real_graphs/results_arxiv.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e0b9624a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seml\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f042ff26",
   "metadata": {},
   "source": [
    "# cora_inductive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "6a6abe85",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_exp():\n",
    "    results = seml.get_results('xy', to_data_frame=False)\n",
    "    dfs = []\n",
    "    for x in results:\n",
    "        df = pd.DataFrame(x['result']['result_df'].copy())\n",
    "        df.loc[:,'model'] = x['config']['model_label']\n",
    "        df.loc[:,'dataset'] = x['config']['dataset']\n",
    "        df.loc[:,'seed'] = x['config']['seed']\n",
    "        df.loc[:,'projection'] = x['config']['projection']\n",
    "        #df.loc[:,'do_lp'] = x['config']['do_lp']\n",
    "        df.loc[:,'lp_alpha'] = x['config']['lp_params']['alpha']\n",
    "        df.loc[:,'lp_num_layers'] = x['config']['lp_params']['num_layers']\n",
    "        df.loc[:,'model_accuracy'] = x['result']['model_accuracy']\n",
    "        df.loc[:,'dense_split'] = x['config']['dense_split']\n",
    "        df.loc[:,'n_per_class'] = str(x['config']['n_per_class'])\n",
    "        df.loc[:,'sample'] = 'frac'\n",
    "        dfs.append(df)\n",
    "    results = seml.get_results('xyz', to_data_frame=False)\n",
    "    for x in results:\n",
    "        df = pd.DataFrame(x['result']['result_df'].copy())\n",
    "        df.loc[:,'model'] = x['config']['model_label']\n",
    "        df.loc[:,'dataset'] = x['config']['dataset']\n",
    "        df.loc[:,'seed'] = x['config']['seed']\n",
    "        df.loc[:,'projection'] = x['config']['projection']\n",
    "        #df.loc[:,'do_lp'] = x['config']['do_lp']\n",
    "        df.loc[:,'lp_alpha'] = x['config']['lp_params']['alpha']\n",
    "        df.loc[:,'lp_num_layers'] = x['config']['lp_params']['num_layers']\n",
    "        df.loc[:,'model_accuracy'] = x['result']['model_accuracy']\n",
    "        df.loc[:,'dense_split'] = x['config']['dense_split']\n",
    "        df.loc[:,'n_per_class'] = str(x['config']['n_per_class'])\n",
    "        df.loc[:,'sample'] = 'equal'\n",
    "        dfs.append(df)\n",
    "    df = pd.concat(dfs)\n",
    "    k = [f'json://{x}'for x in range(40)]\n",
    "    v = [str(x) for x in range(40)]\n",
    "    return df.rename(columns = dict(zip(k,v)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "98add9b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cd2facdb0ea94fceb13b58cc818bf616",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=15.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f5b509d5053c4688ba39f15249c7e899",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=15.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "50bff7586e0b40ad81b1232cb43d9034",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=15.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb0388eb7a4e447182fbe7cb2bf613b8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=15.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "df = load_exp()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2a2a5376",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_min_max_cols(df, clip = None):\n",
    "    if not clip is None:\n",
    "        df[[str(x) for x in range(40) if str(x) in df.columns]] = df[[str(x) for x in range(40) if str(x) in df.columns]].clip(upper=clip)\n",
    "    df['max'] = df[[str(x) for x in range(40) if str(x) in df.columns]].max(axis=1)\n",
