Untitled0.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:331b02880126d1eef6cd3625f78eee3e0895f240b3ec87b31c45ede7344260e3"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from IPython import parallel\n",
      "c = parallel.Client()\n",
      "v = c[:]\n",
      "c.clear()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "c.clear()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 41
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%px\n",
      "import numpy as np\n",
      "from sklearn.datasets import load_digits\n",
      "from sklearn.ensemble import RandomForestClassifier\n",
      "\n",
      "def contain_in(x,y):\n",
      "    for u,v in zip(x,y):\n",
      "        if not (u in v):\n",
      "            return False\n",
      "    return True\n",
      "\n",
      "def error(X,cls):\n",
      "    np.random.shuffle(X)\n",
      "    #X_500 = (set(x) for x in cls.apply(X[:500]).T)\n",
      "    X_500 = cls.apply(X[:500])\n",
      "    error = 0\n",
      "    for t in cls.apply(X[500:]):\n",
      "        if contain_in(t,X_500.T):\n",
      "            error+=1\n",
      "    return error"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "from sklearn.datasets import load_digits\n",
      "from sklearn.ensemble import RandomForestClassifier\n",
      "%matplotlib inline\n",
      "import matplotlib.pyplot as plt\n",
      "import pickle\n",
      "\n",
      "def contain_in(x,y):\n",
      "    for u,v in zip(x,y):\n",
      "        if not (u in v):\n",
      "            return False\n",
      "    return True\n",
      "\n",
      "def error(X,cls):\n",
      "    np.random.shuffle(X)\n",
      "    #X_500 = (set(x) for x in cls.apply(X[:500]).T)\n",
      "    X_500 = cls.apply(X[:500])\n",
      "    error = 0\n",
      "    for t in cls.apply(X[500:]):\n",
      "        if contain_in(t,X_500.T):\n",
      "            error+=1\n",
      "    return error"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "error_list = {}"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "with open(\"error3000.pkl\",\"rb\") as f:\n",
      "    error_list = pickle.load(f)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "with open(\"mnist_train.npy\",\"rb\") as f:\n",
      "    digit = np.load(f)\n",
      "X = digit[:,:-1]\n",
      "X = X*256\n",
      "X.astype(np.uint8)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "digit = load_digits()\n",
      "X = digit.data\n",
      "Y = digit.target\n",
      "X.astype(np.uint8)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "array([[ 0,  0,  5, ...,  0,  0,  0],\n",
        "       [ 0,  0,  0, ..., 10,  0,  0],\n",
        "       [ 0,  0,  0, ..., 16,  9,  0],\n",
        "       ..., \n",
        "       [ 0,  0,  1, ...,  6,  0,  0],\n",
        "       [ 0,  0,  2, ..., 12,  0,  0],\n",
        "       [ 0,  0, 10, ..., 12,  1,  0]], dtype=uint8)"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data_max = X.max()\n",
      "data_min = X.min()\n",
      "label_max =30\n",
      "label_min = 0\n",
      "n_data = 4000\n",
      "data_shape = X.shape[1]\n",
      "n_trees = 600\n",
      "\n",
      "cls = RandomForestClassifier(n_estimators=n_trees,\n",
      "                             max_depth=None,\n",
      "                             max_features=1,\n",
      "                             criterion=\"entropy\",\n",
      "                             n_jobs=-1)\n",
      "\n",
      "cls.fit(np.random.randint(data_min,\n",
      "                          data_max,\n",
      "                          (n_data,data_shape)),\n",
      "        np.random.randint(label_min,\n",
      "                          label_max,\n",
      "                          n_data))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 38,
       "text": [
        "RandomForestClassifier(bootstrap=True, compute_importances=None,\n",
        "            criterion='entropy', max_depth=None, max_features=1,\n",
        "            max_leaf_nodes=None, min_density=None, min_samples_leaf=1,\n",
        "            min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
        "            n_estimators=600, n_jobs=-1, oob_score=False,\n",
        "            random_state=None, verbose=0)"
       ]
      }
     ],
     "prompt_number": 38
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "v.push({\"cls\":cls,\"X\":X})"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 39,
       "text": [
        "<AsyncResult: _push>"
       ]
      }
     ],
     "prompt_number": 39
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "r = v.map(lambda x: error(X,cls),range(1000))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 40
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "error = np.array(r.result)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 41
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "error_list[str(n_trees)] = error"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 42
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "with open(\"error4000.pkl\",\"wb\") as f:\n",
      "    pickle.dump(error_list,f)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 43
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "fig = plt.figure(figsize=(7,7))\n",
      "ax = fig.add_subplot(111)\n",
      "labels = list(error_list.keys())\n",
      "labels.sort()\n",
      "for k in labels:\n",
      "    ax.hist(error_list[k]/12.97,alpha=0.3,bins=10,normed=True,label=k)\n",
      "plt.xlabel(\"Error(%)\")\n",
      "plt.ylabel(\"Normalized Frequency\")\n",
      "plt.title(\"Distribution of Error Rate\")\n",
      "plt.legend(loc=0,title=\"Number of Trees\")\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAb4AAAHBCAYAAAAW+d79AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREREREREREREREREREREREREREQk0P8H4773+iF0eEIAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x24aa00a20>"
       ]
      }
     ],
     "prompt_number": 44
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "labels = list(error_list.keys())\n",
      "labels.sort()\n",
      "fig = plt.figure(figsize=(7,7))\n",
      "ax = fig.add_subplot(111)\n",
      "ax.boxplot([error_list[x] for x in labels],widths=0.2,sym=\"\")\n",
      "ax.spines[\"top\"].set_visible(False)\n",
      "ax.spines[\"right\"].set_visible(False)\n",
      "ax.spines[\"bottom\"].set_visible(False)\n",
      "ax.set_xticklabels(labels)\n",
      "ax.xaxis.set_ticks_position(\"none\")\n",
      "ax.yaxis.set_ticks_position(\"none\")\n",
      "ax.set_xlabel(\"Number of trees\")\n",
      "ax.set_ylabel(\"Error\")\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAbMAAAG2CAYAAAD1IGg0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10841c860>"
       ]
      }
     ],
     "prompt_number": 45
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}