|
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433 |
- {
- "cells": [
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Loading data\n",
- "=====================================================\n",
- "We will work with \"Parkinsons Telemonitoring\" dataset of the University of Oxford. The original study used a range of linear regression methods to predict the clinician's Parkinson's disease symptom score on the UPDRS scale\n",
- "\n",
- "We load the dataset \"Parkinsons Telemonitoring\" using the numpy loadtxt function.\n",
- "\n",
- "The columns are separated by ',' delimiter, which we pass to the loadtxt function."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 1,
- "metadata": {},
- "outputs": [],
- "source": [
- "# matplotlib inline\n",
- "\n",
- "import numpy as np\n",
- "import matplotlib.pyplot as plt\n",
- "\n",
- "data = np.loadtxt(\"data/artifical_lin.txt\")"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "We want to work now with 2-dim data, in order to plot it in 3d space. Therefore we select 2 columns (attributes) from total of 12 columns. \n",
- "\n",
- "In this example we will select attributes \"Clinician's motor UPDRS score\" and \"Clinician's total UPDRS score\", which are 4th and 5th columns. "
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 3,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[[0.00747581 0.43208362]\n",
- " [0.49910584 0.20943748]\n",
- " [0.11935362 0.59634898]\n",
- " [0.47691878 0.91091956]\n",
- " [0.73039367 0.88576849]\n",
- " [0.96646013 0.75029941]\n",
- " [0.0254202 0.74285026]\n",
- " [0.17781366 0.59303845]\n",
- " [0.44925923 0.89314114]\n",
- " [0.08370431 0.26143735]]\n",
- "[1.45054918 1.1025327 1.33827336 2.6022192 2.2101526 2.6110778\n",
- " 1.71069895 2.23335293 3.10281928 0.86406978]\n"
- ]
- }
- ],
- "source": [
- "X = data[:, :-1]\n",
- "y = data[:, -1]\n",
- "print(X[:10, :])\n",
- "print(y[:10])"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "We shuffle examples:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 4,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(500, 2)\n",
- "(500,)\n"
- ]
- },
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bushuhui/.virtualenv/fintech/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
- " return f(*args, **kwds)\n",
- "/home/bushuhui/.virtualenv/fintech/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
- " return f(*args, **kwds)\n",
- "/home/bushuhui/.virtualenv/fintech/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
- " return f(*args, **kwds)\n"
- ]
- }
- ],
- "source": [
- "from sklearn.utils import shuffle\n",
- "X, y = shuffle(X, y, random_state=1)\n",
- "print(X.shape)\n",
- "print(y.shape)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Now we split the data into train and test set:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 7,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "train_set_size = 250 \n",
- "(250, 2)\n",
- "(250, 2)\n"
- ]
- }
- ],
- "source": [
- "train_set_size = int(X.shape[0] / 2)\n",
- "print(\"train_set_size = %d \" % train_set_size)\n",
- "\n",
- "X_train = X[:train_set_size, :] # selects first train_set_size rows (examples) for train set\n",
- "X_test = X[train_set_size:, :] # selects from row train_set_size until the last one for test set\n",
- "print(X_train.shape)\n",
- "print(X_test.shape)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "And we split the targets into train and test set in similar way as we splitted data:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 8,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(250,)\n",
- "(250,)\n"
- ]
- }
- ],
- "source": [
- "y_train = y[:train_set_size] # selects first 15 rows (targets) for train set\n",
- "y_test = y[train_set_size:] # selects from row 250 until the last one for test set\n",
- "print(y_train.shape)\n",
- "print(y_test.shape)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Let's look at the data in the 3d plot. There is some linear relationship in the data:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 9,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 432x288 with 1 Axes>"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "from mpl_toolkits.mplot3d import Axes3D\n",
- "fig = plt.figure()\n",
- "ax = Axes3D(fig)\n",
- "ax.scatter3D(X_train[:500, 0], X_train[:500, 1], y_train[:500])\n",
- "ax.view_init(6,-20)\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Linear regression\n",
- "=================\n",
- "Create linear regression object, which we use later to apply linear regression on data"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 10,
- "metadata": {},
