data_viz.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 122,
   "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 nilearn import image, plotting, datasets\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make 2D plots of predicted vs real scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_path = './results/NEOFAC_{}_{}_predictions.csv'\n",
    "MSE_path = './results/NEOFAC_{}_{}_MSE.csv'\n",
    "traits = pd.read_csv('all_traits.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "N_SVR_MSE = pd.read_csv(MSE_path.format('N','SVR'), index_col=0, names=['type','N'],header=0) #best idx=0 (significant) 54.563558\n",
    "N_LR_MSE = pd.read_csv(MSE_path.format('N','LR'), index_col=0, names=['type','N'],header=0) #best idx=1 66.480085\n",
    "N_CPM_MSE = pd.read_csv(MSE_path.format('N','CPM'), index_col=0, names=['type','N'],header=0) #best idx=0 combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "O_SVR_MSE = pd.read_csv(MSE_path.format('O','SVR'), index_col=0, names=['type','O'],header=0) #best idx=0 (significant) 35.662023\n",
    "O_LR_MSE = pd.read_csv(MSE_path.format('O','LR'), index_col=0, names=['type','O'],header=0)\n",
    "O_CPM_MSE = pd.read_csv(MSE_path.format('O','CPM'), index_col=0, names=['type','O'],header=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "E_SVR_MSE = pd.read_csv(MSE_path.format('E','SVR'), index_col=0, names=['type','E'],header=0) \n",
    "E_LR_MSE = pd.read_csv(MSE_path.format('E','LR'), index_col=0, names=['type','E'],header=0)\n",
    "E_CPM_MSE = pd.read_csv(MSE_path.format('E','CPM'), index_col=0, names=['type','E'],header=0) #best idx=0 (combined) 35.261666"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "A_SVR_MSE = pd.read_csv(MSE_path.format('A','SVR'), index_col=0, names=['type','A'],header=0) #best idx=1 (positive) 31.501072\n",
    "A_LR_MSE = pd.read_csv(MSE_path.format('A','LR'), index_col=0, names=['type','A'],header=0)\n",
    "A_CPM_MSE = pd.read_csv(MSE_path.format('A','CPM'), index_col=0, names=['type','A'],header=0) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "C_SVR_MSE = pd.read_csv(MSE_path.format('C','SVR'), index_col=0, names=['type','C'],header=0) #best idx=2 (negative) 34.086224\n",
    "C_LR_MSE = pd.read_csv(MSE_path.format('C','LR'), index_col=0, names=['type','C'],header=0)\n",
    "C_CPM_MSE = pd.read_csv(MSE_path.format('C','CPM'), index_col=0, names=['type','C'],header=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true,
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>E</th>\n",
       "      <th>A</th>\n",
       "      <th>N</th>\n",
       "      <th>O</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>significant</td>\n",
       "      <td>36.129105</td>\n",
       "      <td>31.737188</td>\n",
       "      <td>54.563558</td>\n",
       "      <td>35.662023</td>\n",
       "      <td>34.334938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>positive</td>\n",
       "      <td>35.457866</td>\n",
       "      <td>31.501072</td>\n",
       "      <td>55.802892</td>\n",
       "      <td>35.975400</td>\n",
       "      <td>34.877439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>negative</td>\n",
       "      <td>35.891109</td>\n",
       "      <td>32.382376</td>\n",
       "      <td>54.702127</td>\n",
       "      <td>37.092324</td>\n",
       "      <td>34.086224</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          type          E          A          N          O          C\n",
       "0  significant  36.129105  31.737188  54.563558  35.662023  34.334938\n",
       "1     positive  35.457866  31.501072  55.802892  35.975400  34.877439\n",
       "2     negative  35.891109  32.382376  54.702127  37.092324  34.086224"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SVR_MSE= pd.merge(E_SVR_MSE,A_SVR_MSE,on='type')\n",
    "SVR_MSE= pd.merge(SVR_MSE,N_SVR_MSE,on='type')\n",
    "SVR_MSE= pd.merge(SVR_MSE,O_SVR_MSE,on='type')\n",
    "SVR_MSE= pd.merge(SVR_MSE,C_SVR_MSE,on='type')\n",
    "SVR_MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true,
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>E</th>\n",
       "      <th>A</th>\n",
       "      <th>N</th>\n",
       "      <th>O</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>significant</td>\n",
       "      <td>59.234352</td>\n",
       "      <td>127.985687</td>\n",
       "      <td>82.488318</td>\n",
