Data Preprocessing Lecture Notes PDF

{ "cells": [ { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "# Lecture 6. Data preprocessing\n", "\n", "**Real-world machine learning pipelines**\n", "\n", "Joaquin Vanschoren" ] }, { "cell_type...

{ "cells": [ { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "# Lecture 6. Data preprocessing\n", "\n", "**Real-world machine learning pipelines**\n", "\n", "Joaquin Vanschoren" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "tags": [ "hide-input" ] }, "outputs": [], "source": [ "# Auto-setup when running on Google Colab\n", "import os\n", "if 'google.colab' in str(get_ipython()) and not os.path.exists('/content/master'):\n", " !git clone -q https://github.com/ML-course/master.git /content/master\n", " !pip --quiet install -r /content/master/requirements_colab.txt\n", " %cd master/notebooks\n", "\n", "# Global imports and settings\n", "%matplotlib inline\n", "from preamble import *\n", "interactive = True # Set to True for interactive plots\n", "if interactive:\n", " fig_scale = 0.9\n", " plt.rcParams.update(print_config)\n", "else: # For printing\n", " fig_scale = 0.35\n", " plt.rcParams.update(print_config)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Data transformations\n", "\n", "* Machine learning models make a lot of assumptions about the data\n", "* In reality, these assumptions are often violated\n", "* We build _pipelines_ that _transform_ the data before feeding it to the learners\n", " * Scaling (or other numeric transformations)\n", " * Encoding (convert categorical features into numerical ones)\n", " * Automatic feature selection\n", " * Feature engineering (e.g. binning, polynomial features,...)\n", " * Handling missing data\n", " * Handling imbalanced data\n", " * Dimensionality reduction (e.g. PCA)\n", " * Learned embeddings (e.g. for text)\n", "* Seek the best combinations of transformations and learning methods\n", " * Often done empirically, using cross-validation\n", " * Make sure that there is no data leakage during this process!" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Scaling\n", "* Use when different numeric features have different scales (different range of values)\n", " * Features with much higher values may overpower the others\n", "* Goal: bring them all within the same range\n", "* Different methods exist" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "tags": [ "hide-input" ] }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "2a3f74459a504d19a14188536ea75856", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(Dropdown(description='scaler', options=(StandardScaler(), RobustScaler(), MinMaxScaler()…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.datasets import fetch_openml\n", "from sklearn.preprocessing import LabelEncoder, MinMaxScaler, StandardScaler, RobustScaler, Normalizer, MaxAbsScaler\n", "import ipywidgets as widgets\n", "from ipywidgets import interact, interact_manual\n", "\n", "# Iris dataset with some added noise\n", "def noisy_iris():\n", " iris = fetch_openml(\"iris\", return_X_y=True, as_frame=False)\n", " X, y = iris\n", " np.random.seed(0)\n", " noise = np.random.normal(0, 0.1, 150)\n", " for i in range(4):\n", " X[:, i] = X[:, i] + noise\n", " X[:, 0] = X[:, 0] + 3 # add more skew \n", " label_encoder = LabelEncoder().fit(y)\n", " y = label_encoder.transform(y)\n", " return X, y\n", "\n", "scalers = [StandardScaler(), RobustScaler(), MinMaxScaler(), Normalizer(norm='l1'), MaxAbsScaler()]\n", "\n", "@interact\n", "def plot_scaling(scaler=scalers):\n", " X, y = noisy_iris()\n", " X = X[:,:2] # Use only first 2 features\n", " \n", " fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(8*fig_scale, 3*fig_scale))\n", " axes.scatter(X[:, 0], X[:, 1], c=y, s=1*fig_scale, cmap=\"brg\")\n", " axes.set_xlim(-15, 15)\n", " axes.set_ylim(-5, 5)\n", " axes.set_title(\"Original Data\")\n", " axes.spines['left'].set_position('zero')\n", " axes.spines['bottom'].set_position('zero')\n", " \n", " X_ = scaler.fit_transform(X)\n", " axes.scatter(X_[:, 0], X_[:, 1], c=y, s=1*fig_scale, cmap=\"brg\")\n", " axes.set_xlim(-2, 2)\n", " axes.set_ylim(-2, 2)\n", " axes.set_title(type(scaler).