale the data first:"
+    "Wow, 86% accuracy on MNIST is a really bad performance. This linear model is certainly too simple for MNIST, but perhaps we just needed to scale the data first:"
    ]
   },
   {
@@ -1474,7 +1485,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "svm_clf = SVC(decision_function_shape=\"ovr\")\n",
+    "svm_clf = SVC(decision_function_shape=\"ovr\", gamma=\"auto\")\n",
     "svm_clf.fit(X_train_scaled[:10000], y_train[:10000])"
    ]
   },
@@ -1505,7 +1516,7 @@
     "from scipy.stats import reciprocal, uniform\n",
     "\n",
     "param_distributions = {\"gamma\": reciprocal(0.001, 0.1), \"C\": uniform(1, 10)}\n",
-    "rnd_search_cv = RandomizedSearchCV(svm_clf, param_distributions, n_iter=10, verbose=2)\n",
+    "rnd_search_cv = RandomizedSearchCV(svm_clf, param_distributions, n_iter=10, verbose=2, cv=3)\n",
     "rnd_search_cv.fit(X_train_scaled[:1000], y_train[:1000])"
    ]
   },
@@ -1536,7 +1547,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 59,
+   "execution_count": 60,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1545,7 +1556,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 60,
+   "execution_count": 61,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1562,7 +1573,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 61,
+   "execution_count": 62,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1600,7 +1611,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 62,
+   "execution_count": 63,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1620,7 +1631,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 63,
+   "execution_count": 64,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1638,7 +1649,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 64,
+   "execution_count": 65,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1658,7 +1669,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 65,
+   "execution_count": 66,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1677,7 +1688,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 66,
+   "execution_count": 67,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1697,7 +1708,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 67,
+   "execution_count": 68,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1713,7 +1724,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 68,
+   "execution_count": 69,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1722,13 +1733,13 @@
     "from scipy.stats import reciprocal, uniform\n",
     "\n",
     "param_distributions = {\"gamma\": reciprocal(0.001, 0.1), \"C\": uniform(1, 10)}\n",
-    "rnd_search_cv = RandomizedSearchCV(SVR(), param_distributions, n_iter=10, verbose=2, random_state=42)\n",
+    "rnd_search_cv = RandomizedSearchCV(SVR(), param_distributions, n_iter=10, verbose=2, cv=3, random_state=42)\n",
     "rnd_search_cv.fit(X_train_scaled, y_train)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 69,
+   "execution_count": 70,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1744,7 +1755,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 70,
+   "execution_count": 71,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1762,7 +1773,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 71,
+   "execution_count": 72,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1771,6 +1782,26 @@
     "np.sqrt(mse)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 73,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cmap = matplotlib.cm.get_cmap(\"jet\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 74,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.datasets import fetch_openml\n",
+    "mnist = fetch_openml(\"mnist_784\", version=1)\n",
+    "print(mnist.data.shape)"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -1795,7 +1826,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.2"
+   "version": "3.6.6"
   },
   "nav_menu": {},
   "toc": {