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bushuhui/.virtualenv/fintech/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
- " return f(*args, **kwds)\n"
- ]
- }
- ],
- "source": [
- "from sklearn import linear_model\n",
- "regr = linear_model.LinearRegression()\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Fit the model using the training set"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 11,
- "metadata": {},
- "outputs": [],
- "source": [
- "regr.fit(X_train, y_train);"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "We found the coefficients and the bias (the intercept)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 12,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[0.08241879 2.96344602]\n",
- "0.09703581706766884\n"
- ]
- }
- ],
- "source": [
- "print(regr.coef_)\n",
- "print(regr.intercept_)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Now we calculate the mean square error on the test set"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 13,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Training error: 0.1527714636459691\n",
- "Test error: 0.16965042383819598\n"
- ]
- }
- ],
- "source": [
- "# The mean square error\n",
- "print(\"Training error: \", np.mean((regr.predict(X_train) - y_train) ** 2))\n",
- "print(\"Test error: \", np.mean((regr.predict(X_test) - y_test) ** 2))"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Plotting data and linear model\n",
- "==============================\n",
- "Now we want to plot the train data and teachers in 3d plot (marked as dots). \n",
- "\n",
- "With plane we represents the data and predictions (linear model that we found).\n",
- "\n",
- "We first scatter the 3d points using mplot3d:\n"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 14,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 432x288 with 1 Axes>"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "from mpl_toolkits.mplot3d import Axes3D\n",
- "fig = plt.figure()\n",
- "ax = Axes3D(fig)\n",
- "\n",
- "\n",
- "ax.scatter3D(X_train[:500, 0], X_train[:500, 1], y_train[:500]) # plots 3d points, 500 is number of points which are visualized\n",
- "\n",
- "# here we create plane which we want to plot, using the train data and predictions (you don't need to understand it)\n",
- "range_x = np.linspace(X_train[:, 0].min(), X_train[:, 0].max(), num=10)\n",
- "range_y = np.linspace(X_train[:, 1].min(), X_train[:, 1].max(), num=10)\n",
- "xx, yy = np.meshgrid(range_x, range_y)\n",
- "zz = np.vstack([xx.ravel(), yy.ravel()]).T\n",
- "pred = regr.predict(zz)\n",
- "pred = pred.reshape(10, 10)\n",
- "\n",
- "ax.plot_surface(xx, yy, pred, alpha=.1) # plots the plane\n",
- "ax.view_init(6,-20)\n",
- "plt.show()\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Now we plot the data and the plane in similar way for test data:"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 15,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 432x288 with 1 Axes>"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "from mpl_toolkits.mplot3d import Axes3D\n",
- "fig = plt.figure()\n",
- "ax = Axes3D(fig)\n",
- "ax.scatter3D(X_test[:500, 0], X_test[:500, 1], y_test[:500]) # plots 3d points 500 is number of points which are visualized\n",
- "\n",
- "# here we create plane which we want to plot, using the train data and predictions (you don't need to understand it)\n",
- "range_x = np.linspace(X_test[:, 0].min(), X_test[:, 0].max(), num=10)\n",
- "range_y = np.linspace(X_test[:, 1].min(), X_test[:, 1].max(), num=10)\n",
- "xx, yy = np.meshgrid(range_x, range_y)\n",
- "zz = np.vstack([xx.ravel(), yy.ravel()]).T\n",
- "pred = regr.predict(zz)\n",
- "pred = pred.reshape(10, 10)\n",
- "\n",
- "ax.plot_surface(xx, yy, pred, alpha=.1) # plots the plane\n",
- "ax.view_init(6,-20)\n",
- "plt.show()\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Playing with this Notebook\n",
- "==========================\n",
- "\n",
- "Do linear regression on dataset named 'artifical_lin\".\n",
- "\n",
- "Try to see what happens with the error, if you change the sizes of train set and test set. \n",
- "\n",
- "Add noise to the data, and fit the model again. How does the error changes when you add more noise?\n",
- "\n",
- "You add noise using normal distribution, with mean 0 and width 0.4 (you can vary this parameters).\n",
- "noise = np.random.normal(0,0.4, (train_set_size,2))\n",
- "\n",
- "Add noise to data:\n",
- "X = X + noise\n"
- ]
- }
- ],
- "metadata": {
- "kernelspec": {
- "display_name": "Python 3",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython",
- "version": 3
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python",
- "pygments_lexer": "ipython3",
- "version": "3.5.2"
- }
- },
- "nbformat": 4,
- "nbformat_minor": 1
- }
|