    "    df['min'] = df[[str(x) for x in range(40) if str(x) in df.columns]].min(axis=1)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "c350d4e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "from cProfile import label\n",
    "from matplotlib.ft2font import LOAD_IGNORE_GLOBAL_ADVANCE_WIDTH\n",
    "from scipy import stats\n",
    "\n",
    "def filter_df(df, filter_dict):\n",
    "    df_copy=df.copy()\n",
    "    for k, v in filter_dict.items():\n",
    "        df_copy = df_copy[df_copy[k]==v]\n",
    "    return df_copy\n",
    "\n",
    "\n",
    "def get_style(label: str):\n",
    "    color_dict = {\n",
    "            \"MLP\": 'r',\n",
    "            \"GCN\": 'tab:green', \n",
    "            \"APPNP\": 'lime', \n",
    "            \"SGC\": \"b\",\n",
    "            \"GAT\": \"slategrey\",\n",
    "            \"GATv2\": \"k\",\n",
    "            \"GraphSAGE\": \"lightsteelblue\",\n",
    "            \"LP\": \"wheat\"\n",
    "        }\n",
    "    linestyle_dict = {\n",
    "            \"LP\": '--'\n",
    "        }\n",
    "    use_color=\"\"\n",
    "    linestyle=\"-\"\n",
    "    for key, color in color_dict.items():\n",
    "        sep_labels = label.split(\"+\")\n",
    "        if sep_labels[0] == key:\n",
    "            use_color = color\n",
    "            if len(sep_labels) == 2 or sep_labels[0] == \"LP\":\n",
    "                linestyle = \"--\"\n",
    "    return use_color, linestyle\n",
    "\n",
    "def compute_means_per_degree(df, filter_dict, type='min', max_degree=10):\n",
    "    # filter model\n",
    "    df = filter_df(df, filter_dict)\n",
    "    # get means per degree\n",
    "    degrees = np.sort(df[df['degree']<=max_degree]['degree'].unique())\n",
    "    values = []\n",
    "    deviations = []\n",
    "    for deg in degrees:\n",
    "        degree_df = df[df['degree']==deg]\n",
    "        means = []\n",
    "        seeds = degree_df['seed'].unique()\n",
    "        for seed in seeds:\n",
    "            if type=='all':\n",
    "                means.append(degree_df[degree_df['seed']==seed][['0','1','2','3','4','5','6']].mean().mean())\n",
    "            else:\n",
    "                means.append(degree_df[degree_df['seed']==seed][type].mean())\n",
    "        deviations.append(stats.sem(means))\n",
    "        values.append(np.mean(means))\n",
    "    return degrees, values, deviations\n",
    "\n",
    "def plot_1(df, type='min', models=['GCN', 'APPNP', 'GCN+LP', 'APPNP+LP'], max_degree=10, plot_lp_alphas=[0.7], filter_dict={}, clip=None):\n",
    "    df = filter_df(df, filter_dict)\n",
    "    df = compute_min_max_cols(df, clip)\n",
    "    fig, axs = plt.subplots(figsize=(4*1.62, 4))\n",
    "    for model in models:\n",
    "        filter_dict['model']= model\n",
    "        color, linestyle = get_style(model)\n",
    "        if 'LP' in model:\n",
    "            for alpha in plot_lp_alphas:\n",
    "                filter_dict['lp_alpha'] = alpha\n",
    "                model_label = f'{model} ($\\\\alpha={alpha}$)'\n",
    "                degrees, values, deviations = compute_means_per_degree(df, filter_dict, type=type, max_degree=max_degree)\n",
    "                plt.errorbar(x=degrees, y=values, yerr=deviations, label=model_label, capsize=5, color=color, linestyle=linestyle, linewidth=2.5, marker=\"o\",markersize=8)\n",
    "        else:\n",
    "            degrees, values, deviations = compute_means_per_degree(df, filter_dict, type=type, max_degree=max_degree)\n",
