Divergence-Figures.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from pylab import rcParams\n",
    "rcParams['figure.figsize'] = 10, 5\n",
    "\n",
    "sns.set(color_codes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "baseDrive = 'data'\n",
    "workingDirectory = baseDrive+'/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "intra_session_cross_entropy = np.load(workingDirectory+'intra-session_cross-entropy.npy')\n",
    "inter_session_cross_entropy = np.load(workingDirectory+'inter-session_cross-entropy.npy')\n",
    "inter_subject_cross_entropy = np.load(workingDirectory+'inter-subject_cross-entropy.npy')\n",
    "\n",
    "inter_session_cross_entropy_999 = np.load(workingDirectory+'inter-session_cross-entropy_999.npy')\n",
    "inter_subject_cross_entropy_999 = np.load(workingDirectory+'inter-subject_cross-entropy_999.npy')\n",
    "\n",
    "inter_session_cross_entropy_1000 = np.load(workingDirectory+'inter-session_cross-entropy_1000.npy')\n",
    "inter_session_cross_entropy_31 = np.load(workingDirectory+'inter-session_cross-entropy_31.npy')\n",
    "\n",
    "inter_subject_cross_entropy_1000 = np.load(workingDirectory+'inter-subject_cross-entropy_1000.npy')\n",
    "inter_subject_cross_entropy_31 = np.load(workingDirectory+'inter-subject_cross-entropy_31.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.16868027, 1.0191177)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intra_session_cross_entropy.mean(), intra_session_cross_entropy.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3.88, 4.4518313)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_session_cross_entropy.mean(), inter_session_cross_entropy.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6.3620415, 4.991061)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_subject_cross_entropy.mean(), inter_subject_cross_entropy.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.49525166, 0.8455143)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_session_cross_entropy_1000.mean(), inter_session_cross_entropy_1000.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.3318869, 0.69579417)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_session_cross_entropy_31.mean(), inter_session_cross_entropy_31.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.28034005, 0.52693576)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_subject_cross_entropy_1000.mean(), inter_subject_cross_entropy_1000.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.2182601, 0.47341797)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inter_subject_cross_entropy_31.mean(), inter_subject_cross_entropy_31.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5200,), (10400,), (46800,))"
      ]
     },
     "execution_count": 264,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intra_session_cross_entropy.shape, inter_session_cross_entropy.shape, inter_subject_cross_entropy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=1, figsize=(7, 7), sharex=True)\n",
    "\n",
    "axes[0].text(1.3, 0.8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r')\n",
    "axes[0].set_title(\"Intra-session\")\n",
    "axes[0].hist(intra_session_cross_entropy, 15, density=True)\n",
    "axes[0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "\n",
    "axes[1].text(4, 0.4, r'$\\mu_1={:.2f}$'.format(np.mean(inter_session_cross_entropy)), color='r')\n",
    "axes[1].text(1, 0.4, r'$\\mu_1^l={:.2f}$'.format(np.mean(inter_session_cross_entropy_1000)), color='grey')\n",
    "axes[1].text(1, 0.8, r'$\\mu_1^d={:.2f}$'.format(np.mean(inter_session_cross_entropy_31)), color='g')\n",
    "axes[1].set_title(\"Inter-session domain shift\")\n",
    "axes[1].hist(inter_session_cross_entropy, 40, density=True)\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy_31), color='g', linestyle='-')\n",
    "\n",
    "axes[2].text(6.5, 0.4, r'$\\mu_2={:.2f}$'.format(np.mean(inter_subject_cross_entropy)), color='r')\n",
    "axes[2].text(1, 0.4, r'$\\mu_2^l={:.2f}$'.format(np.mean(inter_subject_cross_entropy_1000)), color='grey')\n",
    "axes[2].text(1, 0.6, r'$\\mu_2^d={:.2f}$'.format(np.mean(inter_subject_cross_entropy_31)), color='g')\n",
    "axes[2].set_title(\"Inter-subject domain shift\")\n",
