From 9e0f38ae37b542c8921e34137d49a243b43ba371 Mon Sep 17 00:00:00 2001
From: Uwe Schmitt <uwe.schmitt@id.ethz.ch>
Date: Thu, 25 Feb 2021 18:44:04 +0100
Subject: [PATCH] added generalized linear models
---
07_regression.ipynb | 168 +++++++++++++++++++++++++-------------------
1 file changed, 95 insertions(+), 73 deletions(-)
diff --git a/07_regression.ipynb b/07_regression.ipynb
index afd65a3..a9e80d1 100644
--- a/07_regression.ipynb
+++ b/07_regression.ipynb
@@ -150,14 +150,18 @@
],
"source": [
"# IGNORE THIS CELL WHICH CUSTOMIZES LAYOUT AND STYLING OF THE NOTEBOOK !\n",
- "import matplotlib.pyplot as plt\n",
"%matplotlib inline\n",
"%config InlineBackend.figure_format = 'retina'\n",
"import warnings\n",
- "warnings.filterwarnings('ignore', category=FutureWarning)\n",
- "warnings.filterwarnings('ignore', category=DeprecationWarning)\n",
+ "\n",
+ "import matplotlib.pyplot as plt\n",
+ "\n",
+ "warnings.filterwarnings(\"ignore\", category=FutureWarning)\n",
+ "warnings.filterwarnings(\"ignore\", category=DeprecationWarning)\n",
"warnings.filterwarnings = lambda *a, **kw: None\n",
- "from IPython.core.display import HTML; HTML(open(\"custom.html\", \"r\").read())"
+ "from IPython.core.display import HTML\n",
+ "\n",
+ "HTML(open(\"custom.html\", \"r\").read())"
]
},
{
@@ -409,6 +413,7 @@
],
"source": [
"import seaborn as sns\n",
+ "\n",
"sns.set(style=\"ticks\")\n",
"\n",
"sns.pairplot(df, hue=\"kind\", diag_kind=\"hist\");"
@@ -451,7 +456,7 @@
"\n",
"# needs 2d data structure, features.iloc[2] has dimension 1\n",
"encoder = OneHotEncoder(sparse=False)\n",
- "one_hot = encoder.fit_transform(features.iloc[:, 2: 3]) \n",
+ "one_hot = encoder.fit_transform(features.iloc[:, 2:3])\n",
"\n",
"one_hot[:5, :]"
]
@@ -587,9 +592,9 @@
"source": [
"from sklearn.model_selection import train_test_split\n",
"\n",
- "(features_train, features_test, \n",
- " values_train, \n",
- " values_test) = train_test_split(features, values, random_state=42)"
+ "(features_train, features_test, values_train, values_test) = train_test_split(\n",
+ " features, values, random_state=42\n",
+ ")"
]
},
{
@@ -606,7 +611,8 @@
"outputs": [],
"source": [
"from sklearn.kernel_ridge import KernelRidge\n",
- "kr = KernelRidge(alpha=.001, kernel=\"rbf\", gamma=.05)"
+ "\n",
+ "kr = KernelRidge(alpha=0.001, kernel=\"rbf\", gamma=0.05)"
]
},
{
@@ -662,31 +668,33 @@
"\n",
"\n",
"def plot_fit_quality(values_test, predicted):\n",
- " \n",
+ "\n",
" plt.figure(figsize=(12, 4))\n",
" plt.subplot(1, 2, 1)\n",
"\n",
" x = np.arange(len(predicted))\n",
- " plt.scatter(x, predicted - values_test, color='steelblue', marker='o') \n",
+ " plt.scatter(x, predicted - values_test, color=\"steelblue\", marker=\"o\")\n",
"\n",
" plt.plot([0, len(predicted)], [0, 0], \"k:\")\n",
- " \n",
+ "\n",
" max_diff = np.max(np.abs(predicted - values_test))\n",
" plt.ylim([-max_diff, max_diff])\n",
- " \n",
+ "\n",
" plt.ylabel(\"error\")\n",
" plt.xlabel(\"sample id\")\n",
"\n",
" plt.subplot(1, 2, 2)\n",