    "            plt.errorbar(x=degrees, y=values, yerr=deviations, label=model, capsize=5, color=color, linestyle=linestyle, linewidth=2.5, marker=\"o\",markersize=8)\n",
    "    plt.xticks([x+1 for x in range(max_degree)])\n",
    "    plt.legend(fontsize=14)\n",
    "    #plt.title(f'Average Robustness to Structure Changes per Node Degree')\n",
    "    plt.xlabel('Node degree', fontsize=14)\n",
    "    plt.ylabel('$\\max_{c\\\\neq c*}$ $N_c(v)$', fontsize = 14)\n",
    "    #plt.ylabel('Number of Structure Changes', fontsize=12)\n",
    "    plt.grid()\n",
    "\n",
    "    #plt.xlabel('Node degree', fontsize = 15)\n",
    "\n",
    "    return fig\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "3eb75f27",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_1(df, \n",
    "            type='max', \n",
    "            models=['GCN', 'LP', 'GCN+LP'], \n",
    "            max_degree=8, \n",
    "            filter_dict={'dataset': 'ogbn-arxiv', 'projection': False, 'sample': 'frac'}, \n",
    "            plot_lp_alphas=[0.7])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "8ce5c119",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_1(df, \n",
    "            type='max', \n",
    "            models=['GCN', 'LP', 'GCN+LP'], \n",
    "            max_degree=8, \n",
    "            filter_dict={'dataset': 'ogbn-arxiv', 'projection': False, 'sample': 'equal'}, \n",
    "            plot_lp_alphas=[0.7])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a928e222",
   "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></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>model_accuracy</th>\n",
       "      <th>acc std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th>sample</th>\n",
       "      <th>model</th>\n",
       "      <th>lp_alpha</th>\n",
       "      <th>lp_num_layers</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">ogbn-arxiv</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">equal</th>\n",
       "      <th>GCN</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.701581</td>\n",
       "      <td>0.006488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GCN+LP</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.712912</td>\n",
       "      <td>0.006063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LP</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.669829</td>\n",
       "      <td>0.012142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">frac</th>\n",
       "      <th>GCN</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.482383</td>\n",
       "      <td>0.009206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GCN+LP</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.493914</td>\n",
       "      <td>0.012471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LP</th>\n",
       "      <th>0.7</th>\n",
       "      <th>10</th>\n",
       "      <td>0.468674</td>\n",
       "      <td>0.010687</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 model_accuracy   acc std\n",
       "dataset    sample model  lp_alpha lp_num_layers                          \n",
       "ogbn-arxiv equal  GCN    0.7      10                   0.701581  0.006488\n",
       "                  GCN+LP 0.7      10                   0.712912  0.006063\n",
       "                  LP     0.7      10                   0.669829  0.012142\n",
       "           frac   GCN    0.7      10                   0.482383  0.009206\n",
       "                  GCN+LP 0.7      10                   0.493914  0.012471\n",
       "                  LP     0.7      10                   0.468674  0.010687"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_model_acc(df):\n",
