digit.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_svmlight_file\n",
    "from sklearn.model_selection import train_test_split\n",
    "import numpy as np\n",
    "from okl_fct import okl, lrokl, mse, predict\n",
    "import matplotlib.pyplot as plt\n",
    "from numpy import linalg as LA\n",
    "%matplotlib inline\n",
    "from sklearn.metrics.pairwise import rbf_kernel\n",
    "from sklearn.metrics import accuracy_score\n",
    "import seaborn as sns\n",
    "sns.set(font_scale=1.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X, y = load_svmlight_file('data/usps.libsvm')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#construct multi-class labels.\n",
    "Y = np.zeros((len(y),len(np.unique(y))))\n",
    "for i, y0 in enumerate(y):\n",
    "    Y[i,int(y0-1)]=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<9298x256 sparse matrix of type '<class 'numpy.float64'>'\n",
       "\twith 2301812 stored elements in Compressed Sparse Row format>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.2, random_state=42)\n",
    "K_train = rbf_kernel(X_train,gamma=None)\n",
    "K_test = rbf_kernel(X_test,X_train,gamma=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.save('digit_K_train', K_train)\n",
    "np.save('digit_K_test', K_test)\n",
    "np.save('digit_y_train', y_train)\n",
    "np.save('digit_y_test', y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "p_list = np.array(range(10))+1\n",
    "N = 20\n",
    "lam_list = np.logspace(-5,0,N) * np.sqrt(LA.norm(np.dot(np.dot(y_train.T,K_train),y_train)))\n",
    "lam_list = np.sort(lam_list)[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  1.93702093e+03   1.05676013e+03   5.76525503e+02   3.14528952e+02\n",
      "   1.71594251e+02   9.36148701e+01   5.10724797e+01   2.78630754e+01\n",
      "   1.52009648e+01   8.29303045e+00   4.52434136e+00   2.46829731e+00\n",
      "   1.34660299e+00   7.34652027e-01   4.00796376e-01   2.18658262e-01\n",
      "   1.19291087e-01   6.50803833e-02   3.55052199e-02   1.93702093e-02]\n"
     ]
    }
   ],
   "source": [
    "print(lam_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lambda = 1937.02092653\n",
      "lambda = 1056.76012578\n",
      "lambda = 576.525502715\n",
      "lambda = 314.528952382\n",
      "lambda = 171.594251114\n",
      "lambda = 93.6148700858\n",
      "lambda = 51.072479668\n",
      "lambda = 27.8630753538\n",
      "lambda = 15.2009648487\n",
      "lambda = 8.29303044976\n",
      "lambda = 4.52434136418\n",
      "lambda = 2.46829731346\n",
      "lambda = 1.34660299417\n",
      "lambda = 0.73465202673\n",
      "lambda = 0.4007963763\n",
      "lambda = 0.218658261885\n",
      "lambda = 0.11929108724\n",
      "lambda = 0.0650803832987\n",
      "lambda = 0.0355052199481\n",
      "lambda = 0.0193702092653\n"
     ]
    }
   ],
   "source": [
    "model = okl(K_train,y_train,lam_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pickle\n",
    "with open('digit_okl.pickle', 'wb') as handle:\n",
    "    pickle.dump(model, handle, protocol=pickle.HIGHEST_PROTOCOL)\n",
    "#with open('digit_okl.pickle', 'rb') as handle:\n",
    "#    b = pickle.load(handle)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For OKL, when lambda = 0.065,\n",
      "test accuracy reaches the maximum = 0.981\n"
     ]
    }
   ],
   "source": [
    "OKL_acc = np.zeros((N,1))\n",
    "OKL_time = np.zeros((N,1))\n",
    "for k in range(N):\n",
    "    OKL_time[k] = model[k]['time']\n",
    "    pred_test = predict(K_test,model[k]['C'],model[k]['L'])\n",
    "    pred_test_clf = np.zeros(pred_test.shape)\n",
    "    for i in range(len(pred_test)):\n",
    "        n = np.argmax(pred_test[i])\n",
    "        pred_test_clf[i,n]=1\n",
    "    OKL_acc[k] = accuracy_score(y_test,pred_test_clf)\n",
    "jj = np.argmax(OKL_acc)\n",
    "print('For OKL, when lambda = %.3f,\\ntest accuracy reaches the maximum = %.3f' %(lam_list[jj],OKL_acc[jj]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#model2 = {}\n",
    "#for it,p in enumerate(p_list):\n",
    "#    print('p = '+str(p))  \n",
    "#    modelp = lrokl(K_train,y_train,lam_list,p)\n",
    "#    model2[it] = modelp\n",
    "\n",
    "#with open('digit_lrokl.pickle', 'wb') as handle:\n",
    "#    pickle.dump(model2, handle, protocol=pickle.HIGHEST_PROTOCOL)\n",
    "##with open('digit_okl.pickle', 'rb') as handle:\n",
    "##    b = pickle.load(handle)\n",
    "\n",
    "#P = len(p_list)\n",
    "#LROKL_acc= np.zeros((P,N))\n",
    "#LR_time = np.zeros((P,N))\n",
    "#for k in range(N):\n",