__name__)\n", " axes.set_xticks([-1,1])\n", " axes.set_yticks([-1,1])\n", " axes.spines['left'].set_position('center')\n", " axes.spines['bottom'].set_position('center')\n", "\n", " for ax in axes:\n", " ax.spines['right'].set_color('none')\n", " ax.spines['top'].set_color('none')\n", " ax.xaxis.set_ticks_position('bottom')\n", " ax.yaxis.set_ticks_position('left')" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "tags": [ "hide-input" ] }, "outputs": [], "source": [ "if not interactive:\n", " plot_scaling(scalers)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "### Why do we need scaling?\n", "* KNN: Distances depend mainly on feature with larger values\n", "* SVMs: (kernelized) dot products are also based on distances\n", "* Linear model: Feature scale affects regularization\n", " * Weights have similar scales, more interpretable" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "tags": [ "hide-input" ] }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "e518fe21521e441e982cd9bf4e4b2e0b", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(Dropdown(description='classifier', options=(KNeighborsClassifier(), SVC(), LinearSVC(), …" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.svm import SVC, LinearSVC\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.datasets import make_blobs\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.base import clone\n", "\n", "# Example by Andreas Mueller, with some tweaks\n", "def plot_2d_classification(classifier, X, fill=False, ax=None, eps=None, alpha=1):\n", " # multiclass \n", " if eps is None: \n", " eps = X.std(axis=0) / 2.\n", " else:\n", " eps = np.array([eps, eps])\n", "\n", " if ax is None: \n", " ax = plt.gca() \n", "\n", " x_min, x_max = X[:, 0].min() - eps, X[:, 0].max() + eps\n", " y_min, y_max = X[:, 1].min() - eps, X[:, 1].max() + eps\n", " # these should be 1000 but knn predict is unnecessarily slow\n", " xx = np.linspace(x_min, x_max, 100) \n", " yy = np.linspace(y_min, y_max, 100) \n", "\n", " X1, X2 = np.meshgrid(xx, yy) \n", " X_grid = np.c_[X1.ravel(), X2.ravel()] \n", " decision_values = classifier.predict(X_grid) \n", " ax.imshow(decision_values.reshape(X1.shape), extent=(x_min, x_max, \n", " y_min, y_max), \n", " aspect='auto', origin='lower', alpha=alpha, cmap=plt.cm.bwr) \n", "\n", "clfs = [KNeighborsClassifier(), SVC(), LinearSVC(), LogisticRegression(C=10)]\n", "\n", "@interact\n", "def plot_scaling_effect(classifier=clfs, show_test=[False,True]):\n", " X, y = make_blobs(centers=2, random_state=4, n_samples=50)\n", " X = X * np.array([1000, 1])\n", " y, y = 0, 0 \n", " X_train, X_test, y_train, y_test = train_test_split(X,y, stratify=y, random_state=1)\n", " \n", " clf2 = clone(classifier)\n", " clf_unscaled = classifier.fit(X_train, y_train)\n", "\n", " fig, axes = plt.subplots(1, 2, figsize=(7*fig_scale, 3*fig_scale))\n", " axes.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap='bwr', label=\"train\")\n", " axes.set_title(\"Without scaling. Accuracy:{:.2f}\".format(clf_unscaled.score(X_test,y_test)))\n", " if show_test: # Hide test data for simplicity\n", " axes.scatter(X_test[:, 0], X_test[:, 1], c=y_test, marker='^', cmap='bwr', label=\"test\") \n", " axes.legend()\n", " \n", " scaler = StandardScaler().fit(X_train)\n", " X_train_scaled = scaler.transform(X_train)\n", " X_test_scaled = scaler.transform(X_test)\n", " clf_scaled = clf2.fit(X_train_scaled, y_train)\n", "\n", " axes.scatter(X_train_scaled[:, 0], X_train_scaled[:, 1], c=y_train, cmap='bwr', label=\"train\")\n", " axes.set_title(\"With scaling. Accuracy:{:.2f}\".format(clf_scaled.score(X_test_scaled,y_test))) \n", " if show_test: # Hide test data for simplicity\n", " axes.scatter(X_test_scaled[:, 0], X_test_scaled[:, 