    "axes[2].hist(inter_subject_cross_entropy, 80, density=True)\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy), color='r', linestyle='--')\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy_31), color='g', linestyle='-')\n",
    "\n",
    "axes[2].set_xlabel(\"Cross-entropy loss\")\n",
    "\n",
    "plt.xlim(0.0, 16.0)\n",
    "\n",
    "#fig.text(0.5, 0.04, 'Cross-entropy loss', ha='center', va='center')\n",
    "#fig.text(0.06, 0.5, 'Histogram', ha='center', va='center', rotation='vertical')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res.png', dpi=300)\n",
    "fig.savefig('divergences.png', dpi=1200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=2, figsize=(7, 7), sharex=False)\n",
    "\n",
    "axes[0, 0].text(0.1, .8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0, 0].transAxes)\n",
    "axes[0, 0].set_title(\"Intra-session\")\n",
    "axes[0, 0].hist(intra_session_cross_entropy, 15, density=True)\n",
    "axes[0, 0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0, 0].set_xlim(0.0, 15.0)\n",
    "\n",
    "axes[1, 0].text(0.3, 0.8, r'$\\mu_1={:.2f}$'.format(np.mean(inter_session_cross_entropy)), color='r', transform=axes[1, 0].transAxes)\n",
    "axes[1, 0].set_title(\"Inter-session domain shift\")\n",
    "axes[1, 0].hist(inter_session_cross_entropy, 40, density=True)\n",
    "axes[1, 0].axvline(np.mean(inter_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[1, 0].set_xlim(0.0, 15.0)\n",
    "\n",
    "axes[2, 0].text(0.5, 0.8, r'$\\mu_2={:.2f}$'.format(np.mean(inter_subject_cross_entropy)), color='r', transform=axes[2, 0].transAxes)\n",
    "axes[2, 0].set_title(\"Inter-subject domain shift\")\n",
    "axes[2, 0].hist(inter_subject_cross_entropy, 80, density=True)\n",
    "axes[2, 0].axvline(np.mean(inter_subject_cross_entropy), color='r', linestyle='--')\n",
    "\n",
    "axes[2, 0].set_xlabel(\"pre-DA cross-entropy loss\")\n",
    "axes[2, 0].set_xlim(0.0, 15.0)\n",
    "\n",
    "axes[0, 1].text(0.2, 0.8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0, 1].transAxes)\n",
    "axes[0, 1].set_title(\"Intra-session\")\n",
    "axes[0, 1].hist(intra_session_cross_entropy, 1000, density=True)\n",
    "axes[0, 1].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[1, 1].text(0.55, 0.6, r'$\\mu_1^l={:.2f}$'.format(np.mean(inter_session_cross_entropy_1000)), color='grey', transform=axes[1, 1].transAxes)\n",
    "axes[1, 1].text(0.55, 0.8, r'$\\mu_1^d={:.2f}$'.format(np.mean(inter_session_cross_entropy_31)), color='g', transform=axes[1, 1].transAxes)\n",
    "axes[1, 1].set_title(\"Inter-session domain shift\")\n",
    "axes[1, 1].hist(inter_session_cross_entropy_1000, 500, density=True)\n",
    "axes[1, 1].axvline(np.mean(inter_session_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[1, 1].axvline(np.mean(inter_session_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[1, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[2, 1].text(0.3, 0.6, r'$\\mu_2^l={:.2f}$'.format(np.mean(inter_subject_cross_entropy_1000)), color='grey', transform=axes[2, 1].transAxes)\n",
    "axes[2, 1].text(0.3, 0.8, r'$\\mu_2^d={:.2f}$'.format(np.mean(inter_subject_cross_entropy_31)), color='g', transform=axes[2, 1].transAxes)\n",
    "axes[2, 1].set_title(\"Inter-subject domain shift\")\n",
    "axes[2, 1].hist(inter_subject_cross_entropy_1000, 500, density=True)\n",
    "axes[2, 1].axvline(np.mean(inter_subject_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[2, 1].axvline(np.mean(inter_subject_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[2, 1].set_xlabel(\"post-DA cross-entropy loss\")\n",
    "axes[2, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "#fig.text(0.5, 0.04, 'Cross-entropy loss', ha='center', va='center')\n",
    "#fig.text(0.06, 0.5, 'Histogram', ha='center', va='center', rotation='vertical')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res-2.png', dpi=300)\n",
    "fig.savefig('divergences-2.png', dpi=1200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=1, ncols=1, figsize=(7, 7), sharex=False)\n",
    "\n",
    "axes.text(0.1, .8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes.transAxes)\n",
    "axes.set_title(\"Intra-session\")\n",
