figures/ISBI2023SIVA/plot_sim.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import scipy.io as sio\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "datapath=\"/Users/xli77/Documents/MSIVA/results/SIVA\"\n",
    "subspace_struct_list=['234111','2222211','333111','441111']\n",
    "num_iter=11\n",
    "isi_um=np.zeros((4,4,num_iter))\n",
    "isi_ummm=np.zeros((4,4,num_iter))\n",
    "isi_mm=np.zeros((4,4,num_iter))\n",
    "aux_um=np.zeros((4,4,num_iter))\n",
    "aux_ummm=np.zeros((4,4,num_iter))\n",
    "aux_mm=np.zeros((4,4,num_iter))\n",
    "\n",
    "for i,ss in enumerate(subspace_struct_list):\n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"um.mat\"))\n",
    "    isi_um[i,i,:] = np.squeeze(data['isi_um'])\n",
    "    a = np.squeeze(data['aux_um'][1,:])\n",
    "    aux_um[i,i,:] = np.array([a[k][0][0] for k in range(num_iter)])\n",
    "    \n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"ummm.mat\"))\n",
    "    isi_ummm[i,i,:] = np.squeeze(data['isi_ummm'])\n",
    "    a = np.squeeze(data['aux_ummm'][1,:])\n",
    "    aux_ummm[i,i,:] = np.array([a[k][0][0] for k in range(num_iter)])\n",
    "    \n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"mm.mat\"))\n",
    "    isi_mm[i,i,:] = np.squeeze(data['isi_mm'])\n",
    "    a = np.squeeze(data['aux_mm'][1,:])\n",
    "    aux_mm[i,i,:] = np.array([a[k][0][0] for k in range(num_iter)])\n",
    "    \n",
    "    incorrect_ss_list = list(range(4))\n",
    "    incorrect_ss_list.remove(i)\n",
    "    for j in incorrect_ss_list:\n",
    "        data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}_incorrect{subspace_struct_list[j]}\",\"um.mat\"))\n",
    "        isi_um[i,j,:] = np.squeeze(data['isi_um_s1'])\n",
    "        a = np.squeeze(data['aux_um_s1'][1,:])\n",
    "        aux_um[i,j,:] = np.array([a[k][0][0] for k in range(num_iter)])\n",
    "\n",
    "        data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}_incorrect{subspace_struct_list[j]}\",\"ummm.mat\"))\n",
    "        isi_ummm[i,j,:] = np.squeeze(data['isi_ummm_s1'])\n",
    "        a = np.squeeze(data['aux_ummm_s1'][1,:])\n",
    "        aux_ummm[i,j,:] = np.array([a[k][0][0] for k in range(num_iter)])\n",
    "\n",
    "        data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}_incorrect{subspace_struct_list[j]}\",\"mm.mat\"))\n",
    "        isi_mm[i,j,:] = np.squeeze(data['isi_mm_s1'])\n",
    "        a = np.squeeze(data['aux_mm_s1'][1,:])\n",
    "        aux_mm[i,j,:] = np.array([a[k][0][0] for k in range(num_iter)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[42.69163868 42.88364054 42.76165113 42.99218699]\n",
      " [42.64939864 42.2999005  42.85097599 42.8680394 ]\n",
      " [42.71959192 42.85787947 42.63506397 43.0995388 ]\n",
      " [43.09076913 43.23929925 43.17352401 42.9764492 ]]\n",
      "[[42.67687912 42.86470472 42.75052939 43.03843311]\n",
      " [42.65588586 42.22912355 42.62781847 42.76436291]\n",
      " [42.6947989  42.86178266 42.62022519 43.04032558]\n",
      " [42.68928778 42.39724179 41.1201708  39.93684739]]\n",
      "[[23.82415208 23.94737096 23.81949718 24.02788477]\n",
      " [27.76622709 27.44226515 27.80274844 28.16152244]\n",
      " [23.93111632 24.02946237 23.7786396  24.03622762]\n",
      " [17.26454813 18.66018158 17.28986256 17.5637204 ]]\n"
     ]
    }
   ],
   "source": [
    "isi_um_min = np.min(isi_um,axis=2)\n",
    "isi_ummm_min = np.min(isi_ummm,axis=2)\n",
    "isi_mm_min = np.min(isi_mm,axis=2)\n",
    "\n",
    "isi_um_min_ind = np.argmin(isi_um,axis=2)\n",
    "isi_ummm_min_ind = np.argmin(isi_ummm,axis=2)\n",
    "isi_mm_min_ind = np.argmin(isi_mm,axis=2)\n",
    "\n",
    "aux_um_min = np.min(aux_um,axis=2)\n",
    "aux_ummm_min = np.min(aux_ummm,axis=2)\n",
    "aux_mm_min = np.min(aux_mm,axis=2)\n",
    "\n",
    "aux_um_min_ind = np.argmin(aux_um,axis=2)\n",
    "aux_ummm_min_ind = np.argmin(aux_ummm,axis=2)\n",
    "aux_mm_min_ind = np.argmin(aux_mm,axis=2)\n",
    "\n",
    "print(aux_um_min)\n",
    "print(aux_ummm_min)\n",
    "print(aux_mm_min)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use loss to get ISI\n",
    "isi_um_min_loss = np.zeros((4,4))\n",
    "isi_ummm_min_loss = np.zeros((4,4))\n",
    "isi_mm_min_loss = np.zeros((4,4))\n",
    "\n",
    "for i in range(4):\n",
    "    for j in range(4):\n",
    "        isi_um_min_loss[i,j] = isi_um[i,j,aux_um_min_ind[i,j]]\n",
    "        isi_ummm_min_loss[i,j] = isi_ummm[i,j,aux_ummm_min_ind[i,j]]\n",
