Read MSRAction3D Data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import necessary modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from norm import *\n",
    "from utils import *\n",
    "from MostJoints.getMostInformativeJoints import *\n",
    "import os\n",
    "from itertools import chain\n",
    "import setting\n",
    "\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.svm import LinearSVC\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make skeleton coordinators matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_PATH = 'data/'\n",
    "ACTION_SETS_PATH = 'actionsets/'\n",
    "ACTION_SETS  = ['actionsets/ActionSet1.txt', 'actionsets/ActionSet2.txt', 'actionsets/ActionSet3.txt']\n",
    "ACTION_NAMES = ['Wave','Hammer','Smash','Catch','Forward Punch','Throw','Draw X','Draw Tick','Draw ',\n",
    "               'Circle','Clapping Hand','Two hand Wave','Side Boxing','Bend','Forward Kick','Side Kick','Jogging',\n",
    "               'Tennis Swing','Tennis Serve','Golf Swing','Pickup&throw']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of video files:  564\n"
     ]
    }
   ],
   "source": [
    "TRAINING_SUBJECTS = ['s01','s03','s05','s07','s09']\n",
    "TESTING_SUBJECTS = ['s02','s04','s06','s08','s10']\n",
    "data_file_names = os.listdir(DATA_PATH)\n",
    "print(\"Number of video files: \", len(data_file_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data_file_names = [data_file_names[i] for i in range(len(data_file_names)) if data_file_names[i][4:7] in TRAINING_SUBJECTS]\n",
    "test_data_file_names = [data_file_names[i] for i in range(len(data_file_names)) if data_file_names[i][4:7] in TESTING_SUBJECTS]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of training videos:  291\n",
      "Number of testing videos:  270\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of training videos: \", len(train_data_file_names))\n",
    "print(\"Number of testing videos: \", len(test_data_file_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_labels = ['training videos', 'testing videos']\n",
    "y_labesl = [len(train_data_file_names), len(test_data_file_names)]\n",
    "plt.bar(x_labels, y_labesl)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get most informative joints of each action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_most_informative_joints(filenames):\n",
    "    # Notice that filenames shoube be sorted before fitting into this method\n",
    "    # For training phase\n",
    "    noFramesList = []\n",
    "    d1 = {'a0' + str(i) : [] for i in range(1,10)}\n",
    "    d2 = {'a' + str(i) : [] for i in range(10, 21)}\n",
    "    mostInformativeJointsList = dict(chain.from_iterable(d.items() for d in (d1,d2)))\n",
    "    for i in range(len(filenames)):\n",
    "        with open('data/' + filenames[i], 'r') as f:\n",
    "            skeleton_i_th = [line.rstrip('\\n') for line in f]\n",
    "        for j in range(len(skeleton_i_th)):\n",
    "            skeleton_i_th[j] = [float(coord) for coord in skeleton_i_th[j].split(' ')]\n",
    "        skeleton_i_th = np.asarray(skeleton_i_th)\n",
    "        noFrames = int(skeleton_i_th.shape[0]/setting.NUMBER_OF_JOINTS)\n",
    "        x = skeleton_i_th[:,0]\n",
    "        y = skeleton_i_th[:,1]\n",
    "        z = skeleton_i_th[:,2]\n",
    "        x = x.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_i_th.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "        y = y.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_i_th.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "        z = z.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_i_th.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "        t = np.arange(1, noFrames+1)\n",
    "        noFramesList.append(noFrames)\n",
    "        listIdxMostJoints = getMostJoints(x.T,y.T,z.T,1,setting.NUMBER_OF_JOINTS)\n",