    "axes.hist(intra_session_cross_entropy, 2000, density=True)\n",
    "axes.axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes.set_xlim(0.0, 0.3)\n",
    "\n",
    "\n",
    "\n",
    "#fig.text(0.5, 0.04, 'Cross-entropy loss', ha='center', va='center')\n",
    "#fig.text(0.06, 0.5, 'Histogram', ha='center', va='center', rotation='vertical')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=1, figsize=(7, 7), sharex=True)\n",
    "\n",
    "axes[0].text(0.1, .8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0].transAxes)\n",
    "axes[0].set_title(\"Intra session\")\n",
    "axes[0].hist(intra_session_cross_entropy, 15, density=True)\n",
    "axes[0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0].set_xlim(0.0, 15.0)\n",
    "\n",
    "axes[1].text(0.3, 0.8, r'$\\mu_1={:.2f}$'.format(np.mean(inter_session_cross_entropy)), color='r', transform=axes[1].transAxes)\n",
    "axes[1].set_title(\"Inter-session domain shift\")\n",
    "axes[1].hist(inter_session_cross_entropy, 40, density=True)\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[1].set_xlim(0.0, 15.0)\n",
    "\n",
    "axes[2].text(0.5, 0.8, r'$\\mu_2={:.2f}$'.format(np.mean(inter_subject_cross_entropy)), color='r', transform=axes[2].transAxes)\n",
    "axes[2].set_title(\"Inter-subject domain shift\")\n",
    "axes[2].hist(inter_subject_cross_entropy, 80, density=True)\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy), color='r', linestyle='--')\n",
    "\n",
    "axes[2].set_xlabel(\"Cross-entropy loss before DA\")\n",
    "axes[2].set_xlim(0.0, 15.0)\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res-3.png', dpi=300)\n",
    "fig.savefig('divergences-3.png', dpi=1200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=1, figsize=(7, 7), sharex=True)\n",
    "\n",
    "axes[0].text(0.05, .8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0].transAxes)\n",
    "axes[0].set_title(\"Intra session\")\n",
    "axes[0].hist(intra_session_cross_entropy, 200, density=True)\n",
    "axes[0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "#axes[0].set_yscale('log')\n",
    "\n",
    "axes[1].text(0.4, 0.8, r'$\\mu_1={:.2f}$'.format(np.mean(inter_session_cross_entropy_999)), color='r', transform=axes[1].transAxes)\n",
    "axes[1].set_title(\"Inter-session domain shift\")\n",
    "axes[1].hist(inter_session_cross_entropy_999, 200, density=True)\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy_999), color='r', linestyle='--')\n",
    "\n",
    "axes[2].text(0.45, 0.8, r'$\\mu_2={:.2f}$'.format(np.mean(inter_subject_cross_entropy_999)), color='r', transform=axes[2].transAxes)\n",
    "axes[2].set_title(\"Inter-subject domain shift\")\n",
    "axes[2].hist(inter_subject_cross_entropy_999, 200, density=True)\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy_999), color='r', linestyle='--')\n",
    "\n",
    "axes[2].set_xlabel(\"Cross-entropy loss before domain adaptation\")\n",
    "axes[2].set_xlim(0.0, 5.0)\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res-6.png', dpi=300)\n",
    "fig.savefig('divergences-6.png', dpi=1200)\n",
    "fig.savefig('divergences-6.eps', dpi=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=1, figsize=(7, 7), sharex=False)\n",
    "\n",
    "axes[0].text(0.2, 0.8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0].transAxes)\n",
    "axes[0].set_title(\"Intra-session\")\n",
    "axes[0].hist(intra_session_cross_entropy, 1000, density=True)\n",
    "axes[0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[1].text(0.51, 0.8, r'$\\mu_1^l={:.2f}$'.format(np.mean(inter_session_cross_entropy_1000)), color='grey', transform=axes[1].transAxes)\n",
    "axes[1].text(0.19, 0.8, r'$\\mu_1^d={:.2f}$'.format(np.mean(inter_session_cross_entropy_31)), color='g', transform=axes[1].transAxes)\n",
    "axes[1].set_title(\"Inter-session domain shift\")\n",
    "axes[1].hist(inter_session_cross_entropy_1000, 500, density=True)\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[1].axvline(np.mean(inter_session_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[1].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[2].text(0.3, 0.8, r'$\\mu_2^l={:.2f}$'.format(np.mean(inter_subject_cross_entropy_1000)), color='grey', transform=axes[2].transAxes)\n",