    "        isi_mm_min_loss[i,j] = isi_mm[i,j,aux_mm_min_ind[i,j]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe4AAADfCAYAAAAwRjDqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAABXQ0lEQVR4nO3dd3gUVdvH8e+9m94hld4SpIOAgIACUgRBimLvoqivvWJ77I+9l0e6vRcUpReR3pXeQi+phEASSpLd8/4xS9hNQrLJLskOnA/XXmRnzs7cC/nt2Zk5MyNKKTRN0zRNMwdLdRegaZqmaZr7dMetaZqmaSaiO25N0zRNMxHdcWuapmmaieiOW9M0TdNMRHfcmqZpmmYiuuP2ESLyuYjsO828niKiRKRPKfOCReSwY35bT9YnIvEi8qGIbBWRYyKSKSKrROQDEQl0ajdPRBZW5P1pmtmJyK2OnCkRaVrK/B5O8/s4TbeIyG0islxEDolInohsF5HvRaRTKctPFBF/EckQkall1NPb0f7WYtO3OqYPqcB7e8HxGj+naREi8qKIbHTUfEhE1onIGBGJc2p32s8u7czQHbf5DQMiHD/fXNmFiEgEsAwYBLwHXAbcBUwFLgeCPStT084aOcBNpUy/xTGvuLeBccB84AZgKPAuEAN0Lm0FSqkC4Fugn4jEn6aOm4E84OeTE0SkK5DkNL9SRMQKzAbuASYAgzHe33dAV6B2ZZetec6v/Caaj7sFyAK2ATeIyCilVGElljMcaAC0U0qtcZr+i4g854U6Ne1s8Stwo4g8pxxXsBKRYIwM/QLcerKhY/q9wEdKqcecljEL+EREytp4+gJ4ALge48t0EREJBa4AJimlcp1m3QIUAnOBQSJSUymVVYn32AO4ABiqlPrdafpk4NVy6tbOMP2Pb2IiUgfoA3wPjAfigUsrubiajr9Ti89QDpVcrqadbb7C+JLb3WnaMIzP01+KtQ0FAiglVwBKKfvpVqKUWg2sp/St+yuAMIzOHQARCQKuBmYCbznWe13Zb+W0Tvt54KjttHVrZ57uuM3tRoz/wy+Bn4DjVH732HLH39+LyKWOb/SappW0G2O3t3OHejMwCXDe+kUplQnsBB4TkbtFpH4F1/UFcL6ItCw2/SZgH8aW9UlDgCiMz4O5jvmV/TxYjbHlPkZEholIjUouRzsDdMdtbrcAW5RSy5RSh4HfgMEiElXRBSml5gPPA92A6cBhEVnpGLRS4eVp2lnuS+AqEQkSkVoYe76+PE3b6zGORX8K7BaR/SIywXlgWhm+AWw4dcAiUhvoDXxdbMv3FuAw8Ltj+tdAJxFpVsH3hlJqB8bx7USMQwMHRWSDiLzlWL9WjXTHbVIicgHQHGO33UlfAEHANZVZplLqJaA+cIdjudEYnfn6MgbIaNq56CcgEGPg5g0Yu5TnlNZQKbUUOA8YALwD7MLoZJeISJlbxEqpFIxd3zc4HVc+uafNeTd5AtAP+Ekpddwx+eT8Wyr43k6uezxQD+P9jXWs8zFgQyl7ALQqpDtu31EIWE8zz+rU5qSTYfxDRKIcW8UrgAw8GE2qlEpVSk1QSt2mlGoE3AfUAR6v7DI17WyjlMrB2MN1E0bevinnePUJpdR0pdRjSqluQAuMzv5dN1b3BUYGL3E8vwlYrpTa7NTmRozPid+dPg9SgX8xBtJV6rNeKXVIKfWtUupupVRzjBHxEcCLlVme5h264/Yd6UCMiASUMu/krqk0AEebk4NO1gCHHI9MIBboKiJJxRdSGUqpTxzLbuGN5WnaWeRLYCDQmtPvJi+VUmor8AMQ7XxO9Gn8jrEL/CYROR9oVcr6ir7Ic+rz4BDQDqjLqU7fI44R5mvQnwfVSp8O5jv+Ap7COF/y52LzrgRSgC2O54MwRn2+CMwr1jYeY5T5zcB/3F25Y1d4RvGtBsfxu0jH+jVNO2UW8COQrZTaUFoDEfEHIpRSB0uZ3Qw4htEpn5ZS6riI/IBxrPw4kI9xPvXJdbTH6MzHYGTfWQDGKVy3YJyX7RYRiQaOOM4nd54eirH7fK27y9K8T3fcvmM2xgfB547BJMuAcOBajNGitzl1qrdgjF59u9g5nACIyMMUO8/UDTcBI0XkG4wR5keBpsCjGB8Un1T6nWnaWUgpZaP8060igV2Ojnc2xkjvaIxcDwDeVEqdcGN1XwAjgTsxzt12Pjf7FkABbyildhZ/oYj8BgwTkbDSPi9OoxfwvuPzYBGQjXEK3P0YGw3u7OLXzhDdcfsIpdTJSxQ+w6mt5XyMY1RFF0EQkViMwH9VRggnYAwm6UHJLfLTmYJxHG0oxkUfIjB2vS8CrnecU6ppWsUcwdgz1hfj3Oo4jK3mDRhXJhznzkKUUotFZBvGVdGKdpM7tuivB/4qrdN2mIAxYHU48LmbdS/FGKB6CXAbUAOj814B9FVKzT39S7UzTfR1NTRN0zTNPPTgNE3TNE0zEd1xa1oVEpH+IrJFRJJF5MlS5l8sIqtFpFBEhhebd4uIbHM8KnVurqZpZ0ZVZlvvKte0KuK449JWjOOd+zCOF16nlNro1KYhxviCx4DJSqmfHdNrAiuBjhgDkVYBHZRSh6ryPWiaVlJVZ1tvcWta1ekEJCuldiil8jFO3XG5Z7JSapdSai1Q/GIelwKzlFJZjkDPAvpXRdGappWrSrOtO25Nqzp1gL1Oz/c5pp3p12qadmZVabZNezpYbsE0n9vH3/j516u7hBJ+uO/a6i6hhAj/wOouoVQdYm+X8tpIvxZl/97N2nQXxvm2J41VSo31sLRzii9mu+unvncZg/9ddXl1l1BCgNX3upROcSPKzTWUk20fy7Xv/Strmi+zlP0Z4Ajz6QK9H+OqUyfVdUxzx36gZ7HXznPztZqmlaeMbJeTa6jibOtd5ZpWERYp+1G2FUCSiDRyXG/+WozLUbpjBtBPRGo47o3czzFN0zRvqHyuoYqzrTtuTasIKedRBqVUIcbd1mYAm4AflVIbROQlERkMxu1aRWQfcBUwRkQ2OF6bBbyM8QGxAnip2GUvNU3zRCVzDVWfbb2rXNMqwurWt+/TUkpNBaYWm/ac088rMHaVlfbaicBEjwrQNK10Jsq27rg1rSLc222maZrZmCjbuuPWtIowT7Y1TasIE2Vbd9yaVhEm+lauaVoFmCjbuuPWtIowUbg1TasAE2Vbd9yaVhFinnBrmlYBJsq27rg1rSKs1V2ApmlnhImyrTtuTasIE+1O0zStAkyUbd1xa1pFmCjcmqZVgImyrTtuTasIEx0H0zStAkyU7bO64168cBNvv/4rNpti6JVduO2OPi7z8/MLee6pr9m0cR+RUSG8/vYt1K4TTXZ2Hk88/Bkb1+/h8qGdGPXMcK/V1CupE/8ddD9Wi4WvV0zho/nfuswPsPrz8VVP07ZOU7KOHmHkdy+yNzsVP4uV9654gta1m+JnsfLjPzP48O9vvFbXhuW7+PHjv7Hb7HQb2Ir+11/gMr8gv5DPX5vBnq3phEYEccfzlxGTEMnOTal8885sAJSCQbd24fyLEr1S05qlO/jygznY7XZ6DWrL4Ju6lKjp01emsHNLKmERwTzw0hBia0WSvPEAE96c4ahJceXt3bmgR1Ov1GSmb+VnM1/MdrcG7Rl18UgsYuHXDTOZuOpnl/n+Vj/+2/cRWsQlcvh4Do9Pe4MDOelcdl5Pbm1/RVG7pjENuea7B9mSudPjmtYt28l3H89D2excNLA1l93QyWV+QX4hE16bzu4taYRGBnP3cwOJqRXJjk0pfPm2I9cohtx6Ie0vSvK4HoC1y3bw1QdzsNsVPQe14fIbS+Z6zH+nsHNLGmERwdz34mBia0WyfWMKE986lesrbu9Gx4u9lGswVbZ95lrlIjLKm8uz2ey8/srPfPjpXfw8+UlmTF3Nju2pLm1++3UpEREh/D7tWW64qScfvvsHAIEBftxz/2U89NiQ0hZdaRax8Mbgh7ju8yfo/v4tXNG2N03jGri0uaHjQA4fy6HzOzcwZtFP/Kf/XQAMbt2LAD9/en54G30/uZObO11OvagEr9Rlt9n57oO/uO/1oTz/+c2smLOFA7sOurRZNHUDIeFBvPzNbfS+qj2TxiwEoE6jaJ4acz3Pjr+RB94cxrfvzsFmK36f+MrV9Nm7s3ji7at46+s7WDx7I/t2Zrq0mffnWkLDg3jvh7sYcE1Hvvt0HgD1GsfyyvhbeO3z2xj1ztVMeGsGtkLPawKMASxlPbQSzpVsP93zHu75/XmGfv1/DGjag8Y167m0uaJFP46cyGPQlyP56p/feajbrQBM3TKPq797gKu/e4BnZr7D/iNpXum07TY733wwl4ffGMbLX9zKsrmbS+R6wdT1hIQF8dq3I+g7vD0/j10AQJ1GMfxnzA28MOEmHn7zCr58Z7ZXMmS32fni3dk8/vZVvPHVCJbM3sT+Yrn+e8o6QsODeOf7kfS/uiM/jJ4HQN3GMbw07mb++9mtPPH2VUx8a6b3cg2mynW1ddwi8qPT4yfgDm8uf8O63dSrH0PdejH4+/vRb8D5zJu7zqXN33PXMWiIsWXZu19bli/bhlKK4JBAzm/fmIBA7+6QaF+3OTsP7mf3oRQKbIVMWjuX/s27u7Tp37wbP6w2vlX+sf5vLmrSHjC+YYb4B2O1WAnyC6TAVkjOiTyv1LVrcypxtSOJrR2Jn7+VCy5pytpF213arF20nQsvbW68jx5JbF69F6UUAUH+WK3Gr1FBfqHXdjclb0ohvm4U8XWi8PO3cmGf5qxauM2lzcqF27hoQCsAOvdsxvpVu1FKERjkj9XPuSavlGQQKfuhnZPZbhXflD3ZKew/kkahvZDp2+bTq7HrlmTPxl2YvGkOALOSF9K5XtsSyxnQtAfTt873Sk07NqcSVyeK2NpGhjpd0ox/iuX630Xb6dq/BQAdezRl06o9pWTIhnjpd3v7phTi60