Homework2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 課題. MNISTをk-NNで学習せよ\n",
    "注意\n",
    "homework関数を完成させて提出してください\n",
    "\n",
    "訓練データはtrain_X, train_y, テストデータはtest_Xで与えられます\n",
    "\n",
    "train_Xとtrain_yをtrain_X, train_yとvalid_X, valid_yに分けるなどしてモデルを学習させてください\n",
    "test_Xに対して予想ラベルpred_yを作り, homework関数の戻り値としてください\n",
    "pred_yのtest_yに対する精度(F値)で評価します\n",
    "全体の実行時間がiLect上で60分を超えないようにしてください\n",
    "homework関数の外には何も書かないでください\n",
    "k-NNの実装にscikit-learnなどのライブラリを使わないでください\n",
    "次のコードが事前に実行されます"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.utils import shuffle\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.datasets import fetch_mldata\n",
    "from sklearn.cross_validation import train_test_split\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "\n",
    "mnist = fetch_mldata('MNIST original')\n",
    "mnist_X, mnist_y = shuffle(mnist.data.astype('float32'),\n",
    "                           mnist.target.astype('int32'), random_state=42)\n",
    "\n",
    "mnist_X = mnist_X / 255.0\n",
    "\n",
    "train_X, test_X, train_y, test_y = train_test_split(mnist_X, mnist_y,\n",
    "                                                    test_size=0.2,\n",
    "                                                    random_state=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "次のセルのhomework関数を完成させて提出してください\n",
    "上記のコード以外で必要なものは全て書いてください"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 4.58257569  8.77496439]\n",
      "[[ 1.   2.   4. ]\n",
      " [ 2.   2.5  3. ]]\n",
      "(3, 2) (2,)\n",
      "[1 4 3]\n",
      "[[1 2 4]\n",
      " [4 5 6]\n",
      " [3 2 3]]\n",
      "[[1 2 4]\n",
      " [3 2 3]\n",
      " [4 5 6]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([1, 2])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = np.array([[1,2,4],[4,5,6]])\n",
    "print(np.sqrt(np.sum(a**2,axis=1)))\n",
    "b=np.array([1,2])\n",
    "print((a.T/b).T)\n",
    "print(a.T.shape,b.shape)\n",
    "f = np.array([[3,2,3]])\n",
    "e = np.append(np.array(a),f,axis = 0)\n",
    "print(e[:,0])\n",
    "e = [e[i] for i in range(0,len(e))]\n",
    "print(np.array(e))\n",
    "e = sorted(e,key=lambda x:x[0])\n",
    "print(np.array(e))\n",
    "e[0][0:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    }
   ],
   "source": [
    "def mode(arr):\n",
    "    r = [0] * (int(max(arr[1])) + 1)\n",
    "    for a in map(int,arr[1]):\n",
    "        r[a] += 1\n",
    "    if r.count(max(r))==1:\n",
    "        return r.index(max(r))\n",
    "    else:\n",
    "        return arr[0]\n",
    "c =np.array([1.,2.,4.,5.],dtype = np.float64).astype(int)\n",
    "#d =[ c[i] for i in range(0,len(c))].astype(np.float64)\n",
    "\n",
    "#d.append(mode(c))\n",
    "print(len(c))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.9110775   3.        ]\n",
      " [ 0.90592611  3.        ]\n",
      " [ 0.90315121  3.        ]\n",
      " [ 0.88856852  3.        ]\n",
      " [ 0.87627614  3.        ]\n",
      " [ 0.87540203  3.        ]]\n",
      "[[ 0.85760391  2.        ]\n",
      " [ 0.85341579  2.        ]\n",
      " [ 0.82683951  2.        ]\n",
      " [ 0.82155979  2.        ]\n",
      " [ 0.81552023  2.        ]\n",
      " [ 0.80747575  2.        ]]\n",
      "[[ 0.88781905  3.        ]\n",
      " [ 0.87655675  3.        ]\n",
      " [ 0.87463689  3.        ]\n",
      " [ 0.86989015  3.        ]\n",
      " [ 0.86876708  3.        ]\n",
      " [ 0.86644202  3.        ]]\n",
      "[[ 0.93091905  7.        ]\n",
      " [ 0.91680372  7.        ]\n",
      " [ 0.90492934  7.        ]\n",
      " [ 0.90451443  7.        ]\n",
      " [ 0.90029764  7.        ]\n",
      " [ 0.89751941  7.        ]]\n",
      "[[ 0.89994973  4.        ]\n",
      " [ 0.8935734   4.        ]\n",