"\n",
- " plt.scatter(x, (predicted - values_test) / values_test, color='steelblue', marker='o') \n",
+ " plt.scatter(\n",
+ " x, (predicted - values_test) / values_test, color=\"steelblue\", marker=\"o\"\n",
+ " )\n",
" plt.plot([0, len(predicted)], [0, 0], \"k:\")\n",
- " plt.ylim([-.5, .5])\n",
- " \n",
+ " plt.ylim([-0.5, 0.5])\n",
+ "\n",
" plt.ylabel(\"relative error\")\n",
" plt.xlabel(\"sample id\")\n",
"\n",
- " \n",
+ "\n",
"plot_fit_quality(values_test, predicted)"
]
},
@@ -901,7 +909,7 @@
"\n",
"- `sklearn.tree.DecisionTreeRegressor` expands the concept of decision trees to regression [is documented here](https://scikit-learn.org/stable/modules/tree.html#regression).\n",
"\n",
- "\n"
+ "- `sklearn.linear_model.TweedieRegressor`, `sklearn.linear_model.PoissonRegressor`, `sklearn.linear_model.GammaRegressor` offer so-called *Generalized Linear Models* which might be of interest if your target values are bounded or discrete. This is for example the case when your target values are rates or probabilities. The [wikipedia article about generalized linear models](https://en.wikipedia.org/wiki/Generalized_linear_model#Intuition) gives a nice intuition and [this discussion](https://stats.stackexchange.com/questions/190763/how-to-decide-which-glm-family-to-use) provides a guide which GLM should be used when."
]
},
{
@@ -942,22 +950,24 @@
}
],
"source": [
- "from sklearn.pipeline import make_pipeline\n",
- "from sklearn.preprocessing import StandardScaler, PolynomialFeatures\n",
+ "from sklearn.decomposition import PCA\n",
"from sklearn.kernel_ridge import KernelRidge\n",
"from sklearn.linear_model import LinearRegression\n",
"from sklearn.model_selection import cross_val_score\n",
- "from sklearn.decomposition import PCA\n",
+ "from sklearn.pipeline import make_pipeline\n",
+ "from sklearn.preprocessing import PolynomialFeatures, StandardScaler\n",
"\n",
"\n",
"def eval_regression(p, features, values):\n",
- " score = cross_val_score(p, features, values, scoring=\"neg_median_absolute_error\", cv=4).mean()\n",
+ " score = cross_val_score(\n",
+ " p, features, values, scoring=\"neg_median_absolute_error\", cv=4\n",
+ " ).mean()\n",
" print(\"cross val score:\", score)\n",
- " \n",
+ "\n",
" predicted = p.fit(features_train, values_train).predict(features_test)\n",
" plot_fit_quality(values_test, predicted)\n",
"\n",
- " \n",
+ "\n",
"p = make_pipeline(PolynomialFeatures(2), PCA(2), LinearRegression())\n",
"eval_regression(p, features, values)"
]
@@ -970,9 +980,10 @@
"source": [
"p = make_pipeline(PolynomialFeatures(), PCA(), LinearRegression())\n",
"\n",
- "param_grid = {'polynomialfeatures__degree': range(3, 6),\n",
- " 'pca__n_components': range(3, 11),\n",
- " }"
+ "param_grid = {\n",
+ " \"polynomialfeatures__degree\": range(3, 6),\n",
+ " \"pca__n_components\": range(3, 11),\n",
+ "}"
]
},
{
@@ -1007,7 +1018,9 @@
"source": [
"from sklearn.model_selection import GridSearchCV\n",
"\n",