    "    #df_1 = df[df['dataset']==dataset]\n",
    "    # plot original acc\n",
    "    df_1 = df[['seed', 'model', 'lp_alpha', 'lp_num_layers', 'dataset', 'model_accuracy', 'sample']].drop_duplicates()\n",
    "    df_1=df_1.set_index(np.array(range(len(df_1))))\n",
    "    #df_1 = df_1[df_1['dataset']=='cora_ml']\n",
    "    #df_1 = df_1.drop(df_1[(df_1['model']=='LP')&((df_1['do_lp']==False))].index)\n",
    "    #df_1 = df_1.drop(df_1[(df_1['model']=='GCN')&((df_1['do_lp']==False))&((df_1['lp_alpha']!=0.3))].index)\n",
    "    res = df_1 .groupby(['dataset', 'sample','model', 'lp_alpha', 'lp_num_layers']).mean()[['model_accuracy']]\n",
    "    res['acc std'] = df_1 .groupby(['dataset', 'sample','model', 'lp_alpha', 'lp_num_layers']).std()[['model_accuracy']]\n",
    "    return res\n",
    "\n",
    "\n",
    "df_1 = get_model_acc(df)\n",
    "# df_2 = get_model_acc(df, 'pubmed')\n",
    "# df = df_1.append(df_2)\n",
    "# df = df.reset_index()\n",
    "# df['accuracy_1'] = (df['model_accuracy']*100).astype(str).str[:4] +' \\pm '+ (df['acc std']*100).astype(str).str[:4]\n",
    "df_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "5c3b6b6c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{l}\n",
      "\\toprule\n",
      "Empty DataFrame\n",
      "Columns: MultiIndex([], names=[None, 'model'])\n",
      "Index: Index([], dtype='object', name='dataset') \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{llll}\n",
      "\\toprule\n",
      "{} & \\multicolumn{3}{l}{accuracy\\_1} \\\\\n",
      "model &            GCN &         GCN+LP &             LP \\\\\n",
      "dataset  &                &                &                \\\\\n",
      "\\midrule\n",
      "Cora     &  81.4 \\textbackslash pm 0.35 &  83.0 \\textbackslash pm 1.80 &  12.2 \\textbackslash pm 1.79 \\\\\n",
      "citeseer &  67.6 \\textbackslash pm 2.13 &  67.3 \\textbackslash pm 1.96 &  13.6 \\textbackslash pm 2.29 \\\\\n",
      "cora\\_ml  &  83.8 \\textbackslash pm 2.48 &  84.4 \\textbackslash pm 3.65 &  12.3 \\textbackslash pm 2.37 \\\\\n",
      "pubmed   &  77.7 \\textbackslash pm 2.67 &  77.7 \\textbackslash pm 4.19 &  38.9 \\textbackslash pm 7.14 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df_1=df_1.reset_index()\n",
    "\n",
    "df_1['accuracy_1'] = (df_1['model_accuracy']*100).astype(str).str[:4] +' \\pm '+ (df_1['acc std']*100).astype(str).str[:4]\n",
    "df_1['accuracy_2'] = (df_1['model_accuracy']*100).astype(str).str[:4] +'\\n\\pm '+ (df_1['acc std']*100).astype(str).str[:4]\n",
    "\n",
    "print(df_1[df_1['dense_split']==True][['model', 'dataset', 'accuracy_1']].pivot(columns=['model'], index=['dataset']).to_latex())\n",
    "print(df_1[df_1['dense_split']==False][['model', 'dataset', 'accuracy_1']].pivot(columns=['model'], index=['dataset']).to_latex())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "b9c57414",
   "metadata": {},
   "outputs": [],
   "source": [
    "# boxplot\n",
    "\n",
    "def plot_all(df, dataset, max_degree, type='min', plot_class = None, filter_dict={}, main=False):\n",
    "    # clip at 100 for plot in main\n",
    "    if main:\n",
    "        clip=100\n",
    "    else:\n",
    "        clip=None\n",
    "    # compute min and max over classes\n",
    "    df = compute_min_max_cols(df, clip)\n",
    "    # filter for filter dict\n",
    "    filter_dict.update({'dataset': dataset})\n",
    "    for k, v in filter_dict.items():\n",
    "        df = df[df[k]==v]\n",
    "    # x-labels and data to plot per x-label\n",