    "        mostInformativeJointsList[filenames[i][0:3]].append(listIdxMostJoints)\n",
    "    return noFramesList, mostInformativeJointsList"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# sorted list of filenames \n",
    "train_data_file_names = sorted(train_data_file_names)\n",
    "test_data_file_names = sorted(test_data_file_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_noFrameList, train_mostInformativeJointsList = get_most_informative_joints(train_data_file_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create histograms to present apperance of joints for each action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_histograms(action_label):\n",
    "    action = train_mostInformativeJointsList[action_label]\n",
    "    Ns_4 = [0 for i in range(20)]\n",
    "    Ns_8 = [0 for i in range(20)]\n",
    "    Ns_12 = [0 for i in range(20)]\n",
    "    \n",
    "    for i in range(len(action)):\n",
    "        for j in range(0, 12):\n",
    "            if j < 4:\n",
    "                Ns_4[action[i][j]-1] += 1\n",
    "            if j < 8:\n",
    "                Ns_8[action[i][j]-1] += 1\n",
    "            if j < 12:\n",
    "                Ns_12[action[i][j]-1] += 1\n",
    "    x_labels = [i for i in range(1,21)]\n",
    "    fig, ax = plt.subplots(figsize=(10,6))\n",
    "    ind = np.arange(20)\n",
    "    width = 0.35\n",
    "    p1 = ax.bar(ind, Ns_4, width, color='indianred')\n",
    "    p2 = ax.bar(ind+width, Ns_8, width, color='dimgray')\n",
    "    p3 = ax.bar(ind+width*2, Ns_12, width, color='lightseagreen')\n",
    "    ax.set_title('Joint apperances of action' + action_label)\n",
    "    ax.set_xticks(ind+width/2)\n",
    "    ax.set_xticklabels(x_labels)\n",
    "    ax.legend((p1[0], p2[0], p3[0]), ('Ns_4', 'Ns_8', 'Ns_12'))\n",
    "    ax.autoscale_view()\n",
    "    plt.show()\n",
    "    return Ns_4, Ns_8, Ns_12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([0, 0, 0, 0, 0, 1, 7, 5, 4, 6, 5, 9, 6, 6, 6, 4, 0, 1, 0, 0],\n",
       " [0, 0, 1, 1, 3, 4, 8, 10, 8, 12, 10, 13, 13, 13, 12, 7, 3, 2, 0, 0],\n",
       " [2, 0, 3, 7, 9, 11, 11, 13, 12, 14, 13, 15, 15, 14, 15, 12, 7, 4, 3, 0])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a01')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([0, 0, 0, 1, 0, 0, 3, 7, 5, 8, 10, 9, 6, 4, 5, 2, 0, 0, 0, 0],\n",
       " [0, 2, 1, 4, 5, 5, 6, 10, 14, 11, 14, 13, 9, 7, 9, 4, 3, 2, 1, 0],\n",
       " [5, 4, 2, 5, 10, 10, 11, 12, 15, 13, 15, 14, 14, 11, 12, 8, 7, 7, 5, 0])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a06')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([0, 0, 0, 0, 1, 0, 0, 2, 1, 4, 5, 8, 9, 9, 11, 5, 1, 3, 1, 0],\n",
       " [6, 2, 0, 0, 2, 1, 1, 5, 4, 5, 9, 11, 13, 13, 12, 7, 9, 11, 8, 1],\n",
       " [12, 7, 2, 2, 3, 2, 4, 7, 8, 7, 12, 13, 14, 14, 13, 14, 13, 15, 14, 4])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a20')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([0, 6, 3, 2, 9, 2, 4, 1, 3, 1, 1, 2, 2, 3, 4, 4, 4, 2, 7, 0],\n",
       " [7, 8, 7, 5, 11, 4, 9, 4, 6, 2, 5, 6, 6, 7, 5, 8, 7, 2, 9, 2],\n",
       " [9, 10, 10, 9, 14, 10, 10, 7, 6, 3, 6, 6, 7, 12, 8, 12, 13, 9, 14, 5])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a14')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([2, 1, 1, 4, 4, 3, 3, 6, 8, 3, 4, 4, 0, 3, 4, 1, 3, 1, 3, 2],\n",
       " [7, 2, 5, 8, 7, 4, 4, 10, 8, 9, 6, 6, 1, 9, 5, 4, 6, 5, 9, 5],\n",
       " [10, 7, 5, 10, 10, 9, 7, 11, 9, 9, 10, 10, 6, 10, 8, 6, 12, 10, 14, 7])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a16')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sort_and_keep_indexes(Ns, numOfJoints):\n",
    "    Ns = sorted(range(len(Ns)), key = Ns.__getitem__)\n",
    "    Ns = Ns[len(Ns)-numOfJoints:]\n",
