l1 lasso LR.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function, division\n",
    "from builtins import range\n",
    "\\\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 50\n",
    "D =50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=(np.random.random((N,D))-0.5)*10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50, 50)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "true_w = np.array([1,0.5,-0.5]+[0]*(D-3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1. ,  0.5, -0.5,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,\n",
       "        0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,\n",
       "        0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,\n",
       "        0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,  0. ,\n",
       "        0. ,  0. ,  0. ,  0. ,  0. ,  0. ])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "true_w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y=X.dot(true_w)+np.random.randn(N)*50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50,)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "C=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "w = np.random.randn(D)/np.sqrt(D)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.03924776, -0.18822971,  0.10805179,  0.00562476, -0.33135933,\n",
       "       -0.08065469,  0.09069553,  0.14620555, -0.05806507,  0.06076208,\n",
       "        0.0256365 , -0.04760616,  0.05512871,  0.05837481, -0.06039311,\n",
       "       -0.3589947 ,  0.01830412,  0.25912335, -0.22481732,  0.15679365,\n",
       "        0.23661724,  0.30055188, -0.10725298,  0.09194437, -0.02547106,\n",
       "       -0.16662468,  0.37287333, -0.01654469,  0.08608063, -0.11172672,\n",
       "       -0.11580545, -0.02563489,  0.05156434, -0.20141678, -0.08521713,\n",
       "        0.07306002,  0.16976859, -0.13380775,  0.04747764,  0.02865973,\n",
       "       -0.11325935, -0.06481549,  0.1506454 ,  0.06412434, -0.15682763,\n",
       "       -0.00051449, -0.19918255,  0.29243391, -0.17712581,  0.01836169])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr=0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "l1=20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "for t in range(500):\n",
    "    Yhat = X.dot(w)\n",
    "    delta = Yhat - Y\n",
    "    w = w- lr*(X.T.dot(delta)+l1*np.sign(w))\n",
    "    \n",
    "    mse = delta.dot(delta)/N\n",
    "    C.append(mse)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x10f669780>]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(C)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 5.71346726e+00, -4.18462123e-01,  5.16917836e+00,  9.25020627e+00,\n",
       "       -4.74745608e+00, -4.79696338e+00, -9.13893477e+00,  5.40420846e-01,\n",
       "        5.19586681e+00,  1.14105652e+01,  3.09301362e-01,  4.90803328e+00,\n",
       "        1.25682469e+00,  6.45228646e-03, -3.83183469e+00, -1.22828837e+01,\n",
       "       -7.29976871e+00,  7.34434689e+00,  8.56822944e+00,  1.17063200e+01,\n",
       "       -3.22445970e+00, -9.31533215e+00, -2.80387239e+00, -4.45707891e+00,\n",
       "        1.19164819e+01, -5.56030751e+00, -3.16517518e+00, -6.37341345e+00,\n",
       "       -1.01350217e+01, -1.89590897e+00, -2.49981540e+00,  4.88295808e+00,\n",
       "       -3.84915504e+00,  4.12884283e+00,  1.06843999e+01,  4.78594757e+00,\n",
       "        1.97272065e+00,  4.97880430e+00,  3.80780064e+00,  5.08205864e+00,\n",
       "        6.31411183e+00, -2.45733355e+00,  5.57652853e+00, -8.41658815e+00,\n",
       "        5.24441884e+00, -1.11409589e-01, -8.28179078e+00,  2.79933039e+00,\n",
       "       -8.33870917e+00, -1.05088671e+01])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x10f6c9940>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(true_w, label = 'true')\n",
    "plt.plot(w, label = 'predicted')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}