Pushing the docs to dev/ for branch: master, commit c5706e6ff03a9f3c105b2c1b7f972527e664b3c1

Author: sklearn-ci

dev/_downloads/auto_examples_jupyter.zip

@@ -51,7 +51,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "A synthetic random regression problem is generated. The targets ``y`` are\nmodified by: (i) translating all targets such that all entries are\nnon-negative and (ii) applying an exponential function to obtain non-linear\ntargets which cannot be fitted using a simple linear model.\n\nTherefore, a logarithmic and an exponential function will be used to\ntransform the targets before training a linear regression model and using it\nfor prediction.\n\n"
+        "A synthetic random regression problem is generated. The targets ``y`` are\nmodified by: (i) translating all targets such that all entries are\nnon-negative and (ii) applying an exponential function to obtain non-linear\ntargets which cannot be fitted using a simple linear model.\n\nTherefore, a logarithmic (`np.log1p`) and an exponential function\n(`np.expm1`) will be used to transform the targets before training a linear\nregression model and using it for prediction.\n\n"
       ]
     },
     {
@@ -62,7 +62,7 @@
       },
       "outputs": [],
       "source": [
-        "def log_transform(x):\n    return np.log(x + 1)\n\n\ndef exp_transform(x):\n    return np.exp(x) - 1\n\n\nX, y = make_regression(n_samples=10000, noise=100, random_state=0)\ny = np.exp((y + abs(y.min())) / 200)\ny_trans = log_transform(y)"
+        "X, y = make_regression(n_samples=10000, noise=100, random_state=0)\ny = np.exp((y + abs(y.min())) / 200)\ny_trans = np.log1p(y)"
       ]
     },
     {
@@ -98,7 +98,7 @@
       },
       "outputs": [],
       "source": [
-        "f, (ax0, ax1) = plt.subplots(1, 2, sharey=True)\n\nregr = RidgeCV()\nregr.fit(X_train, y_train)\ny_pred = regr.predict(X_test)\n\nax0.scatter(y_test, y_pred)\nax0.plot([0, 2000], [0, 2000], '--k')\nax0.set_ylabel('Target predicted')\nax0.set_xlabel('True Target')\nax0.set_title('Ridge regression \\n without target transformation')\nax0.text(100, 1750, r'$R^2$=%.2f, MAE=%.2f' % (\n    r2_score(y_test, y_pred), median_absolute_error(y_test, y_pred)))\nax0.set_xlim([0, 2000])\nax0.set_ylim([0, 2000])\n\nregr_trans = TransformedTargetRegressor(regressor=RidgeCV(),\n                                        func=log_transform,\n                                        inverse_func=exp_transform)\nregr_trans.fit(X_train, y_train)\ny_pred = regr_trans.predict(X_test)\n\nax1.scatter(y_test, y_pred)\nax1.plot([0, 2000], [0, 2000], '--k')\nax1.set_ylabel('Target predicted')\nax1.set_xlabel('True Target')\nax1.set_title('Ridge regression \\n with target transformation')\nax1.text(100, 1750, r'$R^2$=%.2f, MAE=%.2f' % (\n    r2_score(y_test, y_pred), median_absolute_error(y_test, y_pred)))\nax1.set_xlim([0, 2000])\nax1.set_ylim([0, 2000])\n\nf.suptitle(\"Synthetic data\", y=0.035)\nf.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])"
+        "f, (ax0, ax1) = plt.subplots(1, 2, sharey=True)\n\nregr = RidgeCV()\nregr.fit(X_train, y_train)\ny_pred = regr.predict(X_test)\n\nax0.scatter(y_test, y_pred)\nax0.plot([0, 2000], [0, 2000], '--k')\nax0.set_ylabel('Target predicted')\nax0.set_xlabel('True Target')\nax0.set_title('Ridge regression \\n without target transformation')\nax0.text(100, 1750, r'$R^2$=%.2f, MAE=%.2f' % (\n    r2_score(y_test, y_pred), median_absolute_error(y_test, y_pred)))\nax0.set_xlim([0, 2000])\nax0.set_ylim([0, 2000])\n\nregr_trans = TransformedTargetRegressor(regressor=RidgeCV(),\n                                        func=np.log1p,\n                                        inverse_func=np.expm1)\nregr_trans.fit(X_train, y_train)\ny_pred = regr_trans.predict(X_test)\n\nax1.scatter(y_test, y_pred)\nax1.plot([0, 2000], [0, 2000], '--k')\nax1.set_ylabel('Target predicted')\nax1.set_xlabel('True Target')\nax1.set_title('Ridge regression \\n with target transformation')\nax1.text(100, 1750, r'$R^2$=%.2f, MAE=%.2f' % (\n    r2_score(y_test, y_pred), median_absolute_error(y_test, y_pred)))\nax1.set_xlim([0, 2000])\nax1.set_ylim([0, 2000])\n\nf.suptitle(\"Synthetic data\", y=0.035)\nf.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])"
       ]
     },
     {

dev/_downloads/auto_examples_python.zip

@@ -40,22 +40,13 @@
 # non-negative and (ii) applying an exponential function to obtain non-linear
 # targets which cannot be fitted using a simple linear model.
 #
-# Therefore, a logarithmic and an exponential function will be used to
-# transform the targets before training a linear regression model and using it
-# for prediction.
-
-
-def log_transform(x):
-    return np.log(x + 1)
-
-
-def exp_transform(x):
-    return np.exp(x) - 1
-
+# Therefore, a logarithmic (`np.log1p`) and an exponential function
+# (`np.expm1`) will be used to transform the targets before training a linear
+# regression model and using it for prediction.
 
 X, y = make_regression(n_samples=10000, noise=100, random_state=0)
 y = np.exp((y + abs(y.min())) / 200)
-y_trans = log_transform(y)
+y_trans = np.log1p(y)
 
 ###############################################################################
 # The following illustrate the probability density functions of the target
@@ -103,8 +94,8 @@ def exp_transform(x):
 ax0.set_ylim([0, 2000])
 
 regr_trans = TransformedTargetRegressor(regressor=RidgeCV(),
-                                        func=log_transform,
-                                        inverse_func=exp_transform)
+                                        func=np.log1p,
+                                        inverse_func=np.expm1)
 regr_trans.fit(X_train, y_train)
 y_pred = regr_trans.predict(X_test)