1], c=y_test, marker='^', cmap='bwr', label=\"test\")\n", " axes.legend()\n", "\n", " plot_2d_classification(clf_unscaled, X, ax=axes, alpha=.2)\n", " plot_2d_classification(clf_scaled, scaler.transform(X), ax=axes, alpha=.3)" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "tags": [ "hide-input" ] }, "outputs": [], "source": [ "if not interactive:\n", " plot_scaling_effect(classifier=clfs, show_test=False)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "### Standard scaling (standardization)\n", "* Generally most useful, assumes data is more or less normally distributed\n", "* Per feature, subtract the mean value $\\mu$, scale by standard deviation $\\sigma$\n", "* New feature has $\\mu=0$ and $\\sigma=1$, values can still be arbitrarily large\n", "$$\\mathbf{x}_{new} = \\frac{\\mathbf{x} - \\mu}{\\sigma}$$" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "tags": [ "hide-input" ] }, "outputs": [ { "data": { "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1R5cGUgL0NhdGFsb2cgL1BhZ2VzIDIgMCBSID4+CmVuZG9iago4IDAgb2JqCjw8IC9Gb250IDMgMCBSIC9YT2JqZWN0IDcgMCBSIC9FeHRHU3RhdGUgNCAwIFIgL1BhdHRlcm4gNSAwIFIKL1NoYWRpbmcgNiAwIFIgL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9UeXBlIC9QYWdlIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovTWVkaWFCb3ggWyAwIDAgNDE5LjMyNDA2MjUgMTczLjg4NDg3NSBdIC9Db250ZW50cyA5IDAgUiAvQW5ub3RzIDEwIDAgUiA+PgplbmRvYmoKOSAwIG9iago8PCAvTGVuZ3RoIDEyIDAgUiAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJy9m0+THMeNxe/9KepoH7qYyD/IzKO1shXhy1omY/ewsQeGRHOlIOkQZa++vn8ve6a7Z4ieIR2yxRiKXVNdhUoA7z0gUS++fvP/P3z35s/ffLX9x8vDi8un734+2PYjP2+3tP3Izy+bbd/w8/aQtmPa3h+qzb3kmjzr87vrz9bLPkYdvXE8Pfz4f4fDXw4vfseFfuaq3xwOlvbifK1sfc9c12be58i51nXg3fUBa86Vhg6ev3U5ti790+Hb7aft/te5rYvYyLvbOP3hY52785WPb7b/3j5sL36XT9Zg6vbxLX/9yM8vW5Lt25+/Of+G5Tj901reUytjzNxL26z0PfVWevXefPvu/fbiT2n7+q/n84+cXrvP2i3P7Tj3ltscfLWEZ6d9OGe41z421o5zOX9adHLarc02LOWpD2XmanXMUYNz855T9zq797L5PnNrbnPmEl63p2Kdu6bKh1xy9pwG5gcGl92nt957y3yqO8YWL/z04GzbsaHmmSZLYbsTGAUfhxbryrkkTy01zu57tVxq9uaRHWUvzatZSkkfuueBwTm1cC3GwG/dWJFt7I0A61qQHLtvtNqxuKapTy3PMkri+7H7amZZCRIfsjllliWxOhY/YS15FIIJ//Ld2XqqWGKRV5Ri+KWM2hSFO//jwmnGl0478TgSnqu+tZ1H5FMt4ZXxhNkcZpy/Hdvea69WSq3hlVkQfNKINp6Q1SuYUbEq9sr0PlgR0g4rsntRJJXwwpW8JRQ66+2yCTMsFTeLfEiM4l0b7qmczEgNs7zFAW25dR/GJVlH4k8uJQvj58N9XA03ryAFN4pZqW2EoZQIyMbz45QjaDQJI7IqfMK0p9SJs2HZycEEjDih12/4hFBoNgEPnGK2N60cfsrxckysdG+cQkhjAJE3U4oSywRYWG0Lv3YgwyYpPqPwlxU+QfDc6voiSIb7LFoLwjkJq7MZ2LU+DQclZxgaRChhUSY5mE6p4LVMXBjCDBlNPlcMt75VwhlIZT1maIfBGSN1rLBTfhcT4PYYkcwJSCen87KJuCZVRg8hCSrhatxdVHK0Sn5bZf3A1htOaU5EE3kC4GZavBG5WyjaSy+zk1WEBqiv4J5haHAyv4Fz8oo54D+LIfxGViVTMhuAzDcHADw4UqMoIgNNGQg2VqzgLIAP/gmBkWhw7gsSJhf7dJ4tzz5usETmAVvPEM7R8ImJ2ZYuCEMD0ICqRhbJE0i18KfnEPkBNyAjdbBOIDlhTiyZfoteB2lB5MhmggSUNB/huSYrHA4UDSoDB1gG16UcRZ3tyjnpAkAUDdJbaylBWPGluTWSgb8EuJOcYT2sR3aAt9IbJJ7gFDhQJsS+JuB6IXpFwcvZeAc88nR9XU4izfqADOqCJCJeYcqCSAE99WvpoatfHz759mU1WDzsgMkqLhZt2sApBXNucArp2pKbC2amCUZbTG15R8VY77mW9UWkXpVKCFNl7MoN3AJ/s+xguxCMYJlRfrtQZoCaAnt4grNgmRGjjKSO0BjB01gWh8crEVvj9IbogUXOV2Lh75FQoTNcjEEuwYFYmgUyhTwcPK15qC/nTkAYJCkGdiEwRIAmDANkSCUNh7IBLHmFC5c2Y6K3vcKCY4pz+ITnq5adQ5HZTaBEYsHAbcmT0m1U4DJWrqUQwxlf+HILFJpAUpI45s2WSQDpRp6Xhy2pthTTiqne6NMVa8qdOlGRrYZr57twBVptdXAutCxiTLEcRYBWicaTSkJvspQdWVBiXVDB/jQUybIJ9vGpDzEHeUoTqSNyU4VQWq6ee8gU4lU4pTahHKs+UcQs5JzRlSEd+EbRhBvIySqMQU+H4hw8RKw2Xa8tLTOQjCi3UDOyzCANgjxZXppJZEsMhtgo3uR5DLG6uELVQRLER1fOEqsIW9BUgAd+qAAJczDtgKx8xtKdyJu4EOPGAA2O5yK3LCbM0CfiMYWyQEKeLLFE2I2VvtLapYX1hDgWZ+AQ4mwRI6pX5DVu5JVDFNTMJkzSI6IikrDjhphxgkPVtW4kMZ1ZH4subbtkrrkVYJY4USkxvIaYK0EJ2QNfVecCztInMRII502LZacaCArgHkDSrQw0IpK6rStdE5UkkiL0tQiN81TUSPaA+CyyKpboVLCVuhghZ+LMJtQHL2Lcl96gpJGAPuk0tGjrteXQ2SwrNqAnqWIFMBXNiyXhYgjpKwDO1cqqWRLRSeF5S+QiJbOUESpfRRlKhRSL6adSEIshEuIO3SpgAshH/IRZcNRKXcUE4KSsxj2gRvyEVaKHv+paHfIWLFApFi51plBC7aDIQRhIgOjMQEcMGkQkOkpmci7VRENSWQihqo5FC520WvSjyrfPsACqwDEPyPPMuYkgJNRUdsbYTLhhAfJysVwVvHB62DYhklTtEj1I/koEAgpUE9dJsp5sqaH1j9UHsocKB/rHGEl0KiNIeLn1VqGvhUkARK7zROk8ttoVM1TLXSlfZymdoCIokpc0Z3wuvkQ2gSacUYTKJHSrFvdqyGO8SgHaT/0XEBnAJ08j5EGkUxZWwhyAwApXye+hvDlaR8AlNJsaB0srQwtU2as6D7BH4Qgtq9/RiRrqgCE+iMtKamygNK9WysJDPAbkj5CXMnII5JbYXDKExGaZcU0YkCrgxI2QyBLilfJ5IdutNokyqAsoT9ISHsPtocQR5091rFCF8gvYU6kzQ+2k+JAAmEOQubSFmyhh3ihi0L+mWB+CfBCI0mOEkbfwiWCX5iWukORE7LBQDSV18gYrDWVo1QlmalCAIooO3534pBaaejz0XEbtNHgh7r/AbqRjL1o4cxEEXwDpw2VOqslhU+Hwgvos6Vdjid+AawRt4dGqAsugvAL/+A1trTYX0kJdFz4ZFiO1U1gOk6dklro0cLlaPRMIBVfizi0pML0oTFc3lmWj7FYfMiZqeRcayU0aDuhuarpFLiQauhShOo6LQZYCvZHhLt5LncyVD32v8BLgPENWUJm96vDVc6hyCRgaql+JJ/Um0G8rbVTd8oUQ5vktMJM6TsysGqROOI8ZFrcoOEoRar0kZaHrshJF/BsDd1WXjxocV6uJTYFB6IfJOinXp8PUviSASjueDdEQ9s8kHqn9HO2RV6NLnVijtLuhyVyVvpRpWwqOYmPilZCcCGctqrCQJQDbCVAqnRqr+9WOnlRNalcBX1SuYNOMlQh6d8KRoy4IIZJZyBvtCaEocol7d1tKpIFeAH8JO1eQG+oYSLK+umIwpXYPyq3lqGrmdECrny5NOQ7BhcCRtU8Ba9a8ai6K4IkAtLjHBFKgytTVp7zhkxijelxk8FseB8pEUEBeprY9NBGKXwBJ7L1KMikN4IM1h3FvaLiO+qCo7Mskr8qqPOMeDHwsROH3fUFZlv6mRA9FDiKuyQuCuxUoOaNyvccyHCRQVo0TGkHLYqK4UQn7JBFZzr7KBikuKLk/oKpvD+ftr9POFwZo72uT3Hl7tc82tIdVtUF3OZTWoXeHw1fc9ZfDT+vex5N8yVuvPHhe3b3DV68OL/5A9m6v/nJQI0ikqcZf2l59f/if7TfWfrv97/bqj4ffvzr8dGo5rce4/Esm/8GOX7/58fV//f3l6w8/H9//8OHvP/MM27c8xTL7UNXPA3X9Yu7Vodvmlg7K3/2pX2Z4+lUM7+JwJwDOdl+O3Dbbp7QLFTt6pH+R2b/OclMc7yho0Po6PC7HbltONu89J3CKSsdump7OBl8t8/nuJYuR1DC53P3q2BN3L3AZuACXqzB+/u4tuLuzrrUN7bWc73517Im7OwwO8JCdUO9n3N2ih7/sil9uf3XsidvPtfmD8CjawfiM20dPf/FyZsW1m/7+MOfjY9H9p9rCgqHl+YwxWZXQ50Rt/bWjtilavD6w/f7Y87a3tK9t78+yPf/LMu5i+t2h0HK5mQqgLMsJD/5ZvjThLjc30wYbSv769ueDzxsAnUvIzacsyE9bUNceQX5owf3Bz7AAYoCS21MW1GsL8vbHLZ0J8tGszMWuuymWK0PPAyqHl5fJl0Wxj4dd3sfDLocPZ5K+J+eQmD+BhPCWn0zOPH/X9eRXgzF34uBqbQfaus20/gjh8rrIZWERkw+X9j8//vD2hw+v321fv/7b6+tlvjP3evIo3U0eHXKe6nEJX0/LVU0d+zV0dBoWujpweZKrr4XTQuff3wHyPzcxdPjl7kA0MZQHVF2K6qWhNjw01QfVBUt+X2vZQ/E29qptC4rpRH0BKE3NzAAWKTy/731SdK7aXqBatdlW/bwpbQ8ruVoo+yjkXNUZVctcw0senTu1haiClypqK5hFpZyH270+tUcCvKh861bVicJibcvelzkPTS7SshQNKQ3tAOW2xhuQ95ai89XImaji7txbtZ1RaWurukSGHFlhihWK4DI1W1DVuKGs0m5XcHpBpKt/VhElLMlplALDmkdrgqkof3KI6KhbRTxOlePDZugbTu/aVBEclFMrunWUpnb7Q9P3VdxTHDWVllO5rOo4XvIjtqsS0S41sUQRSCGasmZooicd3L2oP5DUgB7aOadMSTWF3tdoAeVZJRbnxpIW1wbIzD003NRcnBqGaMtyU9BSZOZzn+7ThelU+qzDwEUa7yq1DE1aRT7SVNIcrNyQcJgaJ0hwWL6v5h8tyyS81DLKY+2odXVHKcSiRemU0c7yFW2DngxJbc24RA4llEox9RUAzGZqEBWlSIusPmb1Y8gtXDK1nioNregpoksPYMeqWuRqspBuJZdmtZ3L7genJ40ekbpNO1jKp64M6mpHB5ZMUIGrIz1V/jt0SL0rmWJRXI2dsqINbFHvt6kto7kKy5HrgcepIRpTywIQwOFrhyNyO2YU1o/kaavxNWZz6tI4QsA16tbq6vKsHrNmXMq54n0Uf+RZ1exR1VQqyJVZam2qhhcvFP1lzVxoVimrI1m7tzzjTODeE7zO1NCmnNfuJJxgYWibugtTe0kasFL3RDEOuJXIjcdcVkuITDGNDHIn7b4gVEoU2nimpdzLyGlsJrHJgs4wJRUTJtyZE/oqGgcRTmmH9AYkJ08OnCiE9FHJXkYbHkXrsfhuaxOzTdeuoFH+VG2RzzBcCedxai8C+Fx8DvEQwR3iGqTcex2akdEuPuoykZUWBusROKhgE5C59oEtQZjZNICgRQoJM6lUM42/9rVVr2eeIzpZ7Kst86J+y9aa4ttnnaElQ338oQZV0Z5JRW5owKCFVhj46qb2PalFTqoPrL4pvB15fgls1g9aKkCP5j5I+OYjjBNdHfAjYInFfPKuWvRe2ohMB45zV3e0VGgNZ3b1/hLQH9o+do3RkDNJXemshihERXo+oPt/z6hRamtKwHFLr2t3bU6RAykY2p7Fexq8XcNqZKuzhm58OVrIrPgA7U1b+lwdMFyN395CpXIcyn5tdrS1Wd7XWEmrOKtEFOEaQNQQWs0nQjFNCOaWY2mIlAHiE5HbKS8Kwq9r1yXmTGmDIk7AT8o/okuDAy2M82NvkndFPWWNbHXNwfJdwiFE0DEolbX9A6LkhS6QLQtZY3nA1bOyphNlSwbhH3ON1sZqQhMjuIdrg1ZHT9q3UcPTc/SobY2mguYOaEjYIIjV5tCsQOBSBZYmtVzN0a7NK0Kfy+eYZ6t+qzbbUNnvwt6iEYwbaTcFLE2DY0u3soRCplAeNjiNX2n0B0mgEUOb2rGw0O6jd83+JMTKUKddc5QoJmg6XnSFLrpGIxW+UA/QI3gk6CJ8Aa9UPZmaiqeqY20uuMcehZ4pCxqkKHFIETH7RM9FS5i1o9fxJ6HkmxMr2g7UgH50Za6r4jx7UToTt1l4m0cMXCy4Mq2uSeK1/k3DpbnECSS2Ra9pyHdtsICgs2s2PlxAjbGRbVSHwBwoNzX1LhiNztYmrmt3Kq1dEK4qEKoh689dyYAkqOPkSEwmZrlXaLVeoikF8CaoVZVpdiFpZiFUnUVbYwCaj76IZU+zie9IiDBkq0rClFVe+gJzVZIa7Orx+V256RrKWrUb5b90nBRxqJg1GStJDTxQ/RLxoxcNVo5oYVQgIUvT4n4tuTpVCAUPEyJprqI2cexpk0+j+BqHjCWiqzyqmhxSry/prYdCZPm4UXJIW2n3fGNFIQm9JXJ+weCxnIRmDQGUlnZayIvqzyMu21AsK+s1AW9L9Gk6CSyndAthXyM5AwFEhZ1l+YCASLmSIyhHmOllI85Z41pEvGnYH/V6Qw2R6JrUtbup4dQN2qghTUBAXFc7dgiDim7GAVQ0YUIcAaDErTWArEipmjipJo0T4xtlm7veWFl9D8EbNKriqUT+PHVu3KteiMmkEJCC3B0WyduuYVbIKXGqhlVsvaeQYyWs8RyqDlLsVLJrA5KY9ciKorIkreFeqYkqYSRtFItsR6pKYCeQpK9qUx2SuprKN+LKXO9yaBIewzUXmlmQB5XsYqlnppUgyKJw1+o0RYTqKiIwjIi6Ql+vWmnm+NiV4fgA2JuRROqqvVReT813AJoOEKqWD92m1kBbo7AaOVyzI+utNS9xo2BNKcDUxsmqC/SGVLYZchiJBId1MBpfqVOjttmMc4rA3NeWqF7+wDOayXAecPS4b0a4UdrUqcKkS1whC7uQ7Ia6F2+BCPDzKkr0uoQGJnq3UGZWiRLqPg20boMQlbuaRiduQLxYkTS1fmoDpYa0Q/HELcL1QsIwzU1pGSvrCNugr3LoUeSXhi0BPxUP2sVew24j7IitcJG1HZ0xT3mLXgJDvIZL45Rr1D1qRGXASgPKlCU1xDLov2sFpUXnVtWw1Exay6G2d8EkEjDrJcjjmkIXY7cUorZqqPVWStFwLVoMzZDWiwUh8u2reuZ8weQ8vbtHwFgK66+pehjO0zxSWx7q6qPClGHjDyglGe+Hy/qAyfTioLI/xGwRiN72xOl5eaDpblUzsaE/NQmf1f6RgDbVa13Tkjf6Z12vhcFm2ApL6xVQDBn5RrBocLyqz6dXM6kETI/Bp9BFShzAsp3ppkvYqYEWxlbTYJKpUdz1oh+k7Jo6HyluK3b1S5uJSIcY3NdGJ4gRN66obhFF6oMtHQPi68VXD5UJsQe2kMEqzZZmQq7BT2HHqCJKNQFDdPjdpUstsSYtGvHU6xCrep4IoKHXiJBkoW+GBg9BQZd3ViDmobkskimWa+oqtqWo83pZilp5Fcs3tCA4jO9t4Yr2coGUSbSEXTeMnU3zWBKDW60akMwqAWN+B56LJuTUOsiqGqqGeYjcsC9PDKqmaKZrizWqUts0S34DcJ36BcBsfb2gkSE5vcKQQm+uYU5RhWsVpTopShCpN7qLRYUauYwqWkP7Be2gacm43iH39botWK+9ttXnEd2NuNh1aSpVO+oQKsSXWK+jjii0mjrU6y0wpI+as3XNU6UUlmm4RGP7iB+Vg0PvmyY1BWMhONAtIMh6mfVURZB8t5rbZW23TQ0XAq/H1teL7p1ICRuAiiW9H6R3A7Y28KZe2eXnlkLG+QQH4XJ6B7C43qY8z3U+8o/eLNcwI2VSXwmnbhCiLT+ou5+ZeMq9Scw9mCG6PnZ7sCKrXVIn0fHlU0S/ynAA1AurPhyIuT72xPyTE3D2BQMx9unePEm9L27MAgGCTHe38ehYeHexx4IYbSFvFDpV7+b/O1fuYjtEw4L5Q+PvD8bWF82Ma4dP1pMrGoR+crbCPneu4Mqu04b3taFPzBU82hd/H++LR3MFV3vllxG6y5efGCwIdtmfv+/zkwUlFZGXXiNcy/v8aMHLv73+8P3rj79Vn0Ez9Kf/tt98//K71+/efLxe+8M/AAU/9EAKZW5kc3RyZWFtCmVuZG9iagoxMiAwIG9iago1ODQwCmVuZG9iagoxMCAwIG9iagpbIF0KZW5kb2JqCjE5IDAgb2JqCjw8IC9MZW5ndGggMTY0IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nD2QwRFDIQhE71axJYCAQD3JZHL4v/9rQJNcZB1g96k7gZBRhzPDZ+LJg9OxNHBvFYxrCK8j9AhNApPAxMGaeAwLAadhkWMu31WWVaeVrpqNnte9Y0HVaZc1DW3agfKtjz/CNd6j8BrsHkIHsSh0bmVaC5lYPGucO8yjzOd+Ttt3PRitptSsN3LZ1z06y9RQXlr7hM5otP0n1y+7MV4fhRQ5CAplbmRzdHJlYW0KZW5kb2JqCjIwIDAgb2JqCjw8IC9MZW5ndGggMjE1IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nDVQS64DMQjb5xS+QKVAEgjnmap6m95/+zCoixHW4B/x65g4hpcs+BL4VrxlrNgwU3zHUi2kdmDrQHXhXMUz5AqOXohN7LTgXLa5SaSuIEMiSiPbckOXULQpOiV/b5pbuim3J02abtPKwNTLkLMjiBhKBktQ07VowqLtqfid8oy/OosXfscOK3Tc4dNTEfU9w2c+iEu9xvao2ZlEui/IkNOaSwlNNNnR78gUzjw3pJDOIEFSXIJrZZD96ZejAxIwMbcsQHZXopwlacfSvzN40ucfYVxUjQplbmRzdHJlYW0KZW5kb2JqCjIxIDAgb2JqCjw8IC9MZW5ndGggMzQxIC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nDVSO9KbQQjrv1PoAp5Z3st5nMmk+HP/NgI7FSywQgLSAgeZeIkhqlGu+CVPMF4n8He9PI2fx7uQWvBUpB+4Nm3j/VizJgqWRiyF2ce+HyXkeGr8GwI9F2nCjExGDiQDcb/W5896kymH34A0bU4fJUkPogW7W8OOLwsySHpSw5Kd/LCuBVYXoQlzY00kI6dWpub52DNcxhNjJKiaBSTpE/epghFpxmPnrCUPMhxP9eLFr7fxWuYx9bKqQMY2wRxsJzPhFEUE4heUJDdxF00dxdHMWHO70FBS5L67h5OTXveXk6jAKyGcxVrCMUNPWeZkp0EJVK2cADOs174wTtNGCXdqur0r9vXzzCSM2xx2VkqmwTkO7mWTOYJkrzsmbMLjEPPePYKRmDe/iy2CK5c512T6sR9FG+mD4vqcqymzFSX8Q5U8seIa/5/f+/nz/P4HjCh+IwplbmRzdHJlYW0KZW5kb2JqCjIyIDAgb2JqCjw8IC9MZW5ndGggMzA3IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nD2SS24DMQxD9z6FLhDA+tme86Qoupjef9snJemKHNkWRWqWukxZUx6QNJOEf+nwcLGd8jtsz2Zm4Fqil4nllOfQFWLuonzZzEZdWSfF6oRmOrfoUTkXBzZNqp+rLKXdLngO1yaeW/YRP7zQoB7UNS4JN3RXo2UpNGOq+3/Se/yMMuBqTF1sUqt7HzxeRFXo6AdHiSJjlxfn40EJ6UrCaFqIlXdFA0Hu8rTKewnu295qyLIHqZjOOylmsOt0Ui5uF4chHsjyqPDlo9hrQs/4sCsl9EjYhjNyJ+5oxubUyOKQ/t6NBEuPrmgh8+CvbtYuYLxTOkViZE5yrGmLVU73UBTTucO9DBD1bEVDKXOR1epfw84La5ZsFnhK+gUeo90mSw5W2duoTu+tPNnQ9x9a13QfCmVuZHN0cmVhbQplbmRvYmoKMjMgMCBvYmoKPDwgL0xlbmd0aCAyMzIgL0ZpbHRlciAvRmxhdGVEZWNvZGUgPj4Kc3RyZWFtCnicNVFJbsQwDLv7FfzAANbuvCfFoIf2/9dSyhQIQCW2uCViYyMCLzH4OYjc+JI1oyZ+Z3JX/CxPhUfCreBJFIGX4V52gssbxmU/DjMfvJdWzqTGkwzIRTY9PBEy2CUQOjC7BnXYZtqJviHhsyNSzUaW09cS9NIqBMpTtt/pghJtq/pz+6wLbfvaE052e+pJ5ROI55aswGXjFZPFWAY9UblLMX2Q6myhJ6G8KJ+DbD5qiESXKGfgicHBKNAO7LntZ+JVIWhd3adtY6hGSsfTvw1NTZII+UQJZ7Y07hb+f8+9vtf7D04hVBEKZW5kc3RyZWFtCmVuZG9iagoyNCAwIG9iago8PCAvTGVuZ3RoIDIzMSAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJw1TzmSBCEMy3mFPjBVGNtAv6entjbY+X+6kplOkPAhydMTHZl4mSMjsGbH21pkIGbgU0zFv/a0DxOq9+AeIpSLC2GGkXDWrONuno4X/3aVz1gH7zb4illeENjCTNZXFmcu2wVjaZzEOclujF0TsY11radTWEcwoQyEdLbDlCBzVKT0yY4y5ug4kSeei+/22yx2OX4O6ws2jSEV5/gqeoI2g6Lsee8CGnJB/13d+B5Fu+glIBsJFtZRYu6c5YRfvXZ0HrUoEnNCmkEuEyHN6SqmEJpQrLOjoFJRcKk+p+isn3/lX1wtCmVuZHN0cmVhbQplbmRvYmoKMjUgMCBvYmoKPDwgL0xlbmd0aCAyNDkgL0ZpbHRlciAvRmxhdGVEZWNvZGUgPj4Kc3RyZWFtCnicPVA7jkQhDOs5hS/wJPIjcB5Gqy1m79+uA5opUEx+tjMk0BGBRwwxlK/jJa2groG/i0LxbuLrg8Igq0NSIM56D4h07KY2kRM6HZwzP2E3Y47ARTEGnOl0pj0HJjn7wgqEcxtl7FZIJ4mqIo7qM44pnip7n3gWLO3INlsnkj3kIOFSUonJpZ+Uyj9typQKOmbRBCwSueBkE004y7tJUowZlDLqHqZ2In2sPMijOuhkTc6sI5nZ00/bmfgccLdf2mROlcd0Hsz4nLTOgzkVuvfjiTYHTY3a6Oz3E2kqL1K7HVqdfnUSld0Y5xgSl2d/Gd9k//kH/odaIgplbmRzdHJlYW0KZW5kb2JqCjI2IDAgb2JqCjw8IC9MZW5ndGggMjQ5IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nE1RSYoDMAy75xX6QCFek7ynQ5lD5//Xyg6FOQQJr5KTlphYCw8xhB8sPfiRIXM3/Rt+otm7WXqSydn/mOciU1H4UqguYkJdiBvPoRHwPaFrElmxvfE5LKOZc74HH4W4BDOhAWN9STK5qOaVIRNODHUcDlqkwrhrYsPiWtE8jdxu+0ZmZSaEDY9kQtwYgIgg6wKyGCyUNjYTMlnOA+0NyQ1aYNepG1GLgiuU1gl0olbEqszgs+bWdjdDLfLgqH3x+mhWl2CF0Uv1WHhfhT6YqZl27pJCeuFNOyLMHgqkMjstK7V7xOpugfo/y1Lw/cn3+B2vD838XJwKZW5kc3RyZWFtCmVuZG9iagoyNyAwIG9iago8PCAvTGVuZ3RoIDk0IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nEWNwRHAIAgE/1RBCQoK2k8mk4f2/40QMnxg5w7uhAULtnlGHwWVJl4VWAdKY9xQj0C94XItydwFD3Anf9rQVJyW03dpkUlVKdykEnn/DmcmkKh50WOd9wtj+yM8CmVuZHN0cmVhbQplbmRvYmoKMjggMCBvYmoKPDwgL0xlbmd0aCAzNDEgL0ZpbHRlciAvRmxhdGVEZWNvZGUgPj4Kc3RyZWFtCnicRVJLbkQxCNu/U3CBSOGXkPO0qrqY3n9bm0zVzeAJYGx4y1OmZMqwuSUjJNeUT30iQ6ym/DRyJCKm+EkJBXaVj8drS6yN7JGoFJ/a8eOx9Eam2RVa9e7Rpc2iUc3KyDnIEKGeFbqye9QO2fB6XEi675TNIRzL/1CBLGXdcgolQVvQd+wR3w8droIrgmGway6D7WUy1P/6hxZc7333YscugBas577BDgCopxO0BcgZ2u42KWgAVbqLScKj8npudqJso1Xp+RwAMw4wcsCIJVsdvtHeAJZ9XehFjYr9K0BRWUD8yNV2wd4xyUhwFuYGjr1wPMWZcEs4xgJAir3iGHrwJdjmL1euiJrwCXW6ZC+8wp7a5udCkwh3rQAOXmTDraujqJbt6TyC9mdFckaM1Is4OiGSWtI5guLSoB5a41w3seJtI7G5V9/uH+GcL1z26xdL7ITECmVuZHN0cmVhbQplbmRvYmoKMjkgMCBvYmoKPDwgL0xlbmd0aCA3MiAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJwzMrdQMFCwNAEShhYmCuZmBgophlxAvqmJuUIuF0gMxMoBswyAtCWcgohngJggbRDFIBZEsZmJGUQdnAGRy+BKAwAl2xbJCmVuZHN0cmVhbQplbmRvYmoKMzAgMCBvYmoKPDwgL0xlbmd0aCA0NyAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJwzMrdQMFCwNAEShhYmCuZmBgophlyWEFYuF0wsB8wC0ZZwCiKewZUGALlnDScKZW5kc3RyZWFtCmVuZG9iagozMSAwIG9iago8PCAvVHlwZSAvWE9iamVjdCAvU3VidHlwZSAvRm9ybSAvQkJveCBbIC0xMDIxIC00NjMgMTc5NCAxMjMzIF0gL0xlbmd0aCAzOQovRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJzjMjQwUzA2NVXI5TI3NgKzcsAsI3MjIAski2BBZDO40gAV8wp8CmVuZHN0cmVhbQplbmRvYmoKMzIgMCBvYmoKPDwgL0xlbmd0aCAxNjMgL0ZpbHRlciAvRmxhdGVEZWNvZGUgPj4Kc3RyZWFtCnicRZA7EgMhDEN7TqEj+CMDPs9mMik2929j2GxSwNNYIIO7E4LU2oKJ6IKHtiXdBe+tBGdj/Ok2bjUS5AR1gFak42iUUn25xWmVdPFoNnMrC60THWYOepSjGaAQOhXe7aLkcqbuzvlDcPVf9b9i3TmbiYHJyh0IzepT3Pk2O6K6usn+pMfcrNd+K+xVYWlZS8sJt527ZkAJ3FM52qs9Px8KOvYKZW5kc3RyZWFtCmVuZG9iagozMyAwIG9iago8PCAvTGVuZ3RoIDgzIC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nEWMuw3AMAhEe6ZgBH4m9j5RlMLevw0QJW64J909XB0JmSluM8NDBp4MLIZdcYH0ljALXEdQjp3so2HVvuoEjfWmUvPvD5Se7KzihusBAkIaZgplbmRzdHJlYW0KZW5kb2JqCjM0IDAgb2JqCjw8IC9MZW5ndGggMTYwIC9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nEWQORIDMQgEc72CJ0hcgvesy7XB+v+pB9ZHoukCNBy6Fk3KehRoPumxRqG60GvoLEqSRMEWkh1Qp2OIOyhITEhjkki2HoMjmlizXZiZVCqzUuG0acXCv9la1chEjXCN/InpBlT8T+pclPBNg6+SMfoYVLw7g4xJ+F5F3Fox7f5EMLEZ9glvRSYFhImxqdm+z2CGzPcK1zjH8w1MgjfrCmVuZHN0cmVhbQplbmRvYmoKMzUgMCBvYmoKPDwgL0xlbmd0aCAxOCAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJwzNrRQMIDDFEOuNAAd5gNSCmVuZHN0cmVhbQplbmRvYmoKMzYgMCBvYmoKPDwgL0xlbmd0aCAxMzMgL0ZpbHRlciAvRmxhdGVEZWNvZGUgPj4Kc3RyZWFtCnicRY9LDgQhCET3nKKOwMcf53Ey6YVz/+2AnW4TYz2FVIG5gqE9LmsDnRUfIRm28beplo5FWT5UelJWD8ngh6zGyyHcoCzwgkkqhiFQi5gakS1lbreA2zYNsrKVU6WOsIujMI/2tGwVHl+iWyJ1kj+DxCov3OO6Hcil1rveoou+f6QBMQkKZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvTGVuZ3RoIDI1MSAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJwtUUlyA0EIu88r9IRmp99jlyuH5P/XCMoHBg2LQHRa4qCMnyAsV7zlkatow98zMYLfBYd+K9dtWORAVCBJY1A1oXbxevQe2HGYCcyT1rAMZqwP/Iwp3OjF4TEZZ7fXZdQQ7F2vPZlByaxcxCUTF0zVYSNnDj+ZMi60cz03IOdGWJdhkG5WGjMSjjSFSCGFqpukzgRBEoyuRo02chT7pS+PdIZVjagx7HMtbV/PTThr0OxYrPLklB5dcS4nFy+sHPT1NgMXUWms8kBIwP1uD/VzspPfeEvnzhbT43vNyfLCVGDFm9duQDbV4t+8iOP7jK/n5/n8A19gW4gKZW5kc3RyZWFtCmVuZG9iagozOCAwIG9iago8PCAvTGVuZ3RoIDIxNSAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJw1UTkOAyEM7PcV/kAkjC94T6Iozf6/zYzRVh7BXIa0lCGZ8lKTqCHlUz56mS6cutzXzGo055a0LXOAuLa8L62SwIlmiIPBaZi4AZo8AUPX0ahRQxce0NSlUyiw3AQ+irduD91jtYGXtiHniSBiKBksQc2pRRMWbc8npDW/Xosb3pft3chTpcaWGIEGAVY4HNfo1/CVPU8m0XQVMtSrNcsYCRNFIjz5jqbVE+taNNIyEtTGEaxqA7w7/TBOAAATccsCZJ9KlLPkxG+x9LMGV/r+AZ9HVJYKZW5kc3RyZWFtCmVuZG9iagoxNyAwIG9iago8PCAvVHlwZSAvRm9udCAvQmFzZUZvbnQgL0JNUVFEVitEZWphVnVTYW5zIC9GaXJzdENoYXIgMCAvTGFzdENoYXIgMjU1Ci9Gb250RGVzY3JpcHRvciAxNiAwIFIgL1N1YnR5cGUgL1R5cGUzIC9OYW1lIC9CTVFRRFYrRGVqYVZ1U2FucwovRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250TWF0cml4IFsgMC4wMDEgMCAwIDAuMDAxIDAgMCBdCi9DaGFyUHJvY3MgMTggMCBSCi9FbmNvZGluZyA8PCAvVHlwZSAvRW5jb2RpbmcKL0RpZmZlcmVuY2VzIFsgMzIgL3NwYWNlIDQ4IC96ZXJvIC9vbmUgL3R3byA1MiAvZm91ciAvZml2ZSA2OCAvRCA3OSAvTyA4MyAvUyA5NyAvYSA5OQovYyAvZCAvZSAxMDMgL2cgMTA1IC9pIDEwOCAvbCAxMTAgL24gMTE0IC9yIDExNiAvdCBdCj4+Ci9XaWR0aHMgMTUgMCBSID4+CmVuZG9iagoxNiAwIG9iago8PCAvVHlwZSAvRm9udERlc2NyaXB0b3IgL0ZvbnROYW1lIC9CTVFRRFYrRGVqYVZ1U2FucyAvRmxhZ3MgMzIKL0ZvbnRCQm94IFsgLTEwMjEgLTQ2MyAxNzk0IDEyMzMgXSAvQXNjZW50IDkyOSAvRGVzY2VudCAtMjM2IC9DYXBIZWlnaHQgMAovWEhlaWdodCAwIC9JdGFsaWNBbmdsZSAwIC9TdGVtViAwIC9NYXhXaWR0aCAxMzQyID4+CmVuZG9iagoxNSAwIG9iagpbIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwCjYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgMzE4IDQwMSA0NjAgODM4IDYzNgo5NTAgNzgwIDI3NSAzOTAgMzkwIDUwMCA4MzggMzE4IDM2MSAzMTggMzM3IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYKNjM2IDYzNiAzMzcgMzM3IDgzOCA4MzggODM4IDUzMSAxMDAwIDY4NCA2ODYgNjk4IDc3MCA2MzIgNTc1IDc3NSA3NTIgMjk1CjI5NSA2NTYgNTU3IDg2MyA3NDggNzg3IDYwMyA3ODcgNjk1IDYzNSA2MTEgNzMyIDY4NCA5ODkgNjg1IDYxMSA2ODUgMzkwIDMzNwozOTAgODM4IDUwMCA1MDAgNjEzIDYzNSA1NTAgNjM1IDYxNSAzNTIgNjM1IDYzNCAyNzggMjc4IDU3OSAyNzggOTc0IDYzNCA2MTIKNjM1IDYzNSA0MTEgNTIxIDM5MiA2MzQgNTkyIDgxOCA1OTIgNTkyIDUyNSA2MzYgMzM3IDYzNiA4MzggNjAwIDYzNiA2MDAgMzE4CjM1MiA1MTggMTAwMCA1MDAgNTAwIDUwMCAxMzQyIDYzNSA0MDAgMTA3MCA2MDAgNjg1IDYwMCA2MDAgMzE4IDMxOCA1MTggNTE4CjU5MCA1MDAgMTAwMCA1MDAgMTAwMCA1MjEgNDAwIDEwMjMgNjAwIDUyNSA2MTEgMzE4IDQwMSA2MzYgNjM2IDYzNiA2MzYgMzM3CjUwMCA1MDAgMTAwMCA0NzEgNjEyIDgzOCAzNjEgMTAwMCA1MDAgNTAwIDgzOCA0MDEgNDAxIDUwMCA2MzYgNjM2IDMxOCA1MDAKNDAxIDQ3MSA2MTIgOTY5IDk2OSA5NjkgNTMxIDY4NCA2ODQgNjg0IDY4NCA2ODQgNjg0IDk3NCA2OTggNjMyIDYzMiA2MzIgNjMyCjI5NSAyOTUgMjk1IDI5NSA3NzUgNzQ4IDc4NyA3ODcgNzg3IDc4NyA3ODcgODM4IDc4NyA3MzIgNzMyIDczMiA3MzIgNjExIDYwNQo2MzAgNjEzIDYxMyA2MTMgNjEzIDYxMyA2MTMgOTgyIDU1MCA2MTUgNjE1IDYxNSA2MTUgMjc4IDI3OCAyNzggMjc4IDYxMiA2MzQKNjEyIDYxMiA2MTIgNjEyIDYxMiA4MzggNjEyIDYzNCA2MzQgNjM0IDYzNCA1OTIgNjM1IDU5MiBdCmVuZG9iagoxOCAwIG9iago8PCAvRCAxOSAwIFIgL08gMjAgMCBSIC9TIDIxIDAgUiAvYSAyMiAwIFIgL2MgMjMgMCBSIC9kIDI0IDAgUiAvZSAyNSAwIFIKL2ZpdmUgMjYgMCBSIC9mb3VyIDI3IDAgUiAvZyAyOCAwIFIgL2kgMjkgMCBSIC9sIDMwIDAgUiAvbiAzMiAwIFIKL29uZSAzMyAwIFIgL3IgMzQgMCBSIC9zcGFjZSAzNSAwIFIgL3QgMzYgMCBSIC90d28gMzcgMCBSIC96ZXJvIDM4IDAgUiA+PgplbmRvYmoKMyAwIG9iago8PCAvRjEgMTcgMCBSID4+CmVuZG9iago0IDAgb2JqCjw8IC9BMSA8PCAvVHlwZSAvRXh0R1N0YXRlIC9DQSAwIC9jYSAxID4+Ci9BMiA8PCAvVHlwZSAvRXh0R1N0YXRlIC9DQSAxIC9jYSAxID4+ID4+CmVuZG9iago1IDAgb2JqCjw8ID4+CmVuZG9iago2IDAgb2JqCjw8ID4+CmVuZG9iago3IDAgb2JqCjw8IC9GMS1EZWphVnVTYW5zLW1pbnVzIDMxIDAgUiAvUDAgMTMgMCBSIC9QMSAxNCAwIFIgPj4KZW5kb2JqCjEzIDAgb2JqCjw8IC9UeXBlIC9YT2JqZWN0IC9TdWJ0eXBlIC9Gb3JtCi9CQm94IFsgLTEuNDc0MzQxNjQ5IC0xLjQ3NDM0MTY0OSAxLjQ3NDM0MTY0OSAxLjQ3NDM0MTY0OSBdIC9MZW5ndGggMTM4Ci9GaWx0ZXIgL0ZsYXRlRGVjb2RlID4+CnN0cmVhbQp4nG2QOw7EIAxEe5+CCwziY8Juu2WukSaKlPu3+SAMK7tB1tjM8zi6g4Jb6X4QPFfOnNxJwcdU6rdOYvCJFy6fV0qcl0fKuXB8lFZsJD3IvDhAXIdp+9IJo90B4tEAM2jY92HtLuVGBh3aBJoEvQ70zvjLBCM0jNvAOCGMS8PkYIq3E/3oAvHqVzwKZW5kc3RyZWFtCmVuZG9iagoxNCAwIG9iago8PCAvVHlwZSAvWE9iamVjdCAvU3VidHlwZSAvRm9ybQovQkJveCBbIC0xLjQ3NDM0MTY0OSAtMS40NzQzNDE2NDkgMS40NzQzNDE2NDkgMS40NzQzNDE2NDkgXSAvTGVuZ3RoIDEzOAovRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJxtkDsOxCAMRHufggsM4mPCbrtlrpEmipT7t/kgDCu7QdbYzPM4uoOCW+l+EDxXzpzcScHHVOq3TmLwiRcun1dKnJdHyrlwfJRWbCQ9yLw4QFyHafvSCaPdAeLRADNo2Pdh7S7lRgYd2gSaBL0O9M74ywQjNIzbwDghjEvD5GCKtxP96ALx6lc8CmVuZHN0cmVhbQplbmRvYmoKMiAwIG9iago8PCAvVHlwZSAvUGFnZXMgL0tpZHMgWyAxMSAwIFIgXSAvQ291bnQgMSA+PgplbmRvYmoKMzkgMCBvYmoKPDwgL0NyZWF0b3IgKE1hdHBsb3RsaWIgdjMuOC4zLCBodHRwczovL21hdHBsb3RsaWIub3JnKQovUHJvZHVjZXIgKE1hdHBsb3RsaWIgcGRmIGJhY2tlbmQgdjMuOC4zKQovQ3JlYXRpb25EYXRlIChEOjIwMjQwMzA2MDA1ODQ4KzAyJzAwJykgPj4KZW5kb2JqCnhyZWYKMCA0MAowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDAwMTYgMDAwMDAgbiAKMDAwMDAxNDE4NCAwMDAwMCBuIAowMDAwMDEzMzM4IDAwMDAwIG4gCjAwMDAwMTMzNzAgMDAwMDAgbiAKMDAwMDAxMzQ2OSAwMDAwMCBuIAowMDAwMDEzNDkwIDAwMDAwIG4gCjAwMDAwMTM1MTEgMDAwMDAgbiAKMDAwMDAwMDA2NSAwMDAwMCBuIAowMDAwMDAwMzQ1IDAwMDAwIG4gCjAwMDAwMDYyODEgMDAwMDAgbiAKMDAwMDAwMDIwOCAwMDAwMCBuIAowMDAwMDA2MjYwIDAwMDAwIG4gCjAwMDAwMTM1ODIgMDAwMDAgbiAKMDAwMDAxMzg4MyAwMDAwMCBuIAowMDAwMDEyMDU2IDAwMDAwIG4gCjAwMDAwMTE4NDkgMDAwMDAgbiAKMDAwMDAxMTQyMCAwMDAwMCBuIAowMDAwMDEzMTA5IDAwMDAwIG4gCjAwMDAwMDYzMDEgMDAwMDAgbiAKMDAwMDAwNjUzOCAwMDAwMCBuIAowMDAwMDA2ODI2IDAwMDAwIG4gCjAwMDAwMDcyNDAgMDAwMDAgbiAKMDAwMDAwNzYyMCAwMDAwMCBuIAowMDAwMDA3OTI1IDAwMDAwIG4gCjAwMDAwMDgyMjkgMDAwMDAgbiAKMDAwMDAwODU1MSAwMDAwMCBuIAowMDAwMDA4ODczIDAwMDAwIG4gCjAwMDAwMDkwMzkgMDAwMDAgbiAKMDAwMDAwOTQ1MyAwMDAwMCBuIAowMDAwMDA5NTk3IDAwMDAwIG4gCjAwMDAwMDk3MTYgMDAwMDAgbiAKMDAwMDAwOTg4OCAwMDAwMCBuIAowMDAwMDEwMTI0IDAwMDAwIG4gCjAwMDAwMTAyNzkgMDAwMDAgbiAKMDAwMDAxMDUxMiAwMDAwMCBuIAowMDAwMDEwNjAyIDAwMDAwIG4gCjAwMDAwMTA4MDggMDAwMDAgbiAKMDAwMDAxMTEzMiAwMDAwMCBuIAowMDAwMDE0MjQ0IDAwMDAwIG4gCnRyYWlsZXIKPDwgL1NpemUgNDAgL1Jvb3QgMSAwIFIgL0luZm8gMzkgMCBSID4+CnN0YXJ0eHJlZgoxNDQwMQolJUVPRgo=", "image/png": "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", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_scaling(scaler=StandardScaler())" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "### Min-max scaling\n", "* Scales all features between a given $min$ and $max$ value (e.g. 0 and 1)\n", "* Makes sense if min/max values have meaning in your data\n", "* Sensitive to outliers\n", "\n", "$$\\mathbf{x}_{new} = \\frac{\\mathbf{x} - x_{min}}{x_{max} - x_{min}} \\cdot (max - min) + min $$" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "tags": [ "hide-input" ] }, "outputs": [ { "data": { "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1R5cGUgL0NhdGFsb2cgL1BhZ2VzIDIgMCBSID4+CmVuZG9iago4IDAgb2JqCjw8IC9Gb250IDMgMCBSIC9YT2JqZWN0IDcgMCBSIC9FeHRHU3RhdGUgNCAwIFIgL1BhdHRlcm4gNSAwIFIKL1NoYWRpbmcgNiAwIFIgL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9UeXBlIC9QYWdlIC9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUgovTWVkaWFCb3ggWyAwIDAgNDE5LjMyNDA2MjUgMTczLjg4NDg3NSBdIC9Db250ZW50cyA5IDAgUiAvQW5ub3RzIDEwIDAgUiA+PgplbmRvYmoKOSAwIG9iago8PCAvTGVuZ3RoIDEyIDAgUiAvRmlsdGVyIC9GbGF0ZURlY29kZSA+PgpzdHJlYW0KeJy9m82uHEdyhff9FL20F7eYkf+5HFkeAQMMbI2I8cLwgpA5NAWShiSP5cf3d6Ka3X2voi5JQ2MQlNjV1VmRmREnTkSeevH16/9++/3rP33z1fkfvju9uH36/ueTnX/g75tzOv/A31/Odv6Gv29O6fyQzu9P1dZWck096/O7+882yjZnnaNxPT3++B+n019OL37HQD8z6jenk6WtdH5WzmPLjGsrb2vmXKtfeHd/wVpnpKmL11/drvnQP56+Pf94/vh1bj6Izbx1m/sfPta1dX7y0+vzv5w/nF/8Lu/WYOr5pzf85wf+/nJOsv38p2+u37Ac+z+t5S21MufKo7SzlbGl0cqofbR+/v79+cU/p/PX/3m9/4Hb6+irDsvr/LC2ltua/LSEd6dtdu7ovY55Zu24l/uXRTenzdpq01Je+lBWrlbnmjW4N285jV7X6KOc+7Zya93WyiUcd6Rig6emyodccu45TcwPDC5bX72NMVrmU90wtvTC3xHcbRs21LzSYils6zhGYY9DizVyLqmnlhp3j61aLjX31iM7ylZar2YpJX0YPU8MzqmFazEn+zaMFTnPreFgQwuS4+2brQ4srmnpU8urzJL4fbx9NbOsOEmfsjllliWxOhbPsJY8C87E/vLb1UaqWGLRrijE2Jcya5MXbvyPgdOKh04b/jgTO1f7uW1MkU+1hCOzE2ZrmnH/+aFto45qpdQajsyCsCcNb2OGrF7BjIpV8a6sPiYrQthhRe69yJNKOHAlbnGFwXp32YQZlko3i/YQH2V3bfaeym5GapjVW+zQltvo0xiSdcT/tKVEYTw/to/R2GZ3UnCjmJXaZuhKCYdszJ9NeQCNFm5EVIUzTFtKAz+bljsxmICRjuuNgz3BFZotwINNMduaVo59yvFyLKzsvXELLo0BeN5KKQosE2BhtTl+bUCGLUJ8Re4vK/oCwXOr/kOQjO2zaC1w5ySszmZgl3+aHZRcoWvgobhFWcRg2kOh17LYwhBmiGjiuWK4jXPFnYFU1mOFdhg5Y6aBFbbHdzEB7ogRyToO2Ynp7Dbh14TKHCEkkUoYjacrlTxYJb6tsn5g68GmtI5H43kC4GZavBltt1B0lFHWIKpwDVBfzr1C1+BmviHnZPc54D8rQ/SDqEqmYDYAmV9OAHhypUZeRASaIhBsrFjBXQAf+ScERryh81yQMHVln