    "    return Ns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Similar method but it only gets indexes\n",
    "def get_idx_mostJoints_for_each_action(action_label):\n",
    "    action = train_mostInformativeJointsList[action_label]\n",
    "    Ns_4 = [0 for i in range(20)]\n",
    "    Ns_8 = [0 for i in range(20)]\n",
    "    Ns_12 = [0 for i in range(20)]\n",
    "\n",
    "    for i in range(len(action)):\n",
    "        for j in range(0, 12):\n",
    "            if j < 4:\n",
    "                Ns_4[action[i][j] - 1] += 1\n",
    "            if j < 8:\n",
    "                Ns_8[action[i][j] - 1] += 1\n",
    "            if j < 12:\n",
    "                Ns_12[action[i][j] - 1] += 1\n",
    "    Ns_4 = np.asarray(sort_and_keep_indexes(Ns_4, 4))\n",
    "    Ns_8 = np.asarray(sort_and_keep_indexes(Ns_8, 8))\n",
    "    Ns_12 = np.asarray(sort_and_keep_indexes(Ns_12, 12))\n",
    "    return Ns_4, Ns_8, Ns_12"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## For 4 most informative joints / Trash code/ Done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get action labels\n",
    "# Hold this code here\n",
    "# For Ns_4 only\n",
    "action_label_list = list(train_mostInformativeJointsList.keys())\n",
    "Ns4_list = np.zeros((len(action_label_list), 4), dtype=np.int)\n",
    "for i in range(Ns4_list.shape[0]):\n",
    "    Ns_4, _, _ = get_idx_mostJoints_for_each_action(action_label_list[i])\n",
    "    Ns4_list[i] = Ns_4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_MIJ_matrices(filename):\n",
    "    \n",
    "    action_label = int(filename[1:3])-1\n",
    "    \n",
    "    with open('data/' + filename, 'r') as f:\n",
    "        skeleton_matrix = [line.rstrip('\\n') for line in f]\n",
    "    for j in range(len(skeleton_matrix)):\n",
    "        skeleton_matrix[j] = [float(coord) for coord in skeleton_matrix[j].split(' ')]\n",
    "    \n",
    "    skeleton_matrix = np.asarray(skeleton_matrix)\n",
    "    noFrames = int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS)\n",
    "    x = skeleton_matrix[:,0]\n",
    "    y = skeleton_matrix[:,1]\n",
    "    z = skeleton_matrix[:,2]\n",
    "\n",
    "    x = x.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    y = y.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    z = z.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    t = np.arange(1, noFrames+1).T\n",
    "    x = x[Ns4_list[action_label]]\n",
    "    y = y[Ns4_list[action_label]]\n",
    "    z = z[Ns4_list[action_label]]\n",
    "    return x,y,z,t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculateCovarianceMat(X, Y, Z, T, nLevels, overlap = False, timeVar=True):\n",
    "    nFrames = X.shape[0]\n",
    "    nJoins = X.shape[1]\n",
    "    assert Y.shape[0] == nFrames\n",
    "    assert Z.shape[0] == nFrames\n",
    "    assert T.shape[0] == nFrames\n",
    "    assert Y.shape[1] == nJoins\n",
    "    assert Z.shape[1] == nJoins\n",
    "\n",
    "    # Normalize skeleton coordinators\n",
    "\n",
    "    normX = normCord(X)\n",
    "    normY = normCord(Y)\n",
    "    normZ = normCord(Z)\n",
    "    normT = normSeT(T)\n",
    "\n",
    "    # Create a list of full covariance matrices\n",
    "    fullCovMats = [[] for i in range(nLevels)]\n",
    "    covMats = [[] for i in range(nLevels)]\n",
    "\n",
    "    if timeVar:\n",
    "        sizeMatrix = nJoins*3+1\n",
    "    else:\n",
    "        sizeMatrix = nJoins*3\n",
    "\n",
    "    listIdxMatrix = getValueMatrix(sizeMatrix) # get half of covariance matrix indexes\n",
    "\n",
    "\n",
    "    for l in range(1,nLevels+1):\n",
    "        # Compute covariance matrixes for each level\n",
    "        nofMats = 2**(l-1)\n",
    "        sizeWindow = 1/nofMats\n",
    "        stepWindow = sizeWindow\n",
    "        if overlap:\n",
    "            stepWindow = stepWindow/2\n",