    "axes[2].text(0.06, 0.8, r'$\\mu_2^d={:.2f}$'.format(np.mean(inter_subject_cross_entropy_31)), color='g', transform=axes[2].transAxes)\n",
    "axes[2].set_title(\"Inter-subject domain shift\")\n",
    "axes[2].hist(inter_subject_cross_entropy_1000, 500, density=True)\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[2].axvline(np.mean(inter_subject_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[2].set_xlabel(\"post-DA cross-entropy loss\")\n",
    "axes[2].set_xlim(0.0, 1.0)\n",
    "\n",
    "#fig.text(0.5, 0.04, 'Cross-entropy loss', ha='center', va='center')\n",
    "#fig.text(0.06, 0.5, 'Histogram', ha='center', va='center', rotation='vertical')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res-4.png', dpi=300)\n",
    "fig.savefig('divergences-4.png', dpi=1200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=3, ncols=2, figsize=(7, 7), sharex='col', sharey='row')\n",
    "\n",
    "axes[0, 0].text(0.2, 0.8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0, 0].transAxes)\n",
    "#axes[0, 0].set_title(\"Intra-session\")\n",
    "axes[0, 0].hist(intra_session_cross_entropy, 1000, density=True)\n",
    "axes[0, 0].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0, 0].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[1, 0].text(0.55, 0.8, r'$\\mu_1^L={:.2f}$'.format(np.mean(inter_session_cross_entropy_1000)), color='grey', transform=axes[1, 0].transAxes)\n",
    "#axes[1, 0].set_title(\"Inter-session domain shift\")\n",
    "axes[1, 0].hist(inter_session_cross_entropy_1000, 500, density=True)\n",
    "axes[1, 0].axvline(np.mean(inter_session_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[1, 0].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[2, 0].text(0.3, 0.8, r'$\\mu_2^L={:.2f}$'.format(np.mean(inter_subject_cross_entropy_1000)), color='grey', transform=axes[2, 0].transAxes)\n",
    "#axes[2, 0].set_title(\"Inter-subject domain shift\")\n",
    "axes[2, 0].hist(inter_subject_cross_entropy_1000, 500, density=True)\n",
    "axes[2, 0].axvline(np.mean(inter_subject_cross_entropy_1000), color='grey', linestyle=':')\n",
    "axes[2, 0].set_xlabel(\"Cross-entropy loss of Linear-DA\")\n",
    "axes[2, 0].set_xlim(0.0, 1.0)\n",
    "\n",
    "\n",
    "axes[0, 1].text(0.2, 0.8, r'$\\mu_0={:.2f}$'.format(np.mean(intra_session_cross_entropy)), color='r', transform=axes[0, 1].transAxes)\n",
    "#axes[0, 1].set_title(\"Intra-session\")\n",
    "axes[0, 1].hist(intra_session_cross_entropy, 1000, density=True)\n",
    "axes[0, 1].axvline(np.mean(intra_session_cross_entropy), color='r', linestyle='--')\n",
    "axes[0, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[1, 1].text(0.4, 0.8, r'$\\mu_1^D={:.2f}$'.format(np.mean(inter_session_cross_entropy_31)), color='g', transform=axes[1, 1].transAxes)\n",
    "#axes[1, 1].set_title(\"Inter-session domain shift\")\n",
    "axes[1, 1].hist(inter_session_cross_entropy_31, 500, density=True)\n",
    "axes[1, 1].axvline(np.mean(inter_session_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[1, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "axes[2, 1].text(0.3, 0.8, r'$\\mu_2^D={:.2f}$'.format(np.mean(inter_subject_cross_entropy_31)), color='g', transform=axes[2, 1].transAxes)\n",
    "#axes[2, 1].set_title(\"Inter-subject domain shift\")\n",
    "axes[2, 1].hist(inter_subject_cross_entropy_31, 500, density=True)\n",
    "axes[2, 1].axvline(np.mean(inter_subject_cross_entropy_31), color='g', linestyle='-')\n",
    "axes[2, 1].set_xlabel(\"Cross-entropy loss of Deep-DA\")\n",
    "axes[2, 1].set_xlim(0.0, 1.0)\n",
    "\n",
    "fig.text(0.5, 0.98, 'Intra session', ha='center', va='center')\n",
    "fig.text(0.5, 0.68, 'Inter-session domain shift', ha='center', va='center')\n",
    "fig.text(0.5, 0.38, 'Inter-subject domain shift', ha='center', va='center')\n",
    "\n",
    "#fig.text(0.06, 0.5, 'Histogram', ha='center', va='center', rotation='vertical')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()\n",
    "fig.savefig('divergences_lo-res-5.png', dpi=300)\n",
    "fig.savefig('divergences-5.png', dpi=1200)\n",
    "fig.savefig('divergences-5.eps', dpi=600)"
   ]
  },
  {
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}