- "search = GridSearchCV(p, param_grid, scoring=\"neg_median_absolute_error\", cv=4, n_jobs=4)\n",
+ "search = GridSearchCV(\n",
+ " p, param_grid, scoring=\"neg_median_absolute_error\", cv=4, n_jobs=4\n",
+ ")\n",
"\n",
"search.fit(features, values)\n",
"\n",
@@ -1110,9 +1123,9 @@
}
],
"source": [
- "import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
+ "import pandas as pd\n",
"\n",
"sales_data = pd.read_csv(\"data/sales.csv\")\n",
"sales_data.head()"
@@ -1190,8 +1203,8 @@
" features = np.zeros((len(sales) - window_size, window_size))\n",
"\n",
" for i in range(len(sales) - window_size):\n",
- " features[i] = sales[i: i + window_size]\n",
- " \n",
+ " features[i] = sales[i : i + window_size]\n",
+ "\n",
" return features, sales[window_size:]"
]
},
@@ -1210,11 +1223,12 @@
" assert np.all(X[0] == sales[:4])\n",
" assert np.all(X[1] == sales[1:5])\n",
" assert np.all(X[2] == sales[2:6])\n",
- " assert np.all(X[-1] == sales[-5: -1])\n",
- " \n",
+ " assert np.all(X[-1] == sales[-5:-1])\n",
+ "\n",
" assert np.all(y[0] == sales[4])\n",
" assert np.all(y[-1] == sales[-1])\n",
- " \n",
+ "\n",
+ "\n",
"test()"
]
},
@@ -1234,11 +1248,10 @@
"source": [
"from sklearn.kernel_ridge import KernelRidge\n",
"from sklearn.linear_model import Lasso\n",
+ "from sklearn.metrics import r2_score\n",
"from sklearn.model_selection import GridSearchCV\n",
"from sklearn.svm import SVR\n",
"\n",
- "from sklearn.metrics import r2_score\n",
- "\n",
"# ..."
]
},
@@ -1298,54 +1311,62 @@
}
],
"source": [
- "lasso_grid = {'alpha' : 10 ** np.linspace(-2, 3, 30)}\n",
- "svr_grid = {'C': 10 ** np.linspace(-4, 2, 30)}\n",
- "kernel_ridge_grid = {'alpha' : 10 ** np.linspace(-4, 3, 30)}\n",
+ "lasso_grid = {\"alpha\": 10 ** np.linspace(-2, 3, 30)}\n",
+ "svr_grid = {\"C\": 10 ** np.linspace(-4, 2, 30)}\n",
+ "kernel_ridge_grid = {\"alpha\": 10 ** np.linspace(-4, 3, 30)}\n",
"\n",
"WINDOW_SIZE = 36\n",
"X, y = create_feature_matrix_and_target_values(sales, WINDOW_SIZE)\n",
"\n",
+ "\n",
"def main(X, y):\n",
- " \n",
+ "\n",
" regressors = []\n",
- " \n",
- " for regressor, param_grid in [(Lasso(), lasso_grid),\n",
- " (SVR(), svr_grid),\n",
- " (KernelRidge(kernel=\"rbf\"), kernel_ridge_grid)\n",
- " ]:\n",
+ "\n",
+ " for regressor, param_grid in [\n",
+ " (Lasso(), lasso_grid),\n",
+ " (SVR(), svr_grid),\n",
+ " (KernelRidge(kernel=\"rbf\"), kernel_ridge_grid),\n",
+ " ]:\n",
" search = GridSearchCV(regressor, param_grid, scoring=\"r2\", cv=5)\n",
" search.fit(X, y)\n",
- " \n",
+ "\n",
" # we predict on the learning data set to get a general\n",
" # \"feeling\" how well the regressors work\n",
" predicted = search.predict(X)\n",
- " \n",
+ "\n",
" plot_regression(regressor.__class__.__qualname__, predicted)\n",
- " \n",
+ "\n",
" regressors.append(search)\n",
"\n",
" return regressors\n",
"\n",
- " \n",
+ "\n",
"def plot_regression(title, predicted):\n",
" plt.figure(figsize=(14, 4.5))\n",
" plt.suptitle(title)\n",
" plt.subplot(1, 2, 1)\n",
- " \n",
- " plt.plot(months, sales, label='sales', color=\"steelblue\")\n",