    "            nofMats = nofMats*2-1\n",
    "        startFrameTimes = [stepWindow*i for i in range(nofMats)]\n",
    "        fullCovMats[l-1] = [[] for i in range(nofMats)]\n",
    "        covMats[l-1] = [[] for i in range(nofMats)]\n",
    "        for i in range(len(startFrameTimes)):\n",
    "            startTime = startFrameTimes[i]\n",
    "            endTime = startFrameTimes[i] + sizeWindow + 2*np.finfo(float).eps\n",
    "            sliceInds = [i for i in range(T.shape[0]) if normT[i] >= startTime and normT[i] < endTime]\n",
    "            sliceX = normX[sliceInds, :]\n",
    "            sliceY = normY[sliceInds, :]\n",
    "            sliceZ = normZ[sliceInds, :]\n",
    "            sliceT = normT[sliceInds]\n",
    "            if not timeVar:\n",
    "                sliceVars = np.concatenate((np.concatenate((sliceX,sliceY), axis=1), sliceZ), axis=1)\n",
    "            else:\n",
    "                sliceVars = np.concatenate((np.concatenate((sliceX, sliceY), axis=1), np.concatenate((sliceZ, sliceT), axis=1)), axis=1)\n",
    "            covarianceMat = np.cov(sliceVars.T)\n",
    "            fullCovMats[l-1][i] = covarianceMat\n",
    "            # Get half of covarianceMat and save it as a vector (1-D matrix)\n",
    "            one_half_vector = []\n",
    "            mask = np.zeros_like(covarianceMat, dtype=np.bool)\n",
    "            mask[np.triu_indices_from(mask)] = True\n",
    "            for row in range(covarianceMat.shape[0]):\n",
    "                for column in range(covarianceMat.shape[0]):\n",
    "                    if mask[row][column] == True:\n",
    "                        one_half_vector.append(covarianceMat[row][column])\n",
    "            covMats[l-1][i] = np.asarray(one_half_vector)\n",
    "        covMats[l-1] = np.asarray(covMats[l-1])\n",
    "        \n",
    "    covMats = np.asarray(covMats)\n",
    "    vec = np.empty(0)\n",
    "    for i in range(covMats.shape[0]):\n",
    "        for j in range(covMats[i].shape[0]):\n",
    "            vec = np.hstack((vec, covMats[i][j]))\n",
    "    return fullCovMats, vec"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Now the training part is coming"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_covariance_vector(x,y,z,t):\n",
    "    fullCovmat, vec_covMat = calculateCovarianceMat(x.T, y.T, z.T, t, nLevels=3, overlap=True, timeVar=False)\n",
    "    return vec_covMat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_preprocessed_training_data(filenames):\n",
    "    X_train, y_train = [], []\n",
    "    for i in range(len(filenames)):\n",
    "        x, y, z, t = get_MIJ_matrices(filenames[i])\n",
    "        vec_covMat = get_covariance_vector(x,y,z,t)\n",
    "        label = int(filenames[i][1:3])\n",
    "        X_train.append(vec_covMat)\n",
    "        y_train.append(label)\n",
    "    X_train = np.asarray(X_train)\n",
    "    y_train = np.asarray(y_train)\n",
    "    return X_train, y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, y_train = make_preprocessed_training_data(train_data_file_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = OneVsRestClassifier(LinearSVC(random_state=0)).fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test, y_test = make_preprocessed_training_data(test_data_file_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_hat = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_acc(y_test, y_hat):\n",
    "    count = 0\n",
    "    for i in range(y_test.shape[0]):\n",
    "        if y_test[i] == y_hat[i]:\n",
    "            count += 1\n",
    "    acc = (count/y_test.shape[0])*100\n",
    "    return acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "46.666666666666664\n"
     ]
    }
   ],
   "source": [
    "print(get_acc(y_test, y_hat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_nan(X):\n",
    "    Nan = np.isnan(X)\n",