       "      <td>56.476792</td>\n",
       "      <td>49.480258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>positive</td>\n",
       "      <td>42.646618</td>\n",
       "      <td>49.498272</td>\n",
       "      <td>66.480085</td>\n",
       "      <td>46.466701</td>\n",
       "      <td>47.083576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>negative</td>\n",
       "      <td>46.760995</td>\n",
       "      <td>54.987092</td>\n",
       "      <td>81.981513</td>\n",
       "      <td>42.314519</td>\n",
       "      <td>38.824320</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          type          E           A          N          O          C\n",
       "0  significant  59.234352  127.985687  82.488318  56.476792  49.480258\n",
       "1     positive  42.646618   49.498272  66.480085  46.466701  47.083576\n",
       "2     negative  46.760995   54.987092  81.981513  42.314519  38.824320"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LR_MSE= pd.merge(E_LR_MSE,A_LR_MSE,on='type')\n",
    "LR_MSE= pd.merge(LR_MSE,N_LR_MSE,on='type')\n",
    "LR_MSE= pd.merge(LR_MSE,O_LR_MSE,on='type')\n",
    "LR_MSE= pd.merge(LR_MSE,C_LR_MSE,on='type')\n",
    "LR_MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true,
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>type</th>\n",
       "      <th>E</th>\n",
       "      <th>A</th>\n",
       "      <th>N</th>\n",
       "      <th>O</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>combined</td>\n",
       "      <td>35.261666</td>\n",
       "      <td>33.062295</td>\n",
       "      <td>56.267819</td>\n",
       "      <td>39.090016</td>\n",
       "      <td>35.724548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>positive</td>\n",
       "      <td>40.115013</td>\n",
       "      <td>34.139004</td>\n",
       "      <td>60.140462</td>\n",
       "      <td>39.769968</td>\n",
       "      <td>38.443678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>negative</td>\n",
       "      <td>37.351060</td>\n",
       "      <td>32.932149</td>\n",
       "      <td>57.338657</td>\n",
       "      <td>38.429741</td>\n",
       "      <td>35.897714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>multiple_reg</td>\n",
       "      <td>39.122604</td>\n",
       "      <td>34.953438</td>\n",
       "      <td>58.133408</td>\n",
       "      <td>38.090401</td>\n",
       "      <td>36.099302</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           type          E          A          N          O          C\n",
       "0      combined  35.261666  33.062295  56.267819  39.090016  35.724548\n",
       "1      positive  40.115013  34.139004  60.140462  39.769968  38.443678\n",
       "2      negative  37.351060  32.932149  57.338657  38.429741  35.897714\n",
       "3  multiple_reg  39.122604  34.953438  58.133408  38.090401  36.099302"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CPM_MSE= pd.merge(E_CPM_MSE,A_CPM_MSE,on='type')\n",
    "CPM_MSE= pd.merge(CPM_MSE,N_CPM_MSE,on='type')\n",
    "CPM_MSE= pd.merge(CPM_MSE,O_CPM_MSE,on='type')\n",
    "CPM_MSE= pd.merge(CPM_MSE,C_CPM_MSE,on='type')\n",
    "CPM_MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "pred_N_SVR = np.loadtxt(pred_path.format('N','SVR'))[:,0]\n",
    "pred_N_LR = np.loadtxt(pred_path.format('N','LR'))[:,1]\n",
    "pred_N_CPM = np.loadtxt(pred_path.format('N','CPM'))[:,0]\n",
    "pred_O_SVR = np.loadtxt(pred_path.format('O','SVR'))[:,0]\n",
    "pred_O_LR = np.loadtxt(pred_path.format('O','LR'))[:,1]\n",
    "pred_O_CPM = np.loadtxt(pred_path.format('O','CPM'))[:,0]\n",
    "pred_E_SVR = np.loadtxt(pred_path.format('E','SVR'))[:,0]\n",
    "pred_E_LR = np.loadtxt(pred_path.format('E','LR'))[:,1]\n",
    "pred_E_CPM = np.loadtxt(pred_path.format('E','CPM'))[:,0]\n",
    "pred_A_SVR = np.loadtxt(pred_path.format('A','SVR'))[:,0]\n",
    "pred_A_LR = np.loadtxt(pred_path.format('A','LR'))[:,1]\n",
    "pred_A_CPM = np.loadtxt(pred_path.format('A','CPM'))[:,0]\n",
    "pred_C_SVR = np.loadtxt(pred_path.format('C','SVR'))[:,0]\n",
    "pred_C_LR = np.loadtxt(pred_path.format('C','LR'))[:,1]\n",
    "pred_C_CPM = np.loadtxt(pred_path.format('E','CPM'))[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15\n"
     ]
    }
   ],
   "source": [
    "pred_col = [pred_N_SVR,pred_N_LR,pred_N_CPM,pred_O_SVR,pred_O_LR,pred_O_CPM,pred_E_SVR,pred_E_LR,pred_E_CPM,\n",