      " [ 0.87283891  4.        ]\n",
      " [ 0.86992592  4.        ]\n",
      " [ 0.86397767  4.        ]\n",
      " [ 0.86275899  4.        ]]\n",
      "[3 2 3 7 4] [3, 2, 3, 7, 4] (2, 44800)\n"
     ]
    }
   ],
   "source": [
    "def homework(train_X,train_y,test_X):\n",
    "    #最頻値関数の定義\n",
    "    def mode(arr):\n",
    "        r = [0]*10\n",
    "        for a in range(0,len(arr)):\n",
    "            r[int(np.asscalar(arr[a]))] += 1\n",
    "        if r.count(max(r))==1:\n",
    "            return r.index(max(r))\n",
    "        else:\n",
    "            return arr[0].astype(int)\n",
    "    # cross validation をする。　n-fold caross validation　n = 5\n",
    "    for  j in range(0,1):\n",
    "        new_train_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 != j])\n",
    "        val_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 == j])\n",
    "        new_train_y = np.array([train_y[i] for i in range(0,len(train_y) ) if i%5 != j ])\n",
    "        val_y = np.array([train_y[i] for i in range(0,len(train_y)) if i%5 == j])\n",
    "        #パラメーターkのトレーニング\n",
    "        for k in range(11,12):\n",
    "            predicted_list = []\n",
    "            #result = cos(train_X,val_X)\n",
    "            norm1 = np.sqrt(np.sum(new_train_X**2,axis = 1))\n",
    "            norm2 = np.sqrt(np.sum(val_X**2,axis = 1))\n",
    "            result = ((np.dot(val_X,new_train_X.T)/norm1).T/norm2).T\n",
    "            for i in range(0,5):\n",
    "                tmp = np.append(np.array([result[i]]),np.array([new_train_y.T]),axis = 0).T\n",
    "                tmp = [tmp[i] for i in range(0,len(tmp))]##sort のために　list化\n",
    "                tmp = np.array(sorted(tmp,key=lambda x:x[0],reverse = True)).T\n",
    "                predicted_list.append(mode(tmp[1,0:k]))\n",
    "                print(tmp.T[0:6])\n",
    "    print(val_y[0:5],predicted_list,tmp.shape)\n",
    "    return fi_score(val_y,predicted_list)\n",
    "        \n",
    "homework(train_X,train_y,test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1, 2], [3, 5], [2, 3]]\n"
     ]
    }
   ],
   "source": [
    "aa=[[1,2],[3,5],[2,3]]\n",
    "print(aa[0:len(aa)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "1 == 1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.9110775   3.        ]\n",
      " [ 0.90592611  3.        ]\n",
      " [ 0.90315121  3.        ]\n",
      " [ 0.88856852  3.        ]\n",
      " [ 0.87627614  3.        ]\n",
      " [ 0.87540203  3.        ]]\n",
      "[[ 0.85760391  2.        ]\n",
      " [ 0.85341579  2.        ]\n",
      " [ 0.82683951  2.        ]\n",
      " [ 0.82155979  2.        ]\n",
      " [ 0.81552023  2.        ]\n",
      " [ 0.80747575  2.        ]]\n",
      "[[ 0.88781905  3.        ]\n",
      " [ 0.87655675  3.        ]\n",
      " [ 0.87463689  3.        ]\n",
      " [ 0.86989015  3.        ]\n",
      " [ 0.86876708  3.        ]\n",
      " [ 0.86644202  3.        ]]\n",
      "[[ 0.93091905  7.        ]\n",
      " [ 0.91680372  7.        ]\n",
      " [ 0.90492934  7.        ]\n",
      " [ 0.90451443  7.        ]\n",
      " [ 0.90029764  7.        ]\n",
      " [ 0.89751941  7.        ]]\n",
      "[[ 0.89994973  4.        ]\n",
      " [ 0.8935734   4.        ]\n",
      " [ 0.87283891  4.        ]\n",
      " [ 0.86992592  4.        ]\n",
      " [ 0.86397767  4.        ]\n",
      " [ 0.86275899  4.        ]]\n",
      "[3 2 3 7 4] [] (2, 44800)\n"
     ]
    }
   ],
   "source": [
    "def homework(train_X,train_y,test_X):\n",
    "    #f_scoreを格納するリスト\n",
    "    f = []\n",
    "    #最頻値関数の定義\n",
    "    def mode(arr):\n",
    "        r = [0]*10\n",
    "        for a in range(0,len(arr)):\n",
    "            r[int(np.asscalar(arr[a]))] += 1\n",
