Aphrost
diff --git a/‎notebooks/T7 - 1 - Trees - Árboles de Decisión-Colab.ipynb‎
Lines changed: 14 additions & 12 deletions b/‎notebooks/T7 - 1 - Trees - Árboles de Decisión-Colab.ipynb‎
Lines changed: 14 additions & 12 deletions
@@ -142,8 +142,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "data[\"is_train\"] = (data[\"is_train\"].astype(np.float32) - data[\"is_train\"].astype(np.float32)).astype(np.bool)\n",
-    "plt.hist(data.is_train)"
+    "plt.hist(data[\"is_train\"].astype(np.int))"
    ]
   },
   {
@@ -216,7 +215,7 @@
    },
    "outputs": [],
    "source": [
-    "with open(\"/content/python-ml-course/resources/iris_dtree.dot\", \"w\") as dotfile:\n",
+    "with open(\"/content/python-ml-course/notebooks/resources/iris_dtree.dot\", \"w\") as dotfile:\n",
     "    export_graphviz(tree, out_file=dotfile, feature_names=predictors)\n",
     "    dotfile.close()"
    ]
@@ -239,7 +238,7 @@
    },
    "outputs": [],
    "source": [
-    "file = open(\"/content/python-ml-course/resources/iris_dtree.dot\", \"r\")\n",
+    "file = open(\"/content/python-ml-course/notebooks/resources/iris_dtree.dot\", \"r\")\n",
     "text = file.read()\n",
     "text"
    ]
@@ -286,7 +285,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from sklearn.cross_validation import KFold"
+    "from sklearn.model_selection import KFold"
    ]
   },
   {
@@ -295,16 +294,18 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "cv = KFold(n = X.shape[0], n_folds=10, shuffle=True, random_state=1)"
+    "cv = KFold(n_splits=10, shuffle=True, random_state=1)\n",
+    "cv.get_n_splits(X)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
-    "from sklearn.cross_validation import cross_val_score"
+    "from sklearn.model_selection import cross_val_score\n",
+    "from sklearn.metrics import accuracy_score, make_scorer"
    ]
   },
   {
@@ -313,7 +314,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "scores = cross_val_score(tree, X, Y, scoring=\"accuracy\", cv = cv, n_jobs=1)\n",
+    "scores = cross_val_score(tree, X, Y, scoring=make_scorer(accuracy_score), cv = cv, n_jobs=1)\n",
     "scores"
    ]
   },
@@ -336,8 +337,9 @@
     "for i in range(1,11):\n",
     "    tree = DecisionTreeClassifier(criterion=\"entropy\", max_depth=i, min_samples_split=20, random_state=99)\n",
     "    tree.fit(X,Y)\n",
-    "    cv = KFold(n = X.shape[0], n_folds=10, shuffle=True, random_state=1)\n",
-    "    scores = cross_val_score(tree, X, Y, scoring=\"accuracy\", cv = cv, n_jobs=1)\n",
+    "    cv = KFold(n_splits=10, shuffle=True, random_state=1)\n",
+    "    cv.get_n_splits(X)\n",
+    "    scores = cross_val_score(tree, X, Y, scoring=\"accuracy\", cv = cv, n_jobs=-1)\n",
     "    score = np.mean(scores)\n",
     "    print(\"Score para i = \",i,\" es de \", score)\n",
     "    print(\"   \",tree.feature_importances_)"
@@ -374,7 +376,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "forest = RandomForestClassifier(n_jobs=2, oob_score=True, n_estimators=100)\n",
+    "forest = RandomForestClassifier(n_jobs=-1, oob_score=True, n_estimators=100)\n",
     "forest.fit(X,Y)"
    ]
   },