    "    for i in range(Nan.shape[0]):\n",
    "        for j in range(Nan.shape[1]):\n",
    "            if Nan[i][j] == True:\n",
    "                print(i,j)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## For action set 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "from MostJoints.getMostInformativeJoints import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(ACTION_SETS_PATH + 'ActionSet3.txt', 'r') as f:\n",
    "    as3_names = f.readlines()\n",
    "for i in range(len(as3_names)):\n",
    "    as3_names[i] = as3_names[i].rstrip(\"\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_as3_names = [as3_names[i] for i in range(len(as3_names)) if as3_names[i][4:7] in TRAINING_SUBJECTS]\n",
    "test_as3_names = [as3_names[i] for i in range(len(as3_names)) if as3_names[i][4:7] in TESTING_SUBJECTS]\n",
    "train_as3_names = [train_as3_names[i] + '_skeleton3D.txt' for i in range(len(train_as3_names))]\n",
    "test_as3_names = [test_as3_names[i] + '_skeleton3D.txt' for i in range(len(test_as3_names))]\n",
    "as3_names = [as3_names[i] + '_skeleton3D.txt' for i in range(len(as3_names))]\n",
    "train_as3_noFrameList, train_as3_mostInformativeJointsList = get_most_informative_joints(train_as3_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_histograms(action_label):\n",
    "    action = train_as3_mostInformativeJointsList[action_label]\n",
    "    Ns_4 = [0 for i in range(20)]\n",
    "    Ns_8 = [0 for i in range(20)]\n",
    "    Ns_12 = [0 for i in range(20)]\n",
    "    \n",
    "    for i in range(len(action)):\n",
    "        for j in range(0, 20):\n",
    "            if j < 4:\n",
    "                Ns_4[action[i][j]-1] += 1\n",
    "            if j < 8:\n",
    "                Ns_8[action[i][j]-1] += 1\n",
    "            if j < 12:\n",
    "                Ns_12[action[i][j]-1] += 1\n",
    "    x_labels = [i for i in range(1,21)]\n",
    "    fig, ax = plt.subplots(figsize=(10,6))\n",
    "    ind = np.arange(20)\n",
    "    width = 0.35\n",
    "    p1 = ax.bar(ind, Ns_4, width, color='indianred')\n",
    "    p2 = ax.bar(ind+width, Ns_8, width, color='dimgray')\n",
    "    p3 = ax.bar(ind+width*2, Ns_12, width, color='lightseagreen')\n",
    "    ax.set_title('Joint apperances of action' + action_label)\n",
    "    ax.set_xticks(ind+width/2)\n",
    "    ax.set_xticklabels(x_labels)\n",
    "    ax.legend((p1[0], p2[0], p3[0]), ('Ns_4', 'Ns_8', 'Ns_12'))\n",
    "    ax.autoscale_view()\n",
    "    plt.show()\n",
    "    return Ns_4, Ns_8, Ns_12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "([0, 0, 0, 1, 0, 0, 3, 7, 5, 8, 10, 9, 6, 4, 5, 2, 0, 0, 0, 0],\n",
       " [0, 2, 1, 4, 5, 5, 6, 10, 14, 11, 14, 13, 9, 7, 9, 4, 3, 2, 1, 0],\n",
       " [5, 4, 2, 5, 10, 10, 11, 12, 15, 13, 15, 14, 14, 11, 12, 8, 7, 7, 5, 0])"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_histograms('a06')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [],
   "source": [
    "action = train_as3_mostInformativeJointsList['a14']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([19,  6, 16,  9, 13,  2,  3,  5,  0,  7,  8, 18, 15, 12,  1, 17,  4,\n",
       "        14, 10, 11]),\n",
       " array([ 2, 18,  5, 13, 14,  1, 17, 11, 19, 10,  3, 15,  6, 12,  4,  9,  0,\n",
       "        16,  7,  8]),\n",
       " array([15, 16, 14,  2,  0,  1,  9, 17, 18, 19,  3, 13, 11,  8, 10, 12,  4,\n",
       "         7,  5,  6]),\n",
       " array([ 7,  9,  2, 19,  4, 12, 17, 10, 16, 14,  5,  8, 15, 11,  6,  0, 18,\n",
       "        13,  1,  3]),\n",
       " array([ 8,  7,  9, 10,  6, 12, 11,  0,  1, 18,  5,  4,  2, 19,  3, 17, 13,\n",
       "        16, 15, 14]),\n",
       " array([ 5,  2, 19, 16, 11, 14, 13,  8, 17,  0,  4,  6,  3,  1,  7, 18, 10,\n",
       "        15, 12,  9]),\n",
       " array([14,  3,  5, 12,  1, 19, 16,  7, 17, 18,  6,  8, 10, 15,  0, 13, 11,\n",