    "        if r.count(max(r))==1:\n",
    "            return r.index(max(r))\n",
    "        else:\n",
    "            return arr[0].astype(int)\n",
    "    # cross validation をする。　n-fold caross validation　n = 5\n",
    "    for  j in range(0,5):\n",
    "        new_train_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 != j])\n",
    "        val_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 == j])\n",
    "        new_train_y = np.array([train_y[i] for i in range(0,len(train_y) ) if i%5 != j ])\n",
    "        val_y = np.array([train_y[i] for i in range(0,len(train_y)) if i%5 == j])\n",
    "        #パラメーターkのトレーニング\n",
    "        for k in range(11,12):\n",
    "            predicted_list = []\n",
    "            #result = cos(train_X,val_X)\n",
    "            norm1 = np.sqrt(np.sum(new_train_X**2,axis = 1))\n",
    "            norm2 = np.sqrt(np.sum(val_X**2,axis = 1))\n",
    "            result = ((np.dot(val_X,new_train_X.T)/norm1).T/norm2).T\n",
    "            for i in range(0,len(result)):\n",
    "                tmp = np.append(np.array([result[i]]),np.array([new_train_y.T]),axis = 0).T\n",
    "                tmp = [tmp[i] for i in range(0,len(tmp))]##sort のために　list化\n",
    "                tmp = np.array(sorted(tmp,key=lambda x:x[0],reverse = True)).T\n",
    "                predicted_list.append(mode(tmp[1,0:k]))\n",
    "                print(tmp.T[0:6])\n",
    "                f.append(fi_score(val_y,predicted_list))\n",
    "    print(val_y[0:5],predicted_list,tmp.shape)\n",
    "    return \n",
    "        \n",
    "homework(train_X,train_y,test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.9110775   3.        ]\n",
      " [ 0.90592611  3.        ]\n",
      " [ 0.90315121  3.        ]\n",
      " [ 0.88856852  3.        ]\n",
      " [ 0.87627614  3.        ]\n",
      " [ 0.87540203  3.        ]]\n",
      "[[ 0.85760391  2.        ]\n",
      " [ 0.85341579  2.        ]\n",
      " [ 0.82683951  2.        ]\n",
      " [ 0.82155979  2.        ]\n",
      " [ 0.81552023  2.        ]\n",
      " [ 0.80747575  2.        ]]\n",
      "[[ 0.88781905  3.        ]\n",
      " [ 0.87655675  3.        ]\n",
      " [ 0.87463689  3.        ]\n",
      " [ 0.86989015  3.        ]\n",
      " [ 0.86876708  3.        ]\n",
      " [ 0.86644202  3.        ]]\n",
      "[[ 0.93091905  7.        ]\n",
      " [ 0.91680372  7.        ]\n",
      " [ 0.90492934  7.        ]\n",
      " [ 0.90451443  7.        ]\n",
      " [ 0.90029764  7.        ]\n",
      " [ 0.89751941  7.        ]]\n",
      "[[ 0.89994973  4.        ]\n",
      " [ 0.8935734   4.        ]\n",
      " [ 0.87283891  4.        ]\n",
      " [ 0.86992592  4.        ]\n",
      " [ 0.86397767  4.        ]\n",
      " [ 0.86275899  4.        ]]\n",
      "[3 2 3 7 4] [3, 2, 3, 7, 4] (2, 44800)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/iwamotoshinichi/anaconda/lib/python3.5/site-packages/sklearn/metrics/classification.py:756: DeprecationWarning: The default `weighted` averaging is deprecated, and from version 0.18, use of precision, recall or F-score with multiclass or multilabel data or pos_label=None will result in an exception. Please set an explicit value for `average`, one of (None, 'micro', 'macro', 'weighted', 'samples'). In cross validation use, for instance, scoring=\"f1_weighted\" instead of scoring=\"f1\".\n",
      "  sample_weight=sample_weight)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def homework(train_X,train_y,test_X):\n",
    "    #最頻値関数の定義\n",
    "    def mode(arr):\n",
    "        r = [0]*10\n",
    "        for a in range(0,len(arr)):\n",
    "            r[int(np.asscalar(arr[a]))] += 1\n",
    "        if r.count(max(r))==1:\n",
    "            return r.index(max(r))\n",
    "        else:\n",