QR56ipS+/mrFqY7NJm9YJtdO9v5LpTz/PYUEpN+fmF3o+biXJdnbvKjyiligItIp96c+Hp6YeJT6hR9Dw+Por163a7tMlwauPnZyUsLIjs7Dxq1AjzZilFEiJj2H84veh5yuEM2tdrfto2NruNnON51AyJ5I/18+jfohvrnvqVYP9AnpvyCdnHcrxS16HMPGrEhRc9j4oNZ+cm1y2YbKc2VquF4LBA8o4cJywymJ0bU/jyzVlkpeVw69OXFnXkHtWUkUN0XETR85qx4SRvTCnWJpfokzX5WQgJDSTn8DEiokJI3nCAMa9NJTPtCP/37KCiwHvMRLvTqtE5l+34sGjScjOKnqflZtI6/rzTtrEpO7n5R4kKiiD7+JGiNpc2vYgH/3jFKzVlZ+RSM/ZUrmvEhrGzlAydbGP1M3Kde/g44VHB7NiYwmdvzuRg6hHueKa/VzJ0KCOXmk6fNTVjw9m+6YBLm6zM3KLsn8x17uFjhDtyPf71aWSmHeHuZwd6L9dgqmxXZ8f932LPn6mWKkyifd3m2O122rx2BVHB4Uwe+RHzk1ey+1BK+S8+wxq1qMXzn99Myu4sPn99Bq06N8Q/oHqHTyS2rM1bX9/B/l2ZfPrfqbTt4qWtLBOFuxrpbFdC6/imHC84QXLW7vIbV4HGLWrx8ue3cGD3QSa+Np3WnRrh7+U9FRWV2LI2r381gv27DjL21Sm06ezFvScmynaV7SoXkTki0tJpUmsReVZEOkPRPUnLW8ZIEVkpIisnjp9WZtu4uEjSUg8VPU9LyyY2LtKlTaxTm8JCG7m5x4mKCnX7PVVU6uFM6kTGFT2vFRlLypHM07axWqyEB4WSdfQwV7Trw9ytyym028jMy2b57vW0rdvMK3XViAnlUPqprffsjBxqxLj+O0Q5tbHZ7BzLPUFoRJBLm1oNahIU7M+Bna7H0SpVU2w4B9NPbYlkZeRQMzasWJswDp6sqdDO0bwThEcGu7Sp0zCGoGB/9u3MwCs8uB/32UpnG9JyDxIfFlv0PD4shvS8g6dtYxULYQEhLlvb/ZtezLStf3utpqjYMLIyTuX6UEYuUU5b4GBk6GQbW6GR67BI11zXbhBNYHBAiWPRlVEjNowsp8+arIwcasS41lQzJqwo+ydzHVYi10ZNXss1mCrXVXmMu65SagOAiHQFvgLqA5+JyDB3FqCUGquU6qiU6nj7HQPKbNuiVX327slk/76DFBQUMnPaP/To1cqlTY9erfjz9xUAzJm5hgs6J3ntWE5p/tm/mcYxdalfIwF/qx/D2lzCjE2LXNrM2LyIa9pfCsDlrXqwcMc/AOzPTqO743h3iH8QHeq3IDnDO9/MGzRLIH1/NpkphykssLFi7lbadG3i0qZN1yYsmbEJgNV/b+O88+shImSmHC4ajHYw9Qipew4RnRBRYh0V1aRZLVL3HiL9QDaFBTaWzN5Eh26uo9U7dEtiwbT1ACybt5mW7esjIqQfyC4atJKRepgDu7OISYgssY7KsFikzMc56pzP9oa0rTSIqk2diHj8LH70T7qYeTuWubSZt3MZg5v3BqBvYneW71tbNE8Q+iVdxDQvHd8GaHReAmn7sslw5Hr53M2069rYpU27rk1YPH0jACv/3kozR4YyUg4XZSgz9Qgpe7KI9kKGGjerReq+U7leOmcT7bu75vr87oksnG7kevm8LbQoJdeZqYdJ2X2QWC/lGsrOtq+pyv0eR5x+vhkYrZQaJSJxwGRgkjdX5udn5Ymnr+S+u0Zjs9kZMqwzTRJr8enHU2nRsj49erViyBVd+M9TXzNkwCtERobw6ls3F71+UL8Xycs9QUFBIfPmruOTsffQuIlno7htdhtPTn6fH257G6tY+HbVVLak72JUn9v5d99mZmxezDcrp/LJVc+w7NFvOHQ0h7u+fxGAiUt/44Mrn2T+g58jIny/ahobU3d4VM9JVquFax7oxYdPTMJuV3Qd0JLajaKZPHEJDc6Lo223JnQb2JLPXp3Bf274jJCIIO74z2UAJK87wIxvV2D1syAW4bqHepX4dlypmvws3PpIX15/5EfjtJGBranbOJafxi+gcbMEOnRPouegNvzv5T95+JoxhEYEc/8LgwHYsnYfk79eip+fFbEItz3al4ioEI9rArD6YIh9gM62svPqvNF8OuQlrBYLv22YxfasPfxf5xvYmL6NeTuXM2nDTF7t9yh/3jyWw8dzeWL6G0Wv71CnFWm5Gew/kuZRHc6sfhZueLAX7z3+C3a7ovuAVtRpFMNvExfR8LwE2nVrwkWXtWLcq9N46voJhEYEcddzAwHYtm4/075dgdVq5PrGh3oTHuWdXN/8cB/eevQn7HbFxQNbU7dRDL+MX0CjZgm0755Ej4FtGP3KFB69dixhEUHc68j11rX7+fObX7D6WRGBWx7pR7iXcg3myrYopapmRSLfAb8A84G1wBVKqcWOef8opc6vyPJyC6ZVTeEV0Pj516u7hBJ+uO/a6i6hhAj/wOouoVQdYm8vN7mBD3Qt8/fuxIeLzZN+LzkXst3100+qu4QS/nfV5dVdQgkBVt+7NEinuBFuZbKsbPtarqtyV/nDwF3AfmC1U7D9gTMz1FPTvMwiUubjHKWzrZmemXJdZV+PlFKpQF8RsSilnM+a7wX8VVV1aJonLD5zySLfobOtnQ3MlO0q369RLNgopWYCM6u6Dk2rjDM5wMnsdLY1MzNTtn3vgISm+TBfHGGqaZrnzJRtE+0c0LTqZ5WyH+URkf4iskVEkkXkyVLmB4rID475y0SkoWO6v4h8ISLrRGSTiDzl9TenaecwT3INVZtt3XFrWgWISJmPcl5rBT4BBgAtgOtEpEWxZiOAQ0qpROA94OQ5Q1cBgUqp1kAH4K6Twdc0zXOVzbXjtVWabd1xa1oFWCxlP8rRCUhWSu1QSuUD3wPFb1M1BPjC8fPPQG8xPjkUECoifkAwkI/r+dOapnnAg1xDFWdbd9yaVgHlXTnN+dKdjsdIp5fXAfY6Pd/nmEZpbZRShcBhIBoj6HlACrAHeNudS4lqmuYeD3INVZxtPThN0yqgvAEsSqmxwNgzsOpOgA2oDdQAFojIbKWUdy6fp2nnuLKyfQZzDZXIttsdt+OGAb2BOIptqSulHqhUuZpmMh6OO90P1HN6XtcxrbQ2+xy7ziKBg8D1wHSlVAGQLiKLgI6Axx23zrammSvbbu0qF5HHgCXArUA7oLXTo9VpX6hpZxkPbzKyAkgSkUYiEgBci3Etb2eTgVscPw8H5irjusR7gEsARCQU6AJs9vT96GxrmsHDm4xUabbd3eJ+EHhAKfWxm+017azkyY0IlFKFInIfMAOwAhOVUhtE5CVgpVJqMjAB+EpEkoEsjA8AMEasfiYiGzA2Dj5TSq0tuZYK09nWNMyVbXc77ghgasXfjqadXTy9uJJSairFsqSUes7p5+MYp4cUf11uadO9QGdb0zBXtt0dVf4d0L8iC9a0s9FZeD9unW1N4yy5H7eIPOL0dC/wooh0w7htX4FzW6XUu2emPE3zLb54p6CK0tnWtJLMlO2ydpXfX+x5LtDV8XCmgCoP96Bfv6nqVZZrz8vPld+oin2zdUV1l1CCeDp+8wzpEFt+GxNluyw+ne3nl82r6lWWa/m9vje4fuKmldVdQgkX1S5+6rJ5mCnbp+24lVKNqrIQTTMDi9X81yzS2da0ksyUbXdPB3tOREJKmR4sIr63malpZ8jZdoxbZ1vTDGbKtbtfMZ4HwkqZHuKYp2nnBE9uMuKjdLY1Dc9uMlLV3D0d7OSF0Is7H+N8NE07J5hpd5qbdLY1DXNlu8yOW0RyMEKtgB0i4hxwKxAEjD5z5Wmab7G4easgX6ezrWmuzJTt8ra478P4Rj4ReAbjbiYn5QO7lFJLzlBtmuZzxDzZLo/OtqY5MVO2y+y4lVJfAIjITmCx4yLomnbOMtO38rLobGuaKzNl291j3OuA8NMdpNf3BdbOFWKicLtJZ1vTMFe23e24Myl9AMtJVi/Uomk+z0y709yks61pmCvb7nbcvYo998cYdXoP8KxXK9I0H2am3Wlu0tnWNMyVbbc6bqXU36VMni0iO4A7gG+9WpWm+SgznTLiDp1tTTOYKdvubnGfzr/AxV6oQ9NMwRevonSG/IvOtnYOMVO2K91xi0gY8BDG3YU07Zzgi1dR8jadbe1cZKZsu9VxO12soWgSxiUR84AbzkBdXtEpoQ33nX8TVrEwZcc8vt38h8t8f4sfT3W+m/NqNOJwfg4vLf6Y1KOZADSOrMejHW8nxD8YpRR3z3qOfLvnZ8wsWrCBN177CbtNMWx4V0bceanL/Pz8Ap558gs2bdhLZFQob747gjp1osnOzuXRh8axYd0eBg/rwtPPXuNxLc52rN7HnPHLUHZFm75N6XJlG5f5hQU2prw/n7TtBwkOD2TwYz2JjA8vmn8kI5cJ90+i27Xt6DS0tRdrWordrmjbtyldrmxbak2p2zMJDg9kyGO9StQ0/v5f6Xbt+XT2Uk1m2p3mDrNmO31dGhu+W4dSivoXNSDxsqYu820FNv6dsJrDu7MJCA2g/d0dCYkJ5WhmHvOenUNYgnGV16jGNWlzczuv1LRowUbeeO0X7DY7w4ZfyIg7+7nMN7L9lVO2b3NkO49HH5rAhnW7GTysM08/e7VX6gHY9c9+/p64HGVXtOydxAVXuOagsMDGzA8Xkr7jIEHhgVz2SA8i4k5dAfdIRi5fP/Q7na9