       "         9,  4,  2]),\n",
       " array([19, 15, 17, 13,  1,  3,  5,  7, 11,  6, 14,  4, 12,  2, 16,  0, 18,\n",
       "         9,  8, 10]),\n",
       " array([11,  5, 15, 18,  1, 12,  2,  4, 14, 19, 17, 16, 13,  3,  6, 10,  9,\n",
       "         0,  7,  8]),\n",
       " array([14,  3,  5, 12,  1, 16, 19,  7, 18, 17,  6, 15, 13,  0, 11,  4,  2,\n",
       "        10,  9,  8]),\n",
       " array([15, 17,  5, 19,  1,  3, 13,  7, 14,  6, 16,  4,  2,  0, 18,  8, 12,\n",
       "        11,  9, 10]),\n",
       " array([ 4, 17,  5,  3, 16,  6, 15,  7, 18, 19, 14,  2, 13,  0,  1,  8, 12,\n",
       "        10,  9, 11]),\n",
       " array([ 5,  2, 19, 16, 14, 11, 13,  8, 17,  0,  3,  6,  1,  7,  4, 18, 15,\n",
       "        12, 10,  9]),\n",
       " array([ 7,  2, 17, 19,  4, 14,  9, 12,  5, 16,  1, 15,  0,  6,  8,  3, 10,\n",
       "        18, 13, 11])]"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['a01', 'a02', 'a03', 'a04', 'a05', 'a06', 'a07', 'a08', 'a09', 'a10', 'a11', 'a12', 'a13', 'a14', 'a15', 'a16', 'a17', 'a18', 'a19', 'a20']\n"
     ]
    }
   ],
   "source": [
    "# Get action labels\n",
    "# Hold this code here\n",
    "# For Ns_4 only\n",
    "def get_idx_mostJoints_for_each_action_as1(action_label):\n",
    "    action = train_as1_mostInformativeJointsList[action_label]\n",
    "    Ns_4 = [0 for i in range(20)]\n",
    "    Ns_8 = [0 for i in range(20)]\n",
    "    Ns_12 = [0 for i in range(20)]\n",
    "\n",
    "    for i in range(len(action)):\n",
    "        for j in range(0, 12):\n",
    "            if j < 4:\n",
    "                Ns_4[action[i][j] - 1] += 1\n",
    "            if j < 8:\n",
    "                Ns_8[action[i][j] - 1] += 1\n",
    "            if j < 12:\n",
    "                Ns_12[action[i][j] - 1] += 1\n",
    "    Ns_4 = np.asarray(sort_and_keep_indexes(Ns_4, 4))\n",
    "    Ns_8 = np.asarray(sort_and_keep_indexes(Ns_8, 8))\n",
    "    Ns_12 = np.asarray(sort_and_keep_indexes(Ns_12, 12))\n",
    "    return Ns_4, Ns_8, Ns_12\n",
    "\n",
    "action_label_list = list(train_as1_mostInformativeJointsList.keys())\n",
    "print(action_label_list)\n",
    "Ns4_as1_list = np.zeros((len(action_label_list), 4), dtype=np.int)\n",
    "for i in range(Ns4_as1_list.shape[0]):\n",
    "    Ns_4, _, _ = get_idx_mostJoints_for_each_action_as1(action_label_list[i])\n",
    "    Ns4_as1_list[i] = Ns_4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-146-821c7cdacef9>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mzeros_array\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m16\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m17\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m18\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m19\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mNs4_as1_list\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m     \u001b[0;32mif\u001b[0m \u001b[0mNs4_as1_list\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mall\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mzeros_array\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m         \u001b[0mNs4_as1_list\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0mNs4_as1_list\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()"
     ]
    }
   ],
   "source": [
    "# Hold this code here\n",
    "# Find out Why I extracted the wrong most joints\n",
    "zeros_array = np.array([16,17,18,19])\n",
    "for i in range(Ns4_as1_list.shape[0]):\n",
    "    if Ns4_as1_list[i].all() == zeros_array:\n",
    "        Ns4_as1_list[i] = np.array([0,0,0,0])\n",
    "Ns4_as1_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_MIJ_matrices(filename):\n",
    "    \n",
    "    action_label = int(filename[1:3])-1\n",
    "    \n",
    "    with open('data/' + filename, 'r') as f:\n",