    "#    for j in range(P):\n",
    "#        LR_time[j][k] = model2[j][k]['time']\n",
    "#        pred_test = predict(K_test,model2[j][k]['A'],model2[j][k]['B'].T)\n",
    "#        pred_test_clf = np.zeros(pred_test.shape)\n",
    "#        for i in range(len(pred_test)):\n",
    "#            n = np.argmax(pred_test[i])\n",
    "#        pred_test_clf[i,n]=1\n",
    "#        LROKL_acc[j][k] =accuracy_score(y_test,pred_test_clf)\n",
    "#kk,hh = np.unravel_index(LROKL_acc.argmax(), LROKL_acc.shape)\n",
    "#print('For low rank OKL, when p = %i, lambda = %.3f \\ntest accuracy reaches the maximum = %.3f' %(p_list[kk],lam_list[hh],LROKL_acc[kk][hh]))\n",
    "\n",
    "#p_list_2=[1,3,5,7,9]\n",
    "#for i,p in enumerate(p_list_2):\n",
    "#    plt.semilogx(lam_list, LR_time[p-1],label='p=%i'%p)\n",
    "#plt.title('low rank OKL execution time')\n",
    "#plt.grid(True)\n",
    "#plt.xlabel(\"lambda\")\n",
    "#plt.ylabel(\"second\")\n",
    "#plt.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "#plt.plot(p_list, LR_time[:,hh],label='Low rank OKL (lambda=%.3f)'%lam_list[hh])\n",
    "#plt.title('Low rank OKL execution time')\n",
    "#plt.grid(True)\n",
    "#plt.xlabel(\"p\")\n",
    "#plt.ylabel(\"second\")\n",
    "#plt.legend(loc=0)\n",
    "\n",
    "#plt.plot(p_list, LROKL_arr[:,hh],label='Low rank OKL (lambda=%.3f)'%lam_list[hh],color=c[0])\n",
    "#plt.title('Low rank OKL : test accuracy in function of rank')\n",
    "#plt.axvline(x=p_list[kk],color=c[0],linestyle='--')\n",
    "#plt.grid(True)\n",
    "#plt.xlabel(\"p\")\n",
    "#plt.ylabel(\"accuracy\")\n",
    "#plt.legend(loc=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x268842e19e8>"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAswAAAIUCAYAAAAOvHpYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2688411da90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.semilogx(lam_list, OKL_time,label='OKL')\n",
    "plt.title('OKL execution time')\n",
    "plt.grid(True)\n",
    "plt.xlabel(\"lambda\")\n",
    "plt.ylabel(\"second\")\n",
    "plt.legend(loc=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "c=sns.color_palette()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x26883b4c6a0>"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtIAAAIUCAYAAAA328sDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x268956d9a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.semilogx(lam_list, OKL_acc,label='OKL',color=c[0])\n",
    "#plt.semilogx(lam_list, LROKL_acc[kk],label='Low rank OKL (p=%i)'%p_list[kk],color=c[2])\n",
    "plt.axvline(x=lam_list[jj],color=c[0],linestyle='--')\n",
    "#plt.axvline(x=lam_list[hh],color=c[2],linestyle='--')\n",
    "plt.title('OKL test accuracy')\n",
    "plt.grid(True)\n",
    "plt.xlabel(\"lambda\")\n",
    "plt.ylabel(\"accuracy\")\n",
    "plt.legend(loc=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import random\n",
    "import networkx as nx\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def visualize(ker, label, threshold=0.6):\n",
    "    \n",
    "    \"\"\"\n",
    "    L is the output kernel\n",
    "    labels come form caltech_101_classes.txt\n",
    "    \"\"\"\n",
    "    G = nx.Graph()\n",
    "    \n",
    "    #add nodes\n",
    "    G.add_nodes_from(list(range(len(ker))))\n",
    "\n",
    "    #add edges if similarity > 0.5\n",
    "    for i in range(len(ker)):\n",
    "        for j in range(i+1,len(ker)):\n",
    "            if ker[i][j] > threshold:\n",
    "                G.add_edges_from([(i,j,{'weight':ker[i][j]})])\n",
    "    pos = nx.spring_layout(G)\n",
    "\n",
    "    #draw the graph\n",
    "    plt.figure(figsize=(10,10))\n",
    "    nx.draw_networkx_nodes(G,pos,node_color='grey',\n",
    "                          node_size=500,alpha=0.8)\n",
    "\n",
    "    nx.draw_networkx_edges(G,pos,width=1.0,alpha=0.5)\n",
    "    nx.draw_networkx_labels(G,pos,label,font_size=16)\n",
    "    plt.axis('off')\n",
    "    plt.savefig('full.png')\n",
    "\n",
    "    #show the figure\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1wAAANBCAYAAAD0rb+BAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJBFcAAAASf4BzCNTGFwMMfkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x295a6e63f28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ker = model[jj]['L']\n",
    "#ker2 = model2[kk][hh]['']\n",
    "label = dict(zip(list(range(len(ker))),[0,1,2,3,4,5,6,7,8,9]))\n",
    "visualize(ker, label, threshold=-0.009)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "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
}