    "    data = []\n",
    "    names = []\n",
    "    for deg in np.sort(df[df['degree']<=max_degree]['degree'].unique()):\n",
    "        names.append(deg)\n",
    "        if type == 'min':\n",
    "            data.append(df[df['degree']==deg]['min'])\n",
    "        elif type == 'max':\n",
    "            data.append(df[df['degree']==deg]['max'])\n",
    "        elif type == 'all':\n",
    "            assert False\n",
    "    names = [str(int(x)) for x in names]\n",
    "    # plot\n",
    "    if not main:\n",
    "        fig, axs = plt.subplots(figsize=(4*1.62, 4))\n",
    "    else:\n",
    "        fig, axs = plt.subplots(figsize=(3*1.62, 3))\n",
    "    axs.boxplot(data, notch=False, labels = names, flierprops={'markersize':5}, medianprops={'linewidth':3, \"solid_capstyle\": \"butt\"})\n",
    "    # custom plot options\n",
    "    if not main:\n",
    "        if type=='max':\n",
    "            plt.ylim(top=135)\n",
    "            plt.ylabel('$\\max_{c\\\\neq c*}$ $N_c(v)$', fontsize = 14)\n",
    "        elif type=='min':\n",
    "            if dataset == 'cora_ml': \n",
    "                plt.ylim(top=40)\n",
    "            elif dataset == 'citeseer': \n",
    "                plt.ylim(top=50)\n",
    "            else: \n",
    "                plt.ylim(top=135)\n",
    "            plt.ylabel('$\\min_{c\\\\neq c*}$ $N_c(v)$', fontsize = 14)\n",
    "    else:\n",
    "        plt.title(\"Structure perturbations until\\nGCN-prediction changes\", fontsize=14,fontweight=\"bold\")\n",
    "        plt.ylabel('$\\max_{c\\\\neq c*}$ $N_c(v)$', fontsize = 14)\n",
    "        plt.yticks([10*x for x in range(11)])\n",
    "    plt.xlabel('Node degree', fontsize = 14)\n",
    "    plt.xticks(fontsize=14)\n",
    "    plt.yticks(fontsize=14)\n",
    "    plt.grid(axis='y')\n",
    "    return fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "d89548c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_all(df, 'Cora', max_degree=8, type='max', plot_class=None, filter_dict={'dense_split': False, 'lp_alpha': 0.7, 'model': 'LP', 'projection': True})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "97423cdd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2c02f00ed05b4e6a9e0edbcb203c0d1b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5b13a7e0da30498b8b353831425f6368",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "574f03dace3e4524913312eba1f6ddda",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "26af297a036544f2aa46eee52753bce7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8474dc18452d4b6cb278968df35cb4ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6f060bf4cee84958b048bb0e73aa9c6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "86af4b8233a74949b77eca2f8b91bc8a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95fb7436744d4d86952ef084ab14c750",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "065d1d5766514bdab60c7ace65c44fbf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cac4f25247004a79a25f8806b0dc1f19",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "854e3a7f6976440a9a233a501daab2af",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "293ff6f83d3e49c591a7b1b4ed43c08e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "09209e64d39647f0b9a5015fd2f0b40a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc4c4850668f4d7981faa5fac64dfe00",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b6b7027ca1b446daaae3afddb63b8c43",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a8807dbd466640c799dd5f3da78cfde1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=48.