    "        skeleton_matrix = [line.rstrip('\\n') for line in f]\n",
    "    for j in range(len(skeleton_matrix)):\n",
    "        skeleton_matrix[j] = [float(coord) for coord in skeleton_matrix[j].split(' ')]\n",
    "    \n",
    "    skeleton_matrix = np.asarray(skeleton_matrix)\n",
    "    noFrames = int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS)\n",
    "    x = skeleton_matrix[:,0]\n",
    "    y = skeleton_matrix[:,1]\n",
    "    z = skeleton_matrix[:,2]\n",
    "\n",
    "    x = x.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    y = y.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    z = z.reshape(setting.NUMBER_OF_JOINTS, int(skeleton_matrix.shape[0]/setting.NUMBER_OF_JOINTS))\n",
    "    t = np.arange(1, noFrames+1).T\n",
    "    x = x[Ns4_as1_list[action_label]]\n",
    "    y = y[Ns4_as1_list[action_label]]\n",
    "    z = z[Ns4_as1_list[action_label]]\n",
    "    return x,y,z,t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_as1_train, y_as1_train = make_preprocessed_training_data(train_as1_names)\n",
    "X_as1_test, y_as1_test = make_preprocessed_training_data(test_as1_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'OneVsRestClassifier' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-1-b267026fab2a>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mOneVsRestClassifier\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mLinearSVC\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkernel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'rbf'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrandom_state\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_as1_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_as1_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'OneVsRestClassifier' is not defined"
     ]
    }
   ],
   "source": [
    "model = OneVsRestClassifier(LinearSVC( random_state=0)).fit(X_as1_train, y_as1_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_as1_hat = model.predict(X_as1_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "63.8095238095238\n"
     ]
    }
   ],
   "source": [
    "print(get_acc(y_as1_test, y_as1_hat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[13, 14,  6, 11],\n",
       "       [13, 15, 10,  8],\n",
       "       [ 8,  5,  7,  6],\n",
       "       [ 7, 10,  8,  9],\n",
       "       [ 9, 11, 13,  8],\n",
       "       [ 7,  9, 11, 10],\n",
       "       [ 6,  4,  5, 11],\n",
       "       [14, 13, 12, 11],\n",
       "       [10,  9, 11,  7],\n",
       "       [ 7,  9,  5,  8],\n",
       "       [ 8,  9, 12, 10],\n",
       "       [ 7,  8, 12, 11],\n",
       "       [ 1, 15,  2, 18],\n",
       "       [16,  1, 18,  4],\n",
       "       [12, 18,  9,  8],\n",
       "       [11, 14,  7,  8],\n",
       "       [12, 15, 10, 14],\n",
       "       [11,  9,  7,  8],\n",
       "       [14, 13, 15, 12],\n",
       "       [11, 12, 13, 14]])"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Ns4_as1_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[13, 14,  6, 11],\n",
       "       [13, 15, 10,  8],\n",
       "       [ 8,  5,  7,  6],\n",
       "       [ 7, 10,  8,  9],\n",
       "       [ 9, 11, 13,  8],\n",
       "       [ 7,  9, 11, 10],\n",
       "       [ 6,  4,  5, 11],\n",
       "       [14, 13, 12, 11],\n",
       "       [10,  9, 11,  7],\n",
       "       [ 7,  9,  5,  8],\n",
       "       [ 8,  9, 12, 10],\n",
       "       [ 7,  8, 12, 11],\n",
       "       [ 1, 15,  2, 18],\n",
       "       [16,  1, 18,  4],\n",
       "       [12, 18,  9,  8],\n",
       "       [11, 14,  7,  8],\n",
       "       [12, 15, 10, 14],\n",
       "       [11,  9,  7,  8],\n",
       "       [14, 13, 15, 12],\n",
       "       [11, 12, 13, 14]])"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Ns4_list"
   ]
  },
  {
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}