    "            return arr[0].astype(int)\n",
    "    # cross validation をする。　n-fold caross validation　n = 5\n",
    "    for  j in range(0,1):\n",
    "        new_train_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 != j])\n",
    "        val_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 == j])\n",
    "        new_train_y = np.array([train_y[i] for i in range(0,len(train_y) ) if i%5 != j ])\n",
    "        val_y = np.array([train_y[i] for i in range(0,len(train_y)) if i%5 == j])\n",
    "        #パラメーターkのトレーニング\n",
    "        for k in range(11,12):\n",
    "            predicted_list = []\n",
    "            #result = cos(train_X,val_X)\n",
    "            norm1 = np.sqrt(np.sum(new_train_X**2,axis = 1))\n",
    "            norm2 = np.sqrt(np.sum(val_X**2,axis = 1))\n",
    "            result = ((np.dot(val_X,new_train_X.T)/norm1).T/norm2).T\n",
    "            for i in range(0,5):\n",
    "                tmp = np.append(np.array([result[i]]),np.array([new_train_y.T]),axis = 0).T\n",
    "                tmp = [tmp[i] for i in range(0,len(tmp))]##sort のために　list化\n",
    "                tmp = np.array(sorted(tmp,key=lambda x:x[0],reverse = True)).T\n",
    "                predicted_list.append(mode(tmp[1,0:k]))\n",
    "                print(tmp.T[0:6])\n",
    "    print(val_y[0:5],predicted_list,tmp.shape)\n",
    "    return f1_score(val_y[0:5],np.array(predicted_list))\n",
    "        \n",
    "homework(train_X,train_y,test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/iwamotoshinichi/anaconda/lib/python3.5/site-packages/sklearn/metrics/classification.py:756: DeprecationWarning: The default `weighted` averaging is deprecated, and from version 0.18, use of precision, recall or F-score with multiclass or multilabel data or pos_label=None will result in an exception. Please set an explicit value for `average`, one of (None, 'micro', 'macro', 'weighted', 'samples'). In cross validation use, for instance, scoring=\"f1_weighted\" instead of scoring=\"f1\".\n",
      "  sample_weight=sample_weight)\n",
      "/Users/iwamotoshinichi/anaconda/lib/python3.5/site-packages/sklearn/metrics/classification.py:1074: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/iwamotoshinichi/anaconda/lib/python3.5/site-packages/sklearn/metrics/classification.py:1076: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no true samples.\n",
      "  'recall', 'true', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.66666666666666663"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(np.array([1,2,3]),np.array([1,23,3]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "Here!!\n",
      "[0 3 9 ..., 9 8 3]\n",
      "0.974082549127 [0 3 9 ..., 9 8 3] 0.974071428571\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/iwamotoshinichi/anaconda/lib/python3.5/site-packages/sklearn/metrics/classification.py:756: DeprecationWarning: The default `weighted` averaging is deprecated, and from version 0.18, use of precision, recall or F-score with multiclass or multilabel data or pos_label=None will result in an exception. Please set an explicit value for `average`, one of (None, 'micro', 'macro', 'weighted', 'samples'). In cross validation use, for instance, scoring=\"f1_weighted\" instead of scoring=\"f1\".\n",
      "  sample_weight=sample_weight)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.utils import shuffle\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.datasets import fetch_mldata\n",
    "from sklearn.cross_validation import train_test_split\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "\n",
    "mnist = fetch_mldata('MNIST original')\n",
    "mnist_X, mnist_y = shuffle(mnist.data.astype('float32'),\n",
    "                           mnist.target.astype('int32'), random_state=42)\n",
    "\n",
    "mnist_X = mnist_X / 255.0\n",
    "\n",
    "train_X, test_X, train_y, test_y = train_test_split(mnist_X, mnist_y,\n",
    "                                                    test_size=0.2,\n",
    "                                                    random_state=None)\n",
    "\n",
    "\n",
    "def homework(train_X,train_y,test_X):\n",
    "    pred_y =[]\n",
    "    #メモリー削減のため手動でガベージコレクション\n",
    "    import gc\n",
    "    #最頻値関数の定義\n",
    "    def mode(arr):\n",
    "        r = [0]*10\n",
    "        for a in range(0,len(arr)):\n",
    "            r[int(np.asscalar(arr[a]))] += 1\n",
    "        if r.count(max(r))== 1:\n",
    "            return r.index(max(r))\n",
    "        else:\n",
    "            return arr[0].astype(int)\n",
    "    \n",
    "    # 予想操作を定義する\n",
    "    def basic_operation(new_train_X,val_X,new_train_y,k):\n",
    "        predicted_list = []\n",
    "        #result = cos(train_X,val_X)\n",
    "        norm1 = np.sqrt(np.sum(new_train_X**2,axis = 1))\n",
    "        norm2 = np.sqrt(np.sum(val_X**2,axis = 1))\n",
    "        result = ((np.dot(val_X,new_train_X.T)/norm1).T/norm2).T\n",
    "        #i番目のvalidation_exampleをtmpに格納してsort\n",
    "        for i in range(0,len(result)):\n",
    "            tmp = np.append(np.array([result[i]]),np.array([new_train_y.T]),axis = 0).T\n",
    "            tmp = [tmp[i] for i in range(0,len(tmp))]      #sort のために　list化\n",
    "            tmp = np.array(sorted(tmp,key=lambda x:x[0],reverse = True)).T   \n",
    "            predicted_list.append(mode(tmp[1,0:k]))\n",
    "        return predicted_list\n",
    "    #f値,kを初期化\n",
    "    f = 0 #f値\n",
    "    k = 9 #k-nnのk\n",
    "    #f > 0.92まで　k をトレーニング(k1 = 9)\n",
    "    while f < 0.92:\n",
    "        f = 0\n",
    "        k += 2\n",
    "        # cross validation をする。　n-fold caross validation　n = 5 \n",
    "        for  j in range(0,5):\n",
    "            new_train_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%5 == (j+1)%5 ])\n",
    "            val_X = np.array([train_X[i] for i in range(0,len(train_X)) if i%10 == j])\n",
    "            new_train_y = np.array([train_y[i] for i in range(0,len(train_y) ) if i%5 == (j+1)%5])\n",
    "            val_y = np.array([train_y[i] for i in range(0,len(train_y)) if i%10 == j])\n",
    "            #print('Here!')\n",
    "            predicted_list = basic_operation(new_train_X,val_X,new_train_y,k)\n",
    "            f += f1_score(val_y,np.array(predicted_list))\n",
    "        f =f/5.0\n",
    "        print(f)\n",
    "    #不要になった val_y, predicted_list 等を消去\n",
    "    del val_y,predicted_list,train_X,train_y,val_X\n",
    "    gc.collect()\n",
    "    for i in range(0,50):\n",
    "        n = len(test_X)\n",
    "        new_test_X = test_X[int(n*i/50):int((i+1)*n/50)]\n",
    "        #print('Here!')\n",
    "        pred_y[int(n*i/50):int((i+1)*n/50)] = basic_operation(train_X,new_test_X,train_y,k)\n",
    "        #print('Here!!')\n",
    "    return pred_y\n",
    "\n",
    "y = np.array(homework(train_X,train_y,test_X))\n",
    "print(y)\n",
    "print(f1_score(y,test_y),y,np.sum((y == test_y)/len(test_y)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnMAAAFSCAYAAACZqFkAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1096e8cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "fig = plt.figure(figsize=(9, 9))\n",
    "fig.subplots_adjust(left=0, right=1, bottom=0,\n",
    "                    top=0.5, hspace=0.05, wspace=0.05)\n",
    "\n",
    "for i in range(81):\n",
    "    ax = fig.add_subplot(9, 9, i + 1, xticks=[], yticks=[])\n",
    "    ax.imshow(train_X[i].reshape((28, 28)), cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "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": 0
}