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0006021ba94147ce86c6ff95b43cf1e3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb4b2a0264864bc09a50e26b1137cfd6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 466.56x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVsAAAD5CAYAAACJbJa8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAABCX0lEQVR4nO2deXxU1fn/308SlgQSBNlBRMQCiqwCKi6gFUSwYqviUhV/1mq1trVaAUVAoFItX7Xu1VZBq8WKBQWKohUKLoAiGBBQWZRVWQSEhCUJz++Pc4fcDJNkkszcmSTP+/W6r5l77rnnPHNn5rnnnuXziKpiGIZhxJeURBtgGIZRHTBnaxiGEQDmbA3DMALAnK1hGEYAmLM1DMMIAHO2hmEYAWDO1jACRET6iIh6W5841tPaV8/QeNWTzIjI197nn+TtJ/SamLMNEBGpISLDRWSliOSIyA8islZE3hCRnr58Y0I/igTZOc+rf14i6k80ib7+ZaEEWw8Ci7xte/CWBUMpN6+luM+/NnDDIpCWaAOqGQ8Bv/PerwH2A8cDPwGmAYvLU6iI1FTVQ7EwMJ4ku50ikgYUxLH8mvEqOxxV3QqcHlR9yYiqXppoG4qgqrYFtAHfAgqM9aUJ7k9xsrc/z8sTvg0FWvv27wamA7nAo0Af37E+vvJDaWN8aU2AZ4ANwCFgGzAjLH/41tqz4ci+l99v01AvzZ/vcuATr57B3vEewEzge1wLbDlwQynXzl/PncA/gX2e7aMB8eXNBB4G1nv1bvU+7zG+PJO8sr727F0PHAaWRXn9h/rK+tpLmxTB1pK+p2uAGd6xjcAtvjLreOetB3K86/QVMBaoWcbfit/WjsC/gR3etVkPTATq+vKEyp0H3OZ9vr3ed9bUl68n8I5X1kHvM8wCTivhexwTssuX5r8mfcLzAX2BT3GNk0+B0yPk8W+TovhehhZnY7w2a9kGS6jb5gIR+Rj4WFW/BRb68qwE2gItvP1F3mv4o+A44ADuESk/WgNE5FivzOO9pDW4J5xBvvpOxjmsvZ494P5M5eEfuD/jN4CKyJnAXKAmzlGuwTmA50Wkgar+XxRl/tEr8wegGe5Ptx14yms9zgO64ZzJKuBE4Gagh4icrqp5vrKaA3/37NgGfAg0pPTrHy0lfU9/Bb7DOduWwNMiskFV/wOkA5d4x7/0bGoL3Ocd+wPR/1YAEJEOwEdAXZwDXwO0w928zhCRs1X1sO+UM4FeOCdaFxgI/B9wjYik4BxrQ9x1+xx3LS/C3Qg/ifYCRcFsnONMA7oCU0SkLbAJ9/128PKtwv0mkqLb4CgS3dqrThuR78Rf4ForGZHyhZ3f2nfeKryWGpBKlC1bYJQv7Wpfvu6+9/O84/PC6h/qO7d1BJuGRsj3MpDis/M9L/1/QA0v/V4v7QegdjHXzl/PXNwfrybOwSjwjZfvOm8/DzjFSzse5+gUuMZLm+Qr72YvTbwtmus/1Jf+NcW3oEr6nl7y0uvh3YyA/3lpNfGednz1vOTl2VjG30roe5ns7ecAx3tpt/jyXRz2/RcAXby0f3tp33r7x/rOO85Xb9tQ2aX9B3xp/mvSJ8J/5XYv7Te+tPbFnRvl9zK0OBvjtdkAWYCo6hjgp8AbOMcC8CNca+XFMhY3WVV3e+WWpZ+xl/f6taq+4rNtSRnrj5bH1WsteXaG6j8HOOQN7Iz30jKBU6Io89+qmq+u/3e6l9ZKRLJ85acBK7zyv8Y5Oji6H3M/8Jxnn6r3r4whJX1P//LS9wBveWkdvdcC4Oci8qWIHPQ+x8+9Y83LaUsP7/UDVf3Ge/+K7/hpYfmXq+oy733oCaexZ/NOXCsZ4EsRWSEi/8I98m8pp33F8VKYDeC6wioV1o0QMKo6DZgmIoJ71H3Wex0kIila9DGuJL4NL9r3PhVAROpV1N7S6sC1ykoi3M4QW3CPp+FE+/lDiO+9+vbzcP174XwXtr+tDNc80ueHkq9BcZ8/HAnbHw6M8N5/45XTEtdlUNFGUrQ3lN2+96EuEL+d5wNXA71xXU8/xfXRdwR+W1rdIpLq3YBK/A2FblYU7YYJv15Jj7VsA0RE/iwiveBIK2oJsNo7nOP70+f6zqkTZfHbfO9P9F4jjcaG+vVai8gVvno6+/KE6g+vO9o6SuJj73ULcL6qnq6qpwMXA4+q6tIoyrhURNJEpAZuJgfABlXdS+GMjjTgd77yzwLux/Uhl0Zx1/+oz+9NNzomijIjcZlXRibQz0tb4b2GWuBfqmprXP/pZ2WwNRKha99bREJ99lf7jkfdz+o1Fs7EPaL/P+8aT/YOn1fCqbH4DfnJ9b2P9r+SEMzZBsu1wEIR2Ssi2SLyDYU/dv/j3Grf+89FZKGItCml7K9wswsAHhaRucDTEfI9iWspAbwqIl+JyFrcnMTw+k/z7Aw94i7CzQAA+KeIzAdGlmJXOCNxrc7TgK0islRENuBabn+KsoyeuFH0rynsNngoZBduRoEAH4rI5yKyCtgD/AfXb1caEa+/qu6n8NH5Lu8az6DsrfEQP/Ou/XoKByxDnyPbe/2RiKzHfbeRpnKV5bfyJ9z3V8fL+znu9wBuYHBWGWxPBd4FdnnXeDmur95veyTmUni93hORRbj/RXlZi/s9Abzoff7LKlBe3DBnGywjcf2123F39aY4J/kAcJcv30xcP+JO3J+wF5BRUsGqmg8MwTnNVKAB8LMI+Xbi/rR/xT3GtwaycI4oxETcH2kfcCpeX56qfg9chRvUq+PVc01Un7yw/veBs73PmI97/AT3R78vymLuxQ3iZOFmJYwDnvLKP4gbNHkY54xPAhrhRsvHU9hyLImSrv9QYIFnewvg10TuDomGm3EDaBnAZuDXqhpyeA/gWoq7cZ9zCt5nLIOtRVDVVcAZuDndB3HjBRtxMwz6l6E7BVyf8jPAOlwf8o9wswOewU0Xi4iqrgZ+iftujgV2AbeWod7w8nbiBs42AvVxn79pecuLJxL78QDDiD0i0hrXAgQ3J3dS4qwxjLJjLVvDMIwAMGdrGIYRANaNYBiGEQDWsjUMwwgAc7ZGlSOSRKRPhm9MDMof6iuvdUXLixXh+q1GcmHOthIjIjVF5DYRmS8i34vIIRHZLCLvi8h93vJVf/62IvKMiKwRkQPeOR+LyEhPXtDvqFRERvvO9Qsv/zrozxoDQtqum6I9oQRd3+2+8sor0GNUM2y5biVFRI7BzYXt7iUdwKlDpePmxfbGzadc4eW/CHiNwjmYm3GqXp29/E9QdHkmwJ0i8qSq7ojX54hEPHRvvRVOsSprFmVbAGAY1rKtxDxBoaN9EjhWVTuq6om45aM/w1saKSINcSurMnCTyM9X1Zaq2sHL+zsKV+H4ycQtICgTYY/Zl4jIB15L+isRubSYfJeLyCcicggn04eI9BCRmV4L/KCILBeRG8Lqaikis0Rkv4h8IyK/LMamo7oRRKSJ19Lf4D0VbBORGaH8wLle1nP93QbFdSOIyE9EZIGI7PM+72ci8itvaWu4HQ+KyBMistOr9y+hp4sSrmumiDzkPZkc9K7LuyLS4Oisco+IbBGRXSLyD3FLgkMH7xKRZd75eV79/xaRH5XwHc73rvFqERkUVtlgcYI5B7x8F/nOHerL9yMRmeLVd8j7PfxBnFxjKM+F3u9ll1ffehGZJiInlHRtKgVBy4zZVvENJ9yRhxP1+AxPwrCE/LdRKC13Ryl553n51uJaugeAVhSVp/t1KWUM9eU9gFtS+oO3nw+cGiHfQVxr+yucjuuZXprixGM+9+W901dXSGLxME4VKge38k3xSUT6zh3j7R9LoQSfevWux5P+w2kMh2z+wdtfiNPP9dvd2sv/c1/ad6GyvG1CBDsO4VZ9bfKl3VTCNa0JLPHl3YBbyVfgsyH0eXI9m9f58v/RV9ZM7xqtxAm35/vKrB3huzmEe2rK9V2PBl6+U33n7/W+632+c4d6+dribvTqvX7m2a44ZThw2rih73wDbtn1Tm//rET/7yr8v020AbaV40tzUnnq/6F66c/40hWY6KU/6UvrXkrZ87x8n+CWFyvwAuV3tuO9tJa+P+HkCPnKrHuLk/MLnf9bL097Cm9E83w2hTvbWOv6hvRoP/ZsE9zTRMhZNQizYx3uplkbd5NRYEoJ1/Q637kjfOknAZne+6991yekDvaxl7bQd84poWvq7f/YV/b5ET7j/3lpP/GlXeilvejt5wCtvLQJvnxDvbTnvf0vfPZe46UVAMfhntRC9vv1nTsBjRL9v6voZt0IlZNwWcEQ6yhU9Yomf2k8gmulXUuhhkFhoSK/ECf8Edoird1/FUBVNwEfeGkdI+Qrj+7tqb7zQ9qwqylZCCVEzHR9RaQxrvUPME1VD6jzEv/00mrg+sb9vKmqe1T1AIXLkEvSaA3Zmwf82WfvV+rUzvy8p6qbvev5RYSyWwFzxQUcPYwLbRMiklZuSXqyoe/yA1UNCSH9k6MJ2f8j4Afv+wwpsKXgxIU+x/2GM4Ft4kSK/oH77QU6bhAPbICscrIa9+iWhpPLE3U8BDwkR0da/dz3/mwi67weharmiMh44HGc2Es4LSn8E4HrMiiJkjRIY6l7m0it02hvZrt97yNpxVaEYssWpwg2HdctsRfXNZEGdPHy+XV6w8srSU+2tM8dyr8TF44nnP2qekBEuuNu7r1wTvZqXAu4Ge7mX2mxlm0lRFV/wKlAgRMenyAlR259lcLIEKNE5NzQARGpLSK3S/FaqH/Ftby6RbBjjKqKb+sT4fzLvXqa4xSnIDrlrWh0b/3lhOppR9EWb3FURNe3CKq6jUJ5y59611SAK720PCJr0ZaFkL01gN/77D1RROqWoZyuOEcLTumrB/BgBexa7r2e6X3H4JThwgnpDOfgwu+Evs9+wNOq+h9xUxXbA0+o6s9VtRtOkhFK1sitFJizrbzcjmuVAAwDdnojzN+EZ1Q3detqnPNoAMwTkY0ishLXankM9yc+CnXBEUdVwM7fi9OTXYULGngYJ+lXGtHo3s6lUPD6UXH6rJ8SXTjyiuj6RiI0a+M0XN/pOgqdzv+pk6esCFMofCJ5UNzMi9W4boKGZSjncwqvz1vidGgfr4BdE73y6gKrPJt+EyHfAzhN4VbAN973uQ7X0p3k5WmM0wve6V3v1RQ62Wi6hpIac7aVFHWhQnoDd+BGyRUXZVRwg0p/wPfYpW5uaBdcS3Ud7ofdHOcEx1AoCh6JVyj/j/0K3CKAWrgZDkNUtdSyNArdW69f9Ke4+F2HcANOIykarbi48sut61tMef/AzaL4ANfn2Ax3zW4F7inNnijsPYTT6f0z7vtrhvsO/0fhU0s05awG/h/uaaUmri80Uks02vKW454qvsJ9xzuBX/iy7PfyfYnrGpiC6744xcs/Dzf1EO/cF3Ch51vj9HnX4lre95fXxmTBhGiMmOPNrXzB2z1BVb9OnDVGvBGRH3nONLR/Hy5iNLgouF9EPrN6YQNkhmFUlEVe99XXuClcof79yeZoCzFnaxhGRZmO61vtgOvy+RQX0ufJEs6pdlg3gmEYRgDYAJlhGEYAmLM1DMMIAOuzjUDDhg21devWiTbDMIxKxpIlS3aoaqNIx8zZRqB169Z88sknpWc0DMPwEWlRUQjrRjAMwwiApHS2InKOiLwpLsRLEQFi77iIyBhPHHm/uPAlp4TlqSUij4vIDhHJ8cprGegHMQzD8EhKZ4tbZ70C+C3ecr8w7gbuxOkD9MBFJHjHr0YPPIqLVnAVbtlnFjBTRCKpGsWM/fv3c+WVV9KmTRuuvPJK9u+PZH714Pvvv6dVq1bUqFGDVq1a8f33FZUHqNx2bN68mYyMDESEjIwMNm/enBA7tm/fTqNGjUhNTaVRo0Zs3749IXZUu/9KogV1S9tw69KH+vYFt3b6Xl9aOm699c3efj3cWvlrfHmOw4mg9C+tzu7du2t5yM3N1Zo1a6qIaPPmzVVEtGbNmpqbm1uu8iozO3fuPCJinpKScuT9zp07q6UdmzZt8ou6H9k2bdoUqB3btm2LaMe2bdsCtaOq/leAT7QKiYefADQF5oQSVHU/MB8XSgWc4nuNsDwbcaIrZxInbrjhBvLy8ti1axebN29m165d5OXlccMNN5R+chWjS5cuADz//POMGTOG559/vkh60HY899xzjBkzhueeey4hdpx00kkAZGdno6pkZ2cXSQ+Kk092ej5r1qxBVVmzZk2R9KCojv+Vyjgboan3+l1Y+ne4UCChPAUcre7+ne/8IogLFPhLgCZNmjBv3rwyGzZ//nzq16/PU089xVdffcVJJ51E/fr1mT9/frnKq8yEHpEnTJhAjx49eOmll46kB3kttmzZAsBDDz1UxI4tW7YEakfoEXnmzJk88sgjR5zs/v37A7Vj586dAGzcuJGNGzcWSQ/Sjvnz59OgQQOWLi1UtGzQoEGg/5W+ffuWmmfu3Lml5oma4pq8ybJxdDfCmbhHn+PC8r0AvOW9vxq3RlvC8swFnimtzvJ2I1x++eUKaOfOnXX48OHauXNnBfTyyy8vV3mVmYYNGyqga9asUVXVNWvWKKANGzaslnakp6croB06dNDhw4drhw4dFND09PRA7UiW6zFkyBAVEd29e7eqqu7evVtFRIcMGRKoHX6cO6xwGcV2IyTcmZa2RXC2bTxn2yMs3ywKAwme5+VpFJbnc+D+0uosr7OdOnWqeiFptHnz5gqoiOjUqVPLVV5lZsSIERH7BkeMGBGoHXffffeRukPfDaDDhg0L1I5JkyZFvB6TJk0K1A7rsy2eeDvbythnux6n1n9BKEFEauNmHHzoJS3Bqfz787TEqRJ9SJxYuXIld955J0OGDKF27doMGTKEO++8k1WrVsWryqSlV69eZGVlkZaWhoiQlpZGZmYmvXr1Kv3kGHLmmWeSmZlJjRouEEWNGjXIzMzkjDPOKOXM2LJhwwZuvfVW0tPTAUhPT+fWW28t8igfBI0aNWLr1q1kZWUhImRlZbF161YaNYq46ClupKens3v3bq644gpq167NFVdcwe7du49cn6pIUjpbEakrIl1EpAvOxlbefivv7vEoMFxEfioiHXFhNfbhIgqgqnuAvwN/FpEfi0hXXITQbJzyflzo2rUr7733Hi+99BJr167lpZde4r///W/ggzHJgojQsWNHhg0bRseOHXFhuYInJSWFU045hWHDhnHKKackxI6uXbuycOFC9uzZg6qyZ88ePvroo8B/GwUFBfzyl7+kbdu2DBs2jLZt2/LLX/6SgoJoIgnFlvT0dKZMmcLatWuZMmVKlXa0kKQSiyLSh8JAb34mq+pQL5jeaOBmoD4uGN5tqrrCV0ZtXAiRq3FTw/4L3KpuVkKJnHbaaVqe5boFBQVceumlbNq0if79+/P222/TsmVLpk2bRmpqXKf3Jh3jxo3jhx9+ID09nQ8++IDevXuzf/9+6tWrx8iRIwO1Y+/evdSuXfuIHQcOHCArKytQO5LltzFz5kxGjx7NwoULqVGjBnl5efTq1YuxY8cyaNCgwOxIRkSEivpDEVmiqhHDJyWls0005XW24P5Us2fPZtmyZXTp0oUBAwZUO0cL8MYbb3D99dfTpk2bI85l7dq1vPjii1xyySXVzg5Ijt/GuHHj2LdvH2effTZLly6la9euLFiwgMzMzEBvPslIvJ1tZZz6ldSkpqYyaNCgat9KCBH68Sb6pp4MdiTDb6NTp05cf/31vPPOO/Tv359Ro0YdufkY8cWcrREXsrOzufnmmzn77LNZtmwZ48aNY8GCBSxfvjzQFmV2djY33XTTke6Miy++mP379wduR7KRDDef6kZSDpAZFefQoUOMGjWK888/n1GjRnHo0KFA6+/atSvvvvsu/fv3Z+TIkfTv35933nkn8AGhTp068dxzz/Hyyy+Tl5fHyy+/zLPPPsupp54aqB3guhFmzpzJuHHjmDlzZkIGpUI3wXHjxlGnTh3GjRvHLbfcwvLlywO3pbphzrYKcujQIVq1asXEiRPJzc1l4sSJtGrVKlCHO2DAAFq0aEGvXr0YMWIEvXr1omXLlgwYMCAwG8A5uJycHLZu3UpeXh5bt24lNzc3cEcXGiAbPXo0ubm5jB49mksvvTRwO5LlJlgdqbTOVkQyReRREfnGk1n8UER6+I6XKsNYVRk/fjw//PADO3bs4KOPPmLHjh388MMPjB8/PjAbUlNTmTZtGmPHjqVOnTqMHTs2IbMy/v3vf9OgQQNeeeUVBg4cyCuvvEKDBg2YNm1aoHbMnj2bzZs3s3DhQiZMmMDChQvZtGkTs2fPDtSOZLkJVkcqc5/t34BOwPXAJuDnwLsicrKqbqZQhnEo8AUwCifD2E5V9ybG5GD44IMP6Ny5MxkZGQBkZGTQuXNnPvwwbus5SiTR/YIiwsCBAxk8eDB5eXkJmWe7dOlS+vXrV2RxRf/+/Vm2bFmgA2ahm2BoVsTYsWOr7YyZwCluaVkyb7h5s/nAJWHpS4DxRCHDWNJW3uW6ycJ9992n6enpmpOTo6qqOTk5mp6ervfdd19gNuTn5+vFF1+s3bp10+HDh2u3bt304osv1vz8/MBsUFWdPn261qtXT7t06aLDhw/XLl26aFZWlk6fPj1QO2bMmKHdunXTQ4cOqarqoUOHtGvXrjpjxoxA7TCKh+qujRDRaMjErenuH5b+PjCPKPQTStoqu7M9ePCgNmnSRNPT0/X000/X9PR0bdKkiR48eDAwG5LFueTn5+ugQYO0TZs2evbZZ2ubNm100KBBgTv90M2na9euOnz4cO3atWtCbj5G8cTb2VbaRQ0i8iFORvFKnFbCVcBkYA1wA/ABcLyqbvCd8zzQQlX7RyjPL7HYfcqUKeWyq6CggMWLFx+RWOzZs2dCHtH279/PxIkTWbVqFR06dOCuu+4KdDnkiy++yP79++nUqdORa5GdnU1GRgbXXnttYHaAGzB8+eWXWbFiBR07duSaa66hZs2agdoAhb+NNWvW0LZt24T9NhJN4NKGUdK3b98K19u3b9+qt4JMRE4EngfOwTndT4EvgW7AL3DOtpX6lueKyAtAM1W9sKSyK7Jcd/DgwaxatYrmzZuzZcsWOnTowPTp0wP9U4VGvjdv3ky/fv2YM2cOLVq0CHSAKllWbiXDtYi2j7iy/hcrSixWbiWLHSWtIKu0sxFUda2qnouLV3acqvbERWcIqYLB0ULhjTladDxmzJw5kwULFpCVlUXv3r3Jyspi/vz5zJw5M15VRiRZRr4h8ZPnk+FahD9ORkpLBmdjxJdK62xDqGqOqm4VkfpAf+ANopNhjDmvvfYatWvXZtGiRUyYMIFFixaRnp7O1KlT41VlREoa+Q6KZJk8nwzXwjCgEk/9EpH+uJvFaqAtTuHrC+AFVVUReRS4V0RW47oXRuKTYYwns2bNYvny5Zx66qkJabF07dqVUaNGkZKSwsKFCzn99NN56623GDduXKA2jB49mvHjxzNo0CDy8vIYNWoUY8eODcwGvx29evU68p28/fbbgdthGAmfWVDeDbgCWAscxE3zegKo5zsuwBjv2AHgf0DHaMou72yE119/XdPS0orMAkhLS9PXX3+9XOWVl2RQwU+W0fdkmJkRDjEY9a5KJMv1iIUdVLFIDQCo6r9U9URVraWqzVT11+pEw0PHVVXHeMdqq+q56tO7jQepqalkZGTQtGlTatSoQdOmTUlPTw98xHnChAmkpKQwbNgw2rdvz7Bhw0hJSWHChAmB2ZAsK8jmzJlD8+bNueOOO8jIyOCOO+6gWbNmzJkzp/STDSOGVFpnm4xkZ2dzyy238Nhjj9GvXz8ee+wxfvWrXwXeT/n+++9Tu3Zt5syZQ8+ePZkzZ84R8exEoAkc/FmyZAl79+7lrbfeomfPnrz11lvs27ePTz/9NGE2GdUTc7YxJFlEPho3bsyePXt47733mDBhAu+99x579uwJNM5UsgivFBQUsHnzZhYsWMCECRNYsGABmzdvJj8/P1A7DMOcbQxJFpGPdu3akZ6eTuPGjTnuuONo3Lgx6enptG/fPjAbkmHKFbjujBYtWnDWWWcxYsQIzjrrLFq0aEFaWqUdGy43IlLqZsQPc7YxJDU1lalTpzJo0CAWL17MoEGDmDp1auD9lN26dSM1NZX69esfeU1JSaFr166B2ZAsU666d+9OZmYmo0aNok6dOowaNYq6devSrVu3QO1IBsIHbIpLM+JDpXS2IpIqIuNEZL2IHPBex4tImi9P4BKLBQUFXHbZZcyaNYuePXsya9YsLrvssoSIRKekpNCsWTOuuuoqmjVrRkpKsF91165dmTNnDnl5eQDk5eXx9ttvB96lMmDAAFq2bMnYsWPJyclh7NixHHfccSYpaAROpVyuKyL3AHfh5BWX46QWJwMPq+o4L88w3NzaoRRKLJ4FlCqxWN7luskSuTQZIsomSzTZkC2JDrTopyotTzU7jiqjamkjiMhMYKeqXu9Lmwwcq6qDvFDnW4AnVPWP3vF0YBtwl6r+taTyy+tskyVyabLoEhw6dIjx48cfcfