{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "\"Open" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Regressione lineare bayesiana su polinomio con determinazione della distribuzione predittiva" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "%matplotlib inline\n", "import numpy as np\n", "import scipy.stats as st" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import warnings\n", "warnings.filterwarnings('ignore')" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import matplotlib.colors as mcolors\n", "from matplotlib import cm\n", "\n", "plt.style.use('fivethirtyeight')\n", "\n", "plt.rcParams['font.family'] = 'sans-serif'\n", "plt.rcParams['font.serif'] = 'Ubuntu'\n", "plt.rcParams['font.monospace'] = 'Ubuntu Mono'\n", "plt.rcParams['font.size'] = 10\n", "plt.rcParams['axes.labelsize'] = 10\n", "plt.rcParams['axes.labelweight'] = 'bold'\n", "plt.rcParams['axes.titlesize'] = 10\n", "plt.rcParams['xtick.labelsize'] = 8\n", "plt.rcParams['ytick.labelsize'] = 8\n", "plt.rcParams['legend.fontsize'] = 10\n", "plt.rcParams['figure.titlesize'] = 12\n", "plt.rcParams['image.cmap'] = 'jet'\n", "plt.rcParams['image.interpolation'] = 'none'\n", "plt.rcParams['figure.figsize'] = (12, 6)\n", "plt.rcParams['lines.linewidth'] = 2\n", "\n", "\n", "colors = ['xkcd:pale orange', 'xkcd:sea blue', 'xkcd:pale red', 'xkcd:sage green', 'xkcd:terra cotta', 'xkcd:dull purple', 'xkcd:teal', 'xkcd:goldenrod', 'xkcd:cadet blue', \n", " 'xkcd:scarlet']\n", "cmap_big = cm.get_cmap('Spectral', 512)\n", "cmap = mcolors.ListedColormap(cmap_big(np.linspace(0.7, 0.95, 256)))\n", "\n", "bbox_props = dict(boxstyle=\"round,pad=0.3\", fc=colors[0], alpha=.5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "$d$ funzioni base gaussiane, con medie intervallate in modo costante nel dominio considerato e varianza unitaria" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "def vphi(x, d, dom):\n", " l = np.linspace(domain[0], domain[1], d+1)\n", " mus = [(l[i]+l[i+1])/2.0 for i in range(len(l)-1)]\n", " return np.array([gaussian_basis(x, mus[i], 1) for i in range(d)]).T" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Funzione base gaussiana" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "def gaussian_basis(x, m, s):\n", " return np.exp(-((x-m)**2)/(2*s**2))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Genera la matrice delle features e il vettore target" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "# dominio della feature\n", "domain=(0,2*np.pi)\n", "# numero di elementi da generare\n", "n=10\n", "# array delle feature generato uniformemente nel dominio\n", "X=np.random.uniform(domain[0], domain[1], n)\n", "\n", "# genera il vettore target mediante la funzione f e l'aggiunta di rumore gaussiano\n", "# funzione \n", "def f(x):\n", " return np.sin(x)\n", "# sd del rumore\n", "noise = .05\n", "\n", "#genera target\n", "t=np.array([(f(v)+np.random.normal(0,noise,1))[0] for v in X]).reshape(-1,1)\n", "\n", "# numero di funzioni base\n", "d=8\n", "# genera immagine di X per la regressione\n", "Phi = vphi(X,d, domain)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Iperparametri" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "# iperparametro per il prior\n", "alfa=.2\n", "# parametri del prior\n", "mu=np.zeros(d+1)\n", "sigma=np.eye(d+1)*alfa\n", "\n", "# parametro per la verosimiglianza\n", "beta=50" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Distribuzione predittiva dato un valore $v$" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "# media della distribuzione predittiva\n", "def m_pred(v):\n", " return m.T.dot(vphi(v,d,domain))\n", " \n", "# varianza della distribuzione predittiva \n", "def var_pred(v):\n", " v1=vphi(v,d,domain)\n", " return 1.0/beta+v1.dot(s.dot(v1.T))" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "# numero elementi considerati per il training\n", "l=10\n", "# estrazione del training set\n", "X_t, t_t = Phi[:l,:], t[:l]\n", "\n", "# derivazione di media e matrice di covarianza a posteriori\n", "s = np.linalg.inv(np.eye(d)+beta*np.dot(X_t.T,X_t))\n", "m=beta*s.dot(X_t.T.dot(t_t))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Plot della distribuzione predittiva" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# insieme dei valori considerati per il plot\n", "xx=np.linspace(min(X),max(X),500)\n", "# loro immagini per il calcolo della regressione\n", "phix=vphi(xx,d,domain)\n", "\n", "# calcolo di media e varianza della distribuzione predittiva per tutti i valori\n", "mx = np.vectorize(m_pred)(xx)\n", "sx= np.vectorize(var_pred)(xx)\n", "\n", "# visualizzazione\n", "fig = plt.figure(figsize=(16,8))\n", "ax = fig.gca()\n", "# plot della media\n", "ax.plot(xx,mx,'-', c=colors[9], alpha=1)\n", "# riempimento della regione a distanza minore di una sd dalla media\n", "ax.fill_between(xx, mx-np.sqrt(sx), mx+np.sqrt(sx), facecolor=colors[9], alpha=.15)\n", "# elementi dell'insieme\n", "ax.scatter(X[l:], t[l:], c=colors[1], marker='o', alpha=1)\n", "ax.scatter(X[:l], t[:l], c=colors[2], marker='o', alpha=1)\n", "# plot funzione originale\n", "ax.plot(xx,f(xx),'--',c=colors[1],alpha=1)\n", "plt.xlabel(u'$x$', fontsize=10)\n", "plt.ylabel(u'$y$', fontsize=10)\n", "plt.xticks(fontsize=10)\n", "plt.yticks(fontsize=10)\n", "plt.show()\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "anaconda-cloud": {}, "kernelspec": { "display_name": "Python 3.8", "language": "python", "name": "python3.8" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.0" } }, "nbformat": 4, "nbformat_minor": 2 }