uS4chrbxS0z9LtvPZ+7Ow2xS9B7dl2M2uN6UryC/ko5f+YMfmVMIig3nklaHE1Ypi24YDjHljKgBKwdUjLqJzz/O8UhOYK9vubnHfV+y5HcgAlimlDlVmxSLSF7ga+EQp9a+IjFRKja3MskpjEeHBDrfw2LzXyTiWxei+L7HowCp2HzlQ1Oayxj3Jzc/jhqmPckm9Loxsey0vLfkYq1h4pss9vLpsNNuz9xAREEahKvS4JpvNzquv/MCY8Q8QHx/F9de8Qc9ebWiSWKuozaRfFhMREcKfM15k2tSVvP/OJN569w4CAvy59/7LSd52gOTkFI9rcWa32Zk9ZilXv3gp4dEhfPn4HyR2qk9MvaiiNutmbSUoLJCRo4ezacEO5n25kiGPnxrXNHfichq3r+vVmmaNWcI1L15KeHQoXzw+2VFTjaI2a2dtJSgsgLtGX8XGUmqa4+WawFy709xkumwru2L9N2vo/Gg3gmsEs+DlecS3SyC8dkRRm70LduMf4s8lr/Vl/7J9bPp5Ix3uvgCA0NhQLn7hEm+VA5zM9k+MGX+vI9tv0bNX62LZXuLI9vNMm7qK99/5nbfevZ2AAD/uvX8gydtSSE4+UMZaKsZuszNv3FKGPdePsOgQvh81hcYX1CPaKdcb5mwjMCyAWz+5gi0Ld7Lwq1Vc9miPovkLPl9Jg/O9d6tOm83O+Hdm8NwH11EzLoInb/+MjhclUa/RqXvqzvljDaHhQXz88z0snLWBrz/5i0deGUb9JrG8MfF2rH4WDmXm8ujN4+nYPQmrn3c6XDNlu9x3LCJ+QCgwSyn1hePxlVJqemWD7XA78Dhwo4hcArTzYFklNKvZhP05aaTkZVBotzF3z1K61eng0qZb7fZM37UAgL/3LadDfEsAOia0Zkf2XrZn7wHgSH4udlXWGTPuWb9uF/Xqx1K3Xgz+AX70H9CBeXPXuLT5a+5aBg/tAkDffuezfOkWlFKEhATSvkMigYH+HtdRXMq2TKJqhROVEI7V30rz7o1JXrbHpc225Xto1SsRgPO6NmTP2hSU499k29LdRMaHu3wgeKemCKISIopq2lZqTUkANOvakN1rDxTVtHXpbqLiw1y+fHjD2XSTEbNmO3vHIULjwgiNDcXiZ6FOp7qk/ZPq0ibt31Tqda0PQK2OtcnclFH0u3EmrF+3m3r1Y4ple51Lm7/mrmPw0M4A9O3XjuVLtzpluwmBgZ4OOXKVlpxJZEIEkY5cN+3eiB0rXI9+7Fi+lxY9mwCQdGED9q47levty/YQERfm1VwnbzxAQt0axNepgb+/lW59WrBi/jaXNisWbKXnZcaegQt7NWfdyl0opQgM8i/qpPPzC/F22syU63I7bqVUIfAWxmki3pSjlMpWSj0G9AMu8ObCY4NrkHHs1LUjMo5mERtcw7VNSA0yjhptbMpObsFRIgPCqBeegELx5sVPMLbfK1zbbKBXakpPyyYh4VQNcQk1SEs/fNo2fn5WwsKDyc7O88r6Tyc36yjhMaFFz8OjQ8jJyivRJsLRxmK1EBgSwLGcE+QfK2DZpHV0u6adV2vKycorWp9RUyi5WUeL1ZRXVHfJmtbS7ZrzvVrTyfWU9TATs2b7WPYxgmoGFz0PqhHEsexjLm2OHzrVxmK14B/sR0FuPgBHM48y/4W/WPzGAg5uzfRKTSWzHUVaenaxNodJSIgCqibbxXMdVjOE3IOu68vLOkqYS4b8Oe7I0Mrf1tP5atfDU57KysghJu7UnpHouHCyMnJKtok32lj9LISEBZJz2Pj/3bphPw9dP5ZHbxzHyCcGeG1rG8rOtq9x9yveUqADsNuL655y8gel1JMicr8Xl+0Rq1hpHdOUu2c9x3FbPu/2fIqtWbtYnb6hukvzOYu+/4eOl7ckINj7ewIqa+EZrElMtDvNTedUtgMjg+j91qUEhAWQvSublR8vo8fLl+DvQ7+/vmDZj/9y/qAWPpVrgKYt6/D+tyPZtyuTj1/6g/MvbEKAl/ZUmCnb7r7jccDbIlIfWIUxcKWIUmp1eQsQkTnAA0qpk72fEpFnMXbTLVNKfeTGMkYCIwGS7uhE7T5Jp22bcewQscE1i57HhtQk45jr3r+Mo4cc07OwioUw/xAO5+eScTSLNRlbOJyfC8DSlDUk1WjocccdFx9FauqpGtJTDxEfF1lqm/iEGhQW2sjNOUZUVGjxRXlVWM0QcjJP/ZfmHDxKeM3QEm2OZBpbuHabnRNH8wkODyRlayZbFu9m3hcrOZGXj1jAz99K+4EtPKopvGYoR1xqyiOsZkixmkLJyTS2zF1rymDL4l0laurgYU1gjJ3whIj0Bz7AuCLZeKXU68XmBwJfYnSmB4FrlFK7HPPaAGOACIxj0RcopY57VJAPZrvP4/1oM7jdadsGRwVzPOvUFvbxQ8cJjgp2aRNUw2gTXDMYu81OwbFC/MMCEBGs/sbF4KIaRhESF0JeWi5RDV33xlVUyWxnEx8XVaxNJKmp2VWW7eK5zs06Sli06/pCa4aQm5lHePTJDBUQFB5I6rZMti3ZzcKvTmZI8PO30vay5h7VVDM2nMz0I0XPD6bnUDM2vGSbtCNEx0VgK7RzNPcE4ZGu/791G8YQFBLAnh0ZJDavhTeYKdtl7gMQkYkiEoFxEYaGwLvA38BKp8cKN99X3ZPBFpGuwFdAfeAzERnmzgKUUmOVUh2VUh3L6rQBtmTtoG54AgmhsfhZrFxSvwuL97t+Bi0+sJr+DS8CoEfdTqxO2wjA8tS1NI6qR6A1AKtYaBfbjN1H9rv5Nk+vZasG7Nmdzr59mRTkFzJ92ip69HIdvd2zVxsm/7YUgFkz/6FT5/PO+DGWWkkxHEo5QnZaDrYCG5sW7iCxUz2XNomd6rP+r2QAtizeRf3WtRARrn/tMu4edxV3j7uKDpe3oMvwNh532qdqOlyspvoubZI61WP9X8bxsc1ONd3w2kDuGXc194y7mo6Xt6DL8LZe6bQBrBZLmY+yiIgV+AQYALQArhOR4oWNAA4ppRKB94A3HK/1A74G7lZKtQR6ApU+zcGXs11Wpw0Q2SiKvLRcjmbkYS+0s3/5PuLbJbi0iW+XwN7FxpiIlJUHiGkWg4hwIucEym4cw83LyCMvLY+QGM87z5at6rNnd0axbLuO4O7ZqzWTf1sGwKyZ/9Kpc9Mzmu34xBiyU45w2JGhrQt30rij62DNxhfUY+O87QBsW7Kbeq0SEBGuemUAt48ezu2jh3P+oBZccEVrjzttgMTmtUnZe4i0A9kUFNhYNHsjF1zk+lnesXsS86Ya4wOW/LWJVh0aICKkHcjGVmgHICPlMPt3HySuVmSJdVRWZXMNVZ/t8ra4bwGeBBqVW3n5jjj9fDMwWik1SkTigMnAJC+so4hN2flg9Re81eMJLGJh2o6/2XVkP7e1upItWTtZfGA1U3f8zdNd7uaby97hSH4uLy35GIDcgqP8tGUao/u+BCiWHljD0pR/Pa7Jz8/KU89cwz13fozdbmfosAtJTKrNJx/9QcuWDeh5SRuGXdmVZ0Z9zqBLnyciKoQ33x5R9PoBfZ4lN/c4BQU2/pqzhtHj7ncZtVpZFquFPnd24acXZ6JsitZ9koipX4MF364mITGGpE71adMniSnvL2Ds3T8TFB7I4Ed7erze8mrqe+eF/PjijKKaYkvU1JQ/35/PmLt/Mk5RO8M1Afh5dlnETkCyUmoHgIh8DwwBNjq1GQK84Pj5Z+BjMT7d+wFrlVJrAJRSBz0pBBNn22K10PKGNix7bzHKrqjXvQHhdSLY8tsmIhtGkdCuFvUuasC/41Yx96lZ+If60/4u4zB71pZMtvy+GYtVQIQ2N7UlICzA45qMbF/FPXf+D7tdMXRYFxKTavHJR1No2bI+PS9pzbArL+SZUV8y6NIXHdm+rej1A/o878h2IX/NWcfocf/ncbYtVgs97+jMby/PRtnttLgkiej6NVjy3T/EJ0bT+IL6tOydxIwPF/D5vb8SFBbAgId7lL9gD1j9LNzxaD9eeeh77HY7lwxqS73GsXw/9m+aNK/FBRc1pffl7fjwxcncN/xTwiKCePjloQBsXrOXSV8twc/Pgohw52OXEhEVUvYKK8BM2ZayRlqKiB1IUEqlV/x9lFjWd8AvwHxgLXCFUmqxY94/SqkKjSTq+cONZ26IaCVNH35b+Y2q2Ddb3d1oqjri9fGg3nF781HlFjbw9xFl/t5NGTLhtMsQkeFAf6XUHY7nNwGdlVL3ObVZ72izz/F8O9AZuBFjF1scEAt8r5R6s9w3dfpafDbbjy4c5XPZ/u+Fvau7hBImblpZ3SWUcFFt75065i2ta97i1gdOWdkuK9dQ9dl25yuGt0L0MHAXsB9Y7RRsfyCsrBdqmq/wEynzISIjRWSl02Okt1YNdMe4KEp3YJiIeNqb6GxrmkM15RoqkW13BqellnccRilV7q3/lFKpQF8RsSil7E6zegF/uVGHplW78o53OS40crqLjewHnAcP1HVMK63NPsexr0iMgSz7gPlKqUwAEZkKtAfmVPAtONPZ1jSHsrJdTq6hirPtTsc9Esh2o51bigUbpdRMYKa3lq9pZ5KHI09XAEki0ggjxNcC1xdrMxnj+PMSYDgwVymlRGQG8ISIhAD5QA+MAS6e0NnWNAczZdudjvsPbxwH07SzgVUqP4BFKVUoIvcBMzBOGZmolNogIi8BK5VSk4EJwFcikgxkYXwAoJQ6JCLvYnxAKGCqUmpKqStyn862pjmYKdvlddw+N0hE06qThyNPUUpNBaYWm/ac08/HgatO89qvMU4b8QadbU1zYqZsl9dx++bwX02rJp5epMGHnDVvRNO8wUzZLrPjVkr53kVaNa0aWU10WcSy6GxrmiszZdu7t6PRtLOcJ8fBNE3zXWbKtu64Na0CPD0OpmmabzJTtnXHrWkVYKbjYJqmuc9M2dYdt6ZVgJmOg2ma5j4zZfu0HbeIXOzuQpRS871Tjqb5Nj8THQc7HZ1tTSvJTNkua4t7Hsa5nie/hpw877P4czBOONe0s96ZvsVqFZmHzramuTBTtsv6ihHLqbuVDAK2YNyyL9HxuBnYDAw+wzVqms/ws1jKfJiEzramFWOmXJ92i9v5nqAi8jLwoFJqllOTHSKSDrwJeHrpRU0zhfJuMmIGOtuaVpKZsu3u4LQWGHcwKW4/0Mx75bjvvd6+d3/cEbN+re4SSpjYd3h1l1DCN1uXV3cJlWaeaLvN57J9TdOm1bHaMnX88IPqLqGElQ88WN0llDB+44rqLqGE1jXda2embLtb6wbgeREJPjnB8fNzjnmadk6wWixlPkxIZ1vTKDvbvsbdLe57gD+B/SKy1jGtNWADBp6JwjTNF1lNNIDFTTrbmoa5su1Wx62UWiEijYEbOLX77BvgW6VU3pkqTtN8jZku0uAOnW1NM5gp225fgMUR4rFnsBZN83lm+lbuLp1tTTNXtt3uuEWkLnAxxmkkLjv9lVLverkuTfNJfia6upK7dLY1zVzZdqvjFpEbgIlAIZCB6wUaFKDDrZ0TzHSRBnfobGuawUzZdneL+yXgHeA/SinbGaxH03yan4nC7SadbU3DXNl2t+OOB8brYGvnOjPds9dNOtuahrmy7W6lU4HOZ7IQTTMDi5T9MCGdbU3DXLl2d4t7FvCGiLQE1gEFzjOVUr53yTBNOwN88WIMHtLZ1jTMlW13O+4xjr+fLmWeQt9BSDtHeHrKiIj0Bz7AyMx4pdTrxeYHAl8CHYCDwDVKqV1O8+sDG4EXlFJve1SMQWdb0zBXtt36iqGUspTx0MHWzhl+FinzURYRsQKfAAMwrhF+nYi0KNZsBHBIKZUIvAe8UWz+u8A0r7wZdLY17aTK5hqqPtvm2TegaT5AyvlTjk5AslJqh1IqH/geGFKszRDgC8fPPwO9xXGeiogMBXairyGuaV7nQa6hirPt7nncj5Q131cv0vDv0h188f5s7HY7l1zeliE3XegyvyC/kE9e/pOdW1IJiwzmwZeGEFcriuSNBxj3xnQAFIrht3enU4/zvFJTm5gW3NR8OBYszNu3iD92znKZ7yd+3NPmZhpG1Ce3II+P1kwg81gWAPXCajOi1XUEW4NR2PnPkjcpsBd6XNPCBRt447UfsdvsXDG8GyPu7O8yPz+/gGee/JyNG/YQGRXKW+/eQZ06MWRn5/LoQ2NZv243Q4Z14elnr/O4Fmc7Vu9jzvil2O2Ktn2b0uXKti7zCwtsTHl/PqnbMwkOD2TIY72IjA8vmn8kI5fx9/9Kt2vPp/PQ1l6pycOLNNQB9jo930fJgWFFbZRShSJyGIgWkePAKKAv8JgnRTgza7bXLtvBVx/MwW5X9BzUhstv7OIyvyC/kDH/ncLOLWmERQRz34uDia0VyfaNKUx8awYASimuuL0bHS/2zt3IujXowJM9R2K1WPhl/UwmrPjJZb6/1Y/XLn2UFvGJZB/L4bGpr3PgSDoDm/Xktg5XFrVrGtuQq755kC0ZOzyuadGCjbzx2i/YbXaGDb+QEXf2c5lvZPsrNm3YS2RUKG++ext16kSTnZ3How9NYMO63Qwe1pmnn73a41pO2v3PfuZ/tgJlV7TonUjHYa7ZtBXYmPnRQjJ2ZBEUFkj/Ry4mIi6saH5ORi7fPDyZTle1pf2Qll6ry0zZdvcY9/3FnvsDtYBjQDqVuEiDiNwNdMUY1XojMEUp9WlFl3M6dpudie/M5Jn3ryU6Lpyn7/icDt2TqNsopqjNX3+uJSw8iA9+vJvFszfy7f/m8dDLQ6nXOJZXJ9yK1c/CocxcRt0ykQ7dkrD6ebaDQhBubXE1r634iKzj2bx84ROsTl/H/rzUojY9615IXsFRHl3wAl0SOnBd06F8tGYiFrHwf21v5dO1X7AnZz9h/qEU2j0/g8dms/PqK98xdvyDxMfX4LprXqNnrzY0Saxd1ObXXxYRERHClBkvM23qCt5/ZxJvvXsnAQH+3Hv/YJK3HSA5eb/HtTiz2+zMGrOEa168lPDoUL54fDKJneoTU69GUZu1s7YSFBbAXaOvYuOCHcz7ciVDHu9VNH/OxOU0bl/Xq3WVdxxMREYCI50mjVVKeeNyoi8A7ymlcr18oQhTZvuLd2cz6r2rqRkbznN3fkn7bonUccr231PWERoexDvfj2TJ7E38MHoe9704hLqNY3hp3M1Y/SxkZ+by9G2fc37XRI+zbRELz15yD3f++iypOZn8cP17/LV9KTuyTn2WX9HyUo6cyOWyz+5kQNOLeaT7bTw29Q2mbJ7HlM3zAEiKbsCHg//jlU7byPZPjBl/L/HxUVx/zVv07NWaJom1itpM+mUJEREh/DnjeaZNXcX77/zOW+/eTkCAH/feP5DkbSkkJx/wuJaT7DY788YvY+hzfQmrGcIPT06lccd61KwXVdRmw5xtBIUGcvPHw9i6cCeLvl7FgEd6FM1f8MVKGrSr47WaTior22cw11CJbLt7jLtRsUddoDYwH3i0ksVeAtwC3KSUGgS0Lad9hSRvSiGhbg3i60Th52+la+8WrFywzaXNygXbuPgy49te557N2LBqN0opAoP8i4JckF+Itz4nm0Q1JO1oBhnHDmJTNpamrqJDfBuXNh3i2zD/wDIAlqf9Q8toY0u/dXRz9uTsZ0+O0UHmFuShXC5yVTnr1+2ifv046taLxT/Aj/4DLuCvuWtd2sybu5bBQ429FX37tWfZ0s0opQgJCaR9h0QCA92+cq7bUrZlElUrgqiECKz+Vpp3b8y2ZXtc2mxbvodWvZIAaNa1IbvXHkAp499k69LdRMWHEeP0geANFpEyH0qpsUqpjk4P53DvB+o5Pa/rmEZpbUTED4jEGMjSGXhTRHYBDwFPi8h9nr4fM2Z7+6YU4utEEVfbyHaX3s1ZtTDZpc3qBdvo3r8VAJ16nseGVXtKZDvfi9lundCUPdkH2Hc4lUJ7IdO2zOeSJq57AS5p0pnfN84BYOa2hXSuX/Kf5bJmPZi2Zb5Xalq/bjf16sdQt16MI9sdmDd3nUubv+auY/BQY8Owb792LF+61SnbTbye7bTkg0QlhBMZH47V30rTbg3ZsWKvS5udK/bSrGcTABIvbMC+dalFud6+fA8RcWHUrBfp1bqg7GyXk2uo4mxX+mumUioNeAZ4s5KLOKiM/43RjucnKltLabIycoiOO7XbtGZcOFkZOadtY/WzEBwaSM7hYwBs23CAx24Yz+M3T2DE45d6/I0coGZgFAePHTq1/uPZ1AiMcmlTIzCKLEcbu7JztPAYYf6h1AqNAwWjOt7LKxeOYlCjPh7XA5CWdoj4hFNbsfEJUaSnHyrWJruojZ+flbDwYLKzz+yNo3Ky8oiICS16Hh4dSm7WUZc2uVl5hDvaWKwWAkMCOJZzgvxjBSybtJZu15zv9bqsImU+yrECSBKRRiISAFwLTC7WZjJGpwcwHJirDBcppRoqpRoC7wOvKqU+9tobc+Lr2T6UkUtN52zHhnMos1i2M3OJjosAjGyHhAaS68h28oYDPHnTBJ6+9TNue6yfV7IdFxZNak5m0fO03EziwqJLaZMBgE3ZyT1xlKigCJc2/ZtezNQtf3tcD0B6WjYJTtmOS4giLT27WJvDJCREAVWT7byso4Q55TosOqSUXB8jPCYEMHIdEOLPcUeuV/+2nk5XefV7YBEPcg1VnG1Pv05ZMK68VBkfACil/nA896nzRZNa1ubtb+5g/65M/vfKFNp1aULAGdiydJdFLDSt0Zj/LHmTfFs+T3d6gJ2H97Iha0u11eSrFn7/Dx0vb0lAsL/Xl+3JuZ6O41r3ATMwThmZqJTaICIvASuVUpOBCcBXIpIMZGF8AFSHszbbiS1r8/pXI9i/6yBjX51Cm86NqzXbJ7VOOI9jhSdIPri7ukvxSct/XEO7QS3OSK7BXNl2d3DaFcUnYRwHuxdY4OYy5gAPKKVOjpprKiLDgVlKqWVKqXK/ZjofZ3jmnZu58uaep21bMzacg+mnvoVnpedQMza81DbRcRHYCu0cyztBeGSwS5s6DWMICvZn744MmjSvhSeyTmQTHXzqG3DNoCgOnch2aXPoRDY1g2uQdSIbi1gI8QsmtyCPrOPZbD6UTG6B8W3434wNNIyo53HHHR9fg7TUU1vYaanZxMXVKNYmirTUQyQk1KCw0EZuzjGiokKLL8qrwmuGciTz1Df/nIN5hNUMcWkTVjOUnExjy9xus3PiaD7B4YGkbM1gy+JdzPtiJSfy8hEL+Plb6TCw+NkZFefptplSairGsV/nac85/XwcuKqcZbzgYRlFfDHbT751E8Nu7nHatjViw8hyznZGDjViimU7JoyD6UeoGReOrdDO0bwThJXIdjSBwQHs25lB42aeZTs99yAJ4aeOsceHxZCee7CUNrGk5R7EKhbCAkPIPn6kaP6A8y5m2mbvbG0DxMVHkeqU7fTUbOLjooq1iSQ11dijVhXZDq0ZQq5TrnMPHi0l18HkZB4lLNrIdf7RAoLCA0ndlkny0t0s+mqVI9eCNcBK2wHNiq+mUsyUbXdr/bnY40fgWeAf4A43l1H3ZLBFpCvwFVAf+ExEhrmzAOfjDGV12gBNmtUidV8W6QeyKSywsXjORjp0T3Rp06F7IvOnGsd8ls3bTMsODRAR0g9kYyu0A5CRepgDu7OIreX5MZUdh3eTEBJHbHA0VrHSJaEDq9JdjzmtTl/HxbWNY06d4s9nw8GtAKzN3Ei9sNoEWPyxiIXmNZLYn5ficU0tWzVg9+509u3LpCC/kOnTVtCzl+tx95692jD5tyUAzJq5mk6dzzvjd9KplRTDoZTDZKflYCuwsWnhDhI71Xdpk9SpHuv/MsYtbF68i/qtayEi3PDaQO4ZdzX3jLuajpe3oMvwtl7ptMH4Vl7Ww4R8LttlddoAjZvVInXfoaJsL52zifbFsn1+90QWTl8PwPJ5W2jRvn6JbGemHiZl90FiEzzP9vrUrdSvUYc6EfH4WfwYcN7F/LVjmUubv3YsY0iL3gD0S+rOsr2nxpIIwqVNuzNtq3eObwO0bFWfPbsznLK9ih69XEdw9+zVmsm/GXXOmvkvnTo3PaPZjk+MJjslh8OOXG9dtItGF9RzadOoYz02z9sOQPKS3dRtlYCIMPyV/tz66ZXc+umVtBvYnI7DWnut04ays+1r3NriVkp5o/IjTj/fDIxWSo0SkTiMff+TvLCOIlY/C7c93I9XH/kBu03Ra1Ab6jWO5cdx82ncrBYdL0qi16C2fPLyHzx49WjCIoJ54EXjtLvNa/cx+aulWP0siEW4/bF+RESFlLPG8tmVnc83/siojvdiEQt/71vC/twUrkwcyM7De1idsY55+xZzT5tbeOeiF8gryOOjNRMBOFp4jGm75vLyhaNQKNZkbODfDM9P5/Xzs/L0M9dwz50fYrPbGTqsK4lJtfnko8m0aNmAXpe0ZdiV3Xh61GcMvPQ/REaF8Obbpz7P+/d5mtzc4xQU2Jg7Zw1jxj3gMiK9sixWC33vvJAfX5yBsila90kitn4NFny7moTEGJI61adNn6b8+f58xtz9E8HhgQx+tKfH6y23LhPdQcgdZs32zQ/34a1Hf8JuV1w8sDV1G8Xwy/gFNGqWQPvuSfQY2IbRr0zh0WvHEhYRxL0vDAZg69r9/PnNL1j9rIjALY/0I9wL2bYpO6/O/ZQxV7yMVSxM2jCL7Qf3cO+FN7IhbRvzdizj1/Uzea3/Y0y9bRyHj+fw+NRTQwg61m1Fak4m+w6nlrGWivHzs/LUM1dxz53/w25XDB3WhcSkWnzy0RRatqxPz0taM+zKC3lm1JcMuvRFIqJCePPt24peP6DP845sF/LXnHWMHvd/LiPSK8NitdDjjk5MfmU2druixSWJRNeLYun3/xLXJJrGF9SjRe8kZn24kC/vm0RgWAD9H77Y038K92ozUbbl5Gi9M74ike+AXzBGq64FrlBKLXbM+0cpVaGRRP9kflY1hVfA2ytXVncJJUzsO7y6Syjhm63Lq7uEUt3efFS5yZ26+6Myf+8ua3C/edLvJd7O9vL0CT6X7du/8anD9ACsfODB6i6hhPEbV1R3CSXc1/oZtzJZVrZ9Ldduf9sWkYEiMl9EMkUkQ0T+FpHLKrCuh4G7MIbEr3YKtj8QVtYLNc1XWC1S5sOMdLY1rexs+xp3B6fdAfwP+IZTl2y7CJgkIvcopSaWtwylVCrQV0QsSim706xewF8VK1vTqoeZdqe5Q2db0wxmyra750CMAh4pdm7ZBBFZBTwJlBvuk4oFG6XUTGCmu6/XtOpkce+6xWais61pmCvb7u4qrw9ML2X6NKCB98rRNN9msUiZDxPS2dY0ys62r3F3i3sPxgXQk4tN7wfoqwVo5wyr+N6pIR7S2dY0zJVtdzvut4GPRKQ9sNgxrRtwEyVvUqBpZy3zRNttOtuahrmy7e553GNEJB3jpgMnr7S0CbhaKfX7mSpO03yNxQcvxuAJnW1NM5gp2+V23I5TOv4LfKKU6n7mS9I03+XmDQdMQWdb004xU7bL/YqhlCoA/g9MNORO084QD+8O5lN0tjXtFDPl2t19AzMw7rGraec0KeePCelsaxplZ9vXuDs4bQ7wqoi0AVYBLjdsVUr53vUANe0M8MWrKHlIZ1vTMFe23e24T16c4YFS5imM+49q2lnPTKeMuElnW9MwV7ar8u5gmmZ65vlO7h6dbU0zmCnb7m5xa5oGPnlvXk3TPGembJfZcYtIKHC5Uup7x/P/AUFOTWzAQ0qpvNJer2lnG18cYVoZOtua5spM2S5vi/s2jBGn3zue3wQsB446nrcF7gXeLPnSMyunIL+qV1muLXtSqruEEr7euqy6SyjhxqadqruESvPFEaaV5LPZ3n74UFWvslwpqTnVXUIJV/35Y3WXUMLESwdVdwmVZqZsl9dxXwu8WmzanUqpHQAich3wENUQbk2rDma69V85dLY1zYmZsl1ex50IbHB6no2xC+2klUBzL9ekaT5LTBTucuhsa5oTM2W7vI47Egg++UQpVa+U1/t7uyhN81Vm2p1WDp1tTXNipmyXN4xuL9C6jPltHW007ZxgESnzUR4R6S8iW0QkWUSeLGV+oIj84Ji/TEQaOqb3FZFVIrLO8benVzvT2dY0J57kGqo22+V13FOAF0QkqPgMx6jU5x1tNO2cYBFLmY+yiIgV+AQYALQArhORFsWajQAOKaUSgfeANxzTMzFGgbcGbgG+8vCt6GxrmpPK5hqqPtvl7Sp/Dbga2CIiHwNbHdObAfdhdPyvlbcSTTtbeLgzrROQ7DQA7HtgCLDRqc0Q4AXHzz8DH4uIKKX+cWqzAQgWkUCl1IlK1qKzrWlOzJTtMjtupVS6iHQFRgOvc+q9KWAm8H9KqXR335mmmZ07377LUAfX3c/7gM6na6OUKhSRw0A0xrfyk64EVnvQaetsa1oxZsp2uVdOU0rtBgaISE2MkahgfLPIKu+1mna2Ke94l4iMBEY6TRqrlBrrrfWLSEuMXWz9PF2WzramnVJWts90rh3rcDvbbl/y1BHm5R7UpWmmV97IU0eYTxfo/YDz6O26jmmltdknIn4Yo78PAohIXWAScLNSanuFiz99zTrb2jmvrGyXk2uo4myb5+KsmuYDPBxVvgJIEpFGIhKAcRGUycXaTMYYoAIwHJirlFIiEoUxWOxJpdQi770jTdPA41HlVZpt3XFrWgVYyvlTFqVUIcbArxnAJuBHpdQGEXlJRAY7mk0AokUkGXgEOHlayX0Yu7OfE5F/HY+4M/EeNe1cVNlcQ9VnW98dTNMqwNOrKymlpgJTi017zunn48BVpbzuFeAVj1auadppmSnbuuPWtAow02URNU1zn5myfVZ33OuX7eT7j+dht9m5aGBrBtzgeleqgvxCJr42nd1b0giLDGbkcwOJqRXJzk0pfPn2bEcrxeW3Xkj7i5K8UtOFdc/nsa63YxELv22ezRdrJrnM97f48WKvB2ke05jDJ3J4avY7pORm0D/xYm5qM6SoXVJ0A2789TG2Htzllbp2rN7HnPHLUHZFm75N6XJlG5f5hQU2prw/n7TtBwkOD2TwYz2JjA8vmn8kI5cJ90+i27Xt6DS0rAtyuW/hgg288dqP2G2KK4Z3Y8Sdl7rMz88v4Jknv2Djhj1ERoXy1rt3UKdONNnZuTz60DjWr9vNkGFdePrZa71SD4DFRJdFPJttW7WX6WOXYLcr2vc7j4uuaucyv7DAxqR353EgOZOQ8ECGj+pNDaff1+z0XD75v5/oeX0Hul3RBm+4pGknXhv8ABax8PWKKXww7xuX+QFWf/53zTO0rdOUQ0ePMOLbF9h7KBU/i5UPho+iTe2m+Fmt/LBqOu8Xe21ltY9ryZ2tr8OChVl7FvDztmku8/0sfjzSfgRNIhuQU5DLmyvGkH7sIAANI+pyb9ubCPELwo7ikb9focBe6HFNSxdt4YM3JmO3KwYNu4CbRvRymZ+fX8grz/zAlk37iYgM4aU3r6dWnZqsWLKVTz+YTmGBDT9/K/c+fBkdOieeZi0VZ6ZsV9sxbhG5zPEYKCKTROQyby7fbrPz7QdzefCNYbz0xa0sn7uZA7sOurRZOHU9IWFBvPrtCPoMb88vYxcAULtRDM+OuYHnJ9zEg29ewdfvzMZWaPe4JotYGNX9Th6Y9gpX/fQglyZeRKOoui5thjTrQ86JXIb9cC/frvuD+zvfDMD05Pnc8Ouj3PDrozz31wccyEn3Wqdtt9mZPWYpVz3XjxEfDWPTgh1k7s12abNu1laCwgIZOXo4HQe3ZN6XK13mz524nMbtXd+LJ2w2O6++8j2fjrmP3/54jmlTV7A92fW2qb/+spiIiBCmzHiJm265hPffMb4EBQT4c+/9l/Po41d4rZ6TPL3k6bmgKrI99dNF3PBif+7933DW/72d9D2utwJdPXMLQaEBPDjuGroMac3sz10Hzc8Yv5SkDsUvz155FrHw5tCHuXri43R992auaNub8+IauLS58YKBZB/L4YK3rufThT/y/IC7ARjSphcBfv5c9P6tXPLhHdzSeTD1aiR4XhPC3W1u4IUl73Pv3P9wcZ1O1Auv5dKmX/3u5Obncdecp/l9+yxubTm86P080v4OPlnzFff+9TxPL3wLm91W2moqxGaz8+6rv/H2/27n60mPMHv6GnZuT3Np8+ekFYRHBPPDn09wzY3d+fR948tGZFQob354K1/+8jDPvnw1Lz/zg8f1ODNTrqtzcNqLGJeGiwFCHH97zc7NqcTWiSK2dhR+/lYuuKQZ/y5yHWX/76LtdO1vXJWuQ4+mbF61B6UUgUH+WP2Mf5qCfBt46T+uZWwiew+nsD8njUJ7ITO3L6RHQ9e9AD0aXMCfW/8CYM6OJXSqU3Lr9dLEi5i5faFXagJI2ZZJVK1wohLCsfpbad69McnL9ri02bZ8D616Gd9uz+vakD1rU1BKGfOW7iYyPpzoelFeq2n9ul3Urx9L3Xqx+Af40X9AR/6au8alzby5axg8tAsAffu1Z9nSzSilCAkJpH2HRAIDvX+PDCnnjwac4Wzv35pBzVoR1EyIwM/fSquLm7Bl6W6XNluW7qJd76YAtOjeiB1r9hf9vm5asosaCeHE1q/htZra12vOzoP72Z2VQoGtkElr5jCgRXeXNgNaduf7VdMBmLzuby5ObA9g/M76B2G1WAnyDyTfVkjO8TyPa0qq0YiUvHTSjmZSqGzM37+czgntXNp0rtWOOXsXA7DowCraxjQD4PzYluw6so9dR/YBkFOQhx3lcU2b1u+lbr1o6tSNxt/fjz7927Jw3kaXNgv/2sCAwR0A6Nm3NauWJ6OUomnzOsTERQDQKDGeEycKyM/3fA/ASWbKdXV23D2AcOA4sEEp9aU3F56dkUvN2FO7xmrEhpGdkVOiTQ1HG6ufheCwQHIPHwdgx8YUnrv1C1687UtufKR3UUfuibjQaNLyTm31p+cdJC605mnb2JSd3PyjRAaGu7Tp16QbM5K913HnZh0lPCa06Hl4dAg5WXkl2kQ42lisFgJDAjiWc4L8YwUsm7SObte081o9AGlp2cQnnPpgjU+oQXp69mnb+PlZCQsPJjvb8w+8snhyrfJzyBnN9pGDeUTEhhU9j4gJ5cjBvGJtjhIRa/y+Wq0WgkICOHrkBCeOFbDo5zX0uK69N0uiVmQM+7NPXWjuwOEMakXGuraJiOHAYaONzW7jyPE8aoZEMnndPI4WHGfjM5NY89RPfDL/e7KPuX5WVUZ0UA0yj53aE3Hw2CGig2qcto1d2ckrPEZEQBh1wuIBxYsXPsT7Pf7DFYn9Pa4HICP9MHEJUUXPY+MiyUg7XKzNEeISIgEj16FhQRzOPurSZt7sdTRtXoeAAO8d7TVTrqvtGLdS6ijwvIj0AY6W176qNW5Ri5c+v4WU3QeZ+Np0WndqhH9g9Q8JaBmbxPHCE2w/tKf8xlVg0ff/0PHylgQEnxt3gPTF3Wa+xpezPe/bVXQZ2opAH/p9bV+vOTa7nZb/HUZUcDhT7vmYv5NXsjsrpfwXnyFWsdCiZiKPzP8vJ2z5vNL1UZKzd7E2c3O11XTSjuRUPn1/Gu+NvsOryzVTtqvsq4SIzHFc0u3k88Ei8iyQo5R61s1ljBSRlSKycvLXC8psGxUbRpbTFvahjFyiYsNLtDnkaGMrtHMs9wRhka43S6rVIJrA4AD278zEU+l5B4kPjS56HhcaTXpe1mnbWMVCWEAIh0+ceh+XJnb36tY2QFjNEHIyT22x5Bw8SnjN0BJtjjja2G12ThzNJzg8kJStmcz7YiWj7/yJVX9sZOnPa1k9xXXXV2XEx0eRlnpqayEt9RBxcVGnbVNYaCM35xhRUa51e5+U8zj3eDvbc75fWmbbiOhQjmTkFj0/kplHRHRosTYhHMkwfl9tNjvHj+YTEhHI/i3pzPpsOe/d/h1LJ69nwY//suyPDW6/19NJOZxJnahTp97Wjowl5XCGa5sjmdSONNpYLVYigkLJOnqY4e36MnfLMgrtNjLzslm2ax3t6jbzuKaDxw8RE3xqCzs6uAYHjx86bRuLWAj1C+ZIfi6Zxw+x/uA2juTncsKWz8q0dTSJcj1mXxmxcZGkp2YXPc9IP0xsfGSxNhGkpxpb4YWFNvJyjxMZFQJAelo2Tz/8Fc++cg116kXjXebJdVXuA6irlNoA4Li5wVdAfWCiiAxzZwFKqbFKqY5KqY6Db7yozLYNz0sgfV82GSmHKSywsWLuZtp2bezSpl3XJiyebnQyq/7eynnt6yMiZKQcLhqMdjD1CKl7sohOiCyxjoramJFMvcha1A6Pw8/iR78m3Zm/e4VLm/m7VzCoqTHKsnfjC1mxf13RPEHo07irV49vA9RKiuFQyhGy03KwFdjYtHAHiZ1cB+4kdqrP+r+SAdiyeBf1W9dCRLj+tcu4e9xV3D3uKjpc3oIuw9vQfmDxu9lVXMtWDdi9O519+zIpyC9k+rSV9OzlOvq3Z682TP7N+JCfNXM1nTqfd8ZP6RCsZT7OUV7Ndu9ru5TZtnbTWA4eOMKh1CMUFthYP38753Wu79LmvM4N+HeOccOzjQt30qhNbUSE298czMMTr+PhidfRZXArLrq6HZ0vb1naairkn32baRxdl/o1auFv9WNY295M2+R6EazpGxdxbQdjl/Pg1j1YsH01APuy07jIcbw7xD+IjvVbsi3d9Zh9ZWzL3kXt0HjiQ2LwEysX1+nE8lTXcSLLUtfQu15XALrV7lC0Rb06fQMNI+oQaA3AIhZaxTRlb84Bj2tq1rIue/cc5MC+LAoKCpk9fQ3dejR3adOtZwumTV4FwLxZ62jfqQkiQs6RYzx+3+fc8+AA2pzf0ONaijNTrqty3+8Rp59vBkYrpUY5rhAzGeM6rV5j9bNw/YO9eP/xX1B2RbcBrajTKIbfJy6iwXkJtOvWhO6XtWLCq9N4+voJhEYEMfK5gQAkr9vPtG9XYLVasFiEGx7qTXhUsMc12ZSdtxaN56MBz2G1WJi8ZQ47Du3lrg7XsilzO/N3r+D3LXN4qdeDTLrmE46cyOXpOe8Wvb59rRak5R5kf05aGWupOIvVQp87u/DTizNRNkXrPknE1K/Bgm9Xk5AYQ1Kn+rTpk8SU9xcw9u6fCQoPZPCjPb1aQ3F+flaefuZa7rnzI2x2O0OHdSUxqTaffPQHLVrWp9clbRl2ZTeeHvU5Ay99jsioEN58e0TR6/v3eYbc3OMUFNiYO2cNY8Y9QJPEWmWs0V2+F2IfULXZtlq47O6ufPXcNJRdcX7f84hrUJO5X6+kdlIszTo34Px+5zHpnXl8cOcPBIcFMnzUJd4soQSb3cao39/npxFvY7VY+HbFVLak7eLJvrfz774tTN+0iK9XTOHTa55hxePfkn0shzu+fQGACUsm8dFVT7LokS8QhG9XTmVj6g6Pa7IrO6PXfsuLFz6ERSzM3rOIPTkHuKHZELZl72J56hpm7V7AI+3vYEzvV8ktyOPNlWMAyCs4ym/bZ/Huxc+ggJVp61iZtq7sFbrBz8/KI08N4ZF7JmC32xk49AIaJyYw/pOZNGtZl+49WzBo2AW8/MwPXDPoTSIignnhzesB+OX7xezfk8lnY2fz2VjjdN33Pr2DGtFhZa2yAsyTbTk50vKMr0jkO+AXYD6wFrhCKbXYMe8fpdT5FVne/JQxVVN4BTzyx4zqLqGEey7qVH6jKnZjU9+rCSDQekm5m+sn7PPL/L0LtFzse/vVzjBvZ/u7bW/7XLbvG1/8stPVr2vXptVdQgkTLx1U3SWUEBs01K1MlpVtX8t1Ve4qfxi4C+MOKaudgu0PeOsrk6adUXpXeal0tjXTM1Ouq2xXuVIqFegrIhallPPVTHoBf1VVHZrmGd8LcXXT2dbODubJdpWf31Qs2CilZgIzq7oOTasMX7wYg6/Q2dbMzEzZrv4TkzXNVMzzrVzTtIowT7Z1x61pFeCLx7s0TfOcmbKtO25NqxDz7E7TNK0izJNt3XFrWkWIeb6Va5pWASbKtu64Na0CpFrvy6Np2plipmzrjlvTKsQ838o1TasI82Rbd9yaVgFm+lauaZr7zJRt3XFrWoWYJ9yaplWEebJtnko1zQeIWMt8lP966S8iW0QkWUSeLGV+oIj84Ji/TEQaOs17yjF9i4hc6t13pmnnNk9ybby+6rKtO25Nq5DK349bjE+AT4ABQAvgOhEpfg/UEcAhpVQi8B7whuO1LYBrgZZAf+B/4u4niqZpbqj8/birOtu649a0CrGW8yhTJyBZKbVDKZUPfA8MKdZmCPCF4+efgd5i3GR8CPC9UuqEUmonkOxYnqZpXlHpXEMVZ1t33JpWAR7eHawOsNfp+T7HtFLbKKUKgcNAtJuv1TStkjy8O1iVZtu0g9MurnWX1y5zIyIjlVJjPV3OypF3eaMcwHs1eZOuCaB2mb93IjISGOk0aayv/Zv5uuuSHvNKtr35u3HdG495YzE+mSHwzbp8Kdu+lmu9xW0YWX6TKqdrco9P1aSUGquU6uj0cA73fqCe0/O6jmmU1kZE/IBI4KCbr9Vc+dTvhoMv1gS+WZfP1FROrqGKs607bk2rOiuAJBFpJCIBGANSJhdrMxm4xfHzcGCuUko5pl/rGJnaCEgClldR3Zqmla1Ks23aXeWaZjZKqUIRuQ+YgTHiZaJSaoOIvASsVEpNBiYAX4lIMpCF8QGAo92PwEagELhXKWWrljeiaZqLqs62GB3+uU0f33GPrkkzE1/83fDFmsA36/LFmnyF7rg1TdM0zUT0MW5N0zRNMxHdcWuapmmaieiOW9M0TdNM5JzsuEWkloh8LyIrRWSriPyla9I1aebmi78buibz1uTLztXTwb4CximlfgAQkdbVXA/omtzlizVpvsEXfzd0Te7xxZp81jk3qtxx15UTQF2lVGp11wO6Jnf5Yk2ab/DF3w1dk3t8sSZfd87tKnec2D4bWCMiY0Skm65J16SZmy/+buiazFuTrzvnOm6HAcCVGHdnmS4iQ0WksYhMEJGffaimoSIyToybr/fzkZqai8hoEflZRO7xhZoARCTUcXxsUDXUpPkGnevK16RzbSLn3K7y4kRkHLBXKfWS4/nPSqnhPlZTDeBtpdQIH6rJAnyplLrRF2oS49KCucBGpdSf1VWT5ht0ritdk861CZxzW9wicqkYF4FHROKA7sAsH6/pWeATX6lJRAYDU4CpvlCTiPTFuM5velXWo/kOnWvPa9K5No9zcVT5cOB/IpKLMSDiP0qpJb5Yk4gI8DowTSm12hdqAnBcMH+yiEwBvq3umkTkv0Ao0AI4JiJTlVL2KqxLq3461x7UBDrXZnLO7yo/SUSigf8CfYHxSqnXqrkkROQBjNvArQD+VUqNruaSEJGewBVAILBWKVWlWwxlEZFbgUy9S007SefaPTrX5qI7bk3TNE0zkXPuGLemaZqmmZnuuDVN0zTNRHTHrWmapmkmojtuTdM0TTMR3XFrmqZpmonojlvTNE3TTER33JqmaZpmIrrjNgkRUeU8Pvdg2S+IyHovlqtpmpt0trWKOhcveWpWtZx+HgSMKzbtWNWWo2mal+hsaxWit7hNQimVevIBZJcy7WIRWSUix0Vkp4j89+SF+wFE5AoRWSsix0QkS0T+FpF4x+UEnwdaOn3Dv7Ua3qKmnZN0trWK0lvcZwERuRT4BngQmA/UB0ZjXHf4MRFJAL4HngJ+AcKALo6X/wC0wvim39Mx7XBV1a5p2unpbGul0R332eEZ4C2l1GeO59tFZBTwtYg8DtQG/IGflVK7HW2Kjns57spT6Ph2r2ma79DZ1krQHffZoQPQyRHokyxAMJAArAFmA+tFZKbj55+VUhlVXqmmaRWhs62VoI9xnx0swItAO6dHGyAJyFBK2YB+jsdaYASwTUTaVkOtmqa5T2dbK0FvcZ8dVgPNlFLJp2ugjPu3LgGWiMhLwAbgGoxv7PmAtSoK1TStQnS2tRJ0x312eAn4U0R2Az8ChRiDUjoppZ4QkS5AH2AGkAacD9QDNjpevwtoICLtgT1AjlLqRNW+BU3TSqGzrZWgd5WfBZRSM4CBQC9guePxJEZQwRhJ2g34E9gGvAO8rJT62jH/F2AqMAfIAK6rsuI1TTstnW2tNGLsZdE0TdM0zQz0FremaZqmmYjuuDVN0zTNRHTHrWmapmkmojtuTdM0TTMR3XFrmqZpmonojlvTNE3TTER33JqmaZpmIrrj1jRN0zQT0R23pmmappnI/wMQ4plLXvnzrwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 504x230.4 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "label=[\"$S_1$\",\"$S_2$\",\"$S_3$\",\"$S_4$\"]\n",
    "fig,axes=plt.subplots(1,2,figsize=(7,3.2))\n",
    "\n",
    "sns.heatmap(isi_um_min,cmap=\"YlGn\",annot=True,fmt=\".2f\",vmin=0,vmax=0.1,xticklabels=label,yticklabels=label,ax=axes[0])\n",
    "axes[0].set_title(\"UA ISI\",fontsize=16)\n",
    "axes[0].set_ylabel(\"Ground Truth\",fontsize=14)\n",
    "axes[0].set_xlabel(\"Test\",fontsize=14)\n",
    "\n",
    "sns.heatmap(isi_ummm_min,cmap=\"YlGn\",annot=True,fmt=\".2f\",vmin=0,vmax=0.1,xticklabels=label,yticklabels=label,ax=axes[1])\n",
    "axes[1].set_title(\"MSIVA ISI\",fontsize=16)\n",
    "axes[1].set_ylabel(\"Ground Truth\",fontsize=14)\n",
    "axes[1].set_xlabel(\"Test\",fontsize=14)\n",
    "\n",
    "plt.tight_layout()\n",
    "# plt.savefig(\"simulation.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "datapath=\"/Users/xli77/Documents/MSIVA/results/SIVA\"\n",
    "subspace_struct_list=['234111','2222211','333111','441111']\n",
    "\n",
    "corr = np.zeros((4,8,12,12))\n",
    "\n",
    "for i,ss in enumerate(subspace_struct_list):\n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"A.mat\"))\n",
    "    A=np.squeeze(data['A'])\n",
    "\n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"um_W.mat\"))\n",
    "    W1=np.squeeze(data['W1'])\n",
    "    \n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"ummm_W.mat\"))\n",
    "    W2=np.squeeze(data['W2'])\n",
    "\n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"mm_W.mat\"))\n",
    "    W3=np.squeeze(data['W3'])\n",
    "\n",
    "    data=sio.loadmat(os.path.join(datapath,f\"subspace_struct_{ss}\",\"W.mat\"))\n",
    "    W=np.squeeze(data['W'])\n",
    "\n",
    "    corr[i,0] = W1[0]@A[0]\n",
    "    corr[i,1] = W1[1]@A[1]\n",
    "    corr[i,2] = W2[0]@A[0]\n",
    "    corr[i,3] = W2[1]@A[1]\n",
    "    corr[i,4] = W3[0]@A[0]\n",
    "    corr[i,5] = W3[1]@A[1]\n",
    "    corr[i,6] = W[0]@A[0]\n",
    "    corr[i,7] = W[1]@A[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x633.6 with 32 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrabsmax = np.max(np.abs(corr))\n",
    "analysis_list = [\"UA\"]*2+[\"MSIVA\"]*2\n",
    "fig,axes=plt.subplots(4,4,figsize=(10,2.2*4))\n",
    "\n",
    "for i in range(4):\n",
    "    for j in range(4):\n",
    "        sns.heatmap(np.abs(corr[i,j]),cmap=\"Reds\",vmin=0,vmax=corrabsmax,ax=axes[j,i])\n",
    "        axes[j,i].set_xlabel(f\"$S_{i+1}$  $M_{j%2+1}$\", fontsize=14)\n",
    "        axes[j,i].set_ylabel(f\"$S_{i+1}$  $M_{j%2+1}$\", fontsize=14)\n",
    "        axes[j,i].set_xticks([])\n",
    "        axes[j,i].set_yticks([])\n",
    "        if i == 0 and j == 0:\n",
    "            axes[j,i].set_title(\"UA\", loc=\"left\", fontsize=18, fontstyle=\"italic\", fontweight=\"bold\")\n",
    "        if i == 0 and j == 2:\n",
    "            axes[j,i].set_title(\"MSIVA\", loc=\"left\", fontsize=18, fontstyle=\"italic\", fontweight=\"bold\")\n",
    "\n",
    "plt.tight_layout()\n",
    "# plt.savefig(\"simulation_absWA_red.pdf\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7.13 ('nhpseg')",
   "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": "3.7.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "30c86d7e87117feea1cc08b035a94da7db90798447807219a06605c175b520ce"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}