    "            pred_A_SVR,pred_A_LR,pred_A_CPM,pred_C_SVR,pred_C_LR,pred_C_CPM]\n",
    "pred_names = ['pred_N_SVR','pred_N_LR','pred_N_CPM','pred_O_SVR','pred_O_LR','pred_O_CPM','pred_E_SVR','pred_E_LR','pred_E_CPM',\n",
    "            'pred_A_SVR','pred_A_LR','pred_A_CPM','pred_C_SVR','pred_C_LR','pred_C_CPM']\n",
    "print(len(pred_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred = pd.DataFrame(np.transpose(np.vstack(pred_col)),\n",
    "                         columns=pred_names)\n",
    "df = pd.concat([pred,traits],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true,
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for trait in ['C']:\n",
    "    line = np.linspace(0,50,5)\n",
    "    g = sns.JointGrid(x=\"NEOFAC_{}\".format(trait), y=\"pred_{}_SVR\".format(trait), data=df, height=10,xlim=(0,50),ylim=(0,50))\n",
    "    g = g.plot_joint(sns.scatterplot, color=\"b\", alpha=.6)\n",
    "    g = g.plot_marginals(sns.distplot, kde=True, color=\"b\")\n",
    "    g.set_axis_labels('Real scores','Predicted scores')\n",
    "    g.ax_joint.plot(line,line,':k')\n",
    "    g.savefig('./figures/{}_SVR.png'.format(trait))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize Connectomes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ICA map\n",
    "path2 = './HCP_PTN1200_recon2/groupICA_3T_HCP1200_MSMAll/groupICA/groupICA_3T_HCP1200_MSMAll_d200.ica/{}'\n",
    "ICA_map = image.load_img(path2.format('melodic_IC_sum.nii'))\n",
    "coords = plotting.find_probabilistic_atlas_cut_coords(ICA_map)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sig</th>\n",
       "      <th>pos</th>\n",
       "      <th>neg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>40000.000000</td>\n",
       "      <td>40000.000000</td>\n",
       "      <td>40000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>12.418700</td>\n",
       "      <td>5.374400</td>\n",
       "      <td>7.044300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>97.576991</td>\n",
       "      <td>64.387514</td>\n",
       "      <td>73.832492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>810.000000</td>\n",
       "      <td>810.000000</td>\n",
       "      <td>810.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                sig           pos           neg\n",
       "count  40000.000000  40000.000000  40000.000000\n",
       "mean      12.418700      5.374400      7.044300\n",
       "std       97.576991     64.387514     73.832492\n",
       "min        0.000000      0.000000      0.000000\n",
       "25%        0.000000      0.000000      0.000000\n",
       "50%        0.000000      0.000000      0.000000\n",
       "75%        0.000000      0.000000      0.000000\n",
       "max      810.000000    810.000000    810.000000"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "edges = pd.DataFrame(np.loadtxt('./results/NEOFAC_N_LR_edge_count.csv'), columns = ['sig','pos','neg'])\n",
    "edges.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [],
   "source": [
    "#connectome of subject with very high neuroticism (max negative score)\n",
    "view = plotting.view_connectome(edges['pos'].to_numpy().reshape(-1,200), coords, edge_cmap = 'Reds', linewidth=4, colorbar_fontsize=10,\n",
    "                                symmetric_cmap=False, title='Connections Positively Correlated with Neuroticism',title_fontsize=15)\n",
    "view.save_as_html('positive_connections.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "#connectome of subject with very high neuroticism (max negative score)\n",
    "view = plotting.view_connectome(edges['neg'].to_numpy().reshape(-1,200), coords, edge_cmap = 'Blues', linewidth=4, colorbar_fontsize=10,\n",
    "                                symmetric_cmap=False, title='Connections Negatively Correlated with Neuroticism',title_fontsize=15)\n",
    "view.save_as_html('negative_connections.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/annabelle/miniconda3/lib/python3.7/site-packages/numpy/ma/core.py:2795: UserWarning: Warning: converting a masked element to nan.\n",
      "  order=order, subok=True, ndmin=ndmin)\n"
     ]
    }
   ],
   "source": [
    "plotting.plot_prob_atlas(ICA_map,black_bg=True,output_file='200-ICA-MAP.png',annotate=True,view_type='continuous')"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}