- " plt.plot(months[WINDOW_SIZE:], predicted, color=\"chocolate\", linestyle=\":\", label='predicted');\n",
+ "\n",
+ " plt.plot(months, sales, label=\"sales\", color=\"steelblue\")\n",
+ " plt.plot(\n",
+ " months[WINDOW_SIZE:],\n",
+ " predicted,\n",
+ " color=\"chocolate\",\n",
+ " linestyle=\":\",\n",
+ " label=\"predicted\",\n",
+ " )\n",
" plt.legend()\n",
- " \n",
+ "\n",
" plt.subplot(1, 2, 2)\n",
" plt.scatter(sales[WINDOW_SIZE:], predicted, color=\"steelblue\")\n",
" r2 = r2_score(sales[WINDOW_SIZE:], predicted)\n",
- " \n",
+ "\n",
" plt.title(\"r2 = {:.3f}\".format(r2))\n",
" plt.plot([0, 3], [0, 3], \"k:\")\n",
- " plt.xlabel('sales')\n",
- " plt.ylabel('predicted')\n",
+ " plt.xlabel(\"sales\")\n",
+ " plt.ylabel(\"predicted\")\n",
+ "\n",
"\n",
- " \n",
"regressors = main(X, y)"
]
},
@@ -1440,13 +1461,14 @@
"# start from line 96 in our feature matrix:\n",
"NLAST = 96\n",
"\n",
+ "\n",
"def forecast(X, y, regressor, n_last):\n",
- " \n",
+ "\n",
" # we crate a copy because we change below content\n",
" # current_window in place. Without copy we would\n",
" # also change X!\n",
- " current_window = X[-n_last, :].copy() \n",
- " \n",
+ " current_window = X[-n_last, :].copy()\n",
+ "\n",
" predicted = []\n",
"\n",
" for k in range(n_last):\n",
@@ -1456,7 +1478,7 @@
" # modify the window data in place:\n",
" current_window[:-1] = current_window[1:]\n",
" current_window[-1] = new\n",
- " \n",
+ "\n",
" return np.array(predicted).flatten(), y[-n_last:]\n",
"\n",
"\n",
@@ -1465,29 +1487,29 @@
"\n",
" x_axis = list(range(len(y)))[-n_last:]\n",
"\n",
- " plt.figure(figsize=(14, 5)) \n",
- " plt.subplot(1, 2, 1) \n",
- " plt.plot(x_axis, predicted, color=\"steelblue\", label='predicted')\n",
- " plt.plot(x_axis, correct, color=\"chocolate\", linestyle=\":\", label='true');\n",
- " plt.legend();\n",
+ " plt.figure(figsize=(14, 5))\n",
+ " plt.subplot(1, 2, 1)\n",
+ " plt.plot(x_axis, predicted, color=\"steelblue\", label=\"predicted\")\n",
+ " plt.plot(x_axis, correct, color=\"chocolate\", linestyle=\":\", label=\"true\")\n",
+ " plt.legend()\n",
" plt.xlabel(\"month\")\n",
" plt.ylabel(\"sales\")\n",
"\n",
" plt.subplot(1, 2, 2)\n",
- " plt.scatter(correct, predicted, color='steelblue')\n",
- " \n",
+ " plt.scatter(correct, predicted, color=\"steelblue\")\n",
+ "\n",
" r2 = r2_score(correct, predicted)\n",
" plt.title(\"r2 = {:.3f}\".format(r2))\n",
" plt.xlabel(\"sales\")\n",
" plt.ylabel(\"predicted\")\n",
"\n",
" mi, ma = np.min(correct), np.max(correct)\n",
- " plt.plot([mi, ma], [mi, ma], 'k:');\n",
+ " plt.plot([mi, ma], [mi, ma], \"k:\")\n",
" plt.suptitle(regressor.estimator.__class__.__qualname__)\n",
- " \n",
- " \n",
+ "\n",
+ "\n",
"for regressor in regressors:\n",
- " forecast_and_plot(regressor, NLAST) "
+ " forecast_and_plot(regressor, NLAST)"
]
},
{
@@ -1515,7 +1537,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.9.1"
+ "version": "3.7.7"
},
"latex_envs": {
"LaTeX_envs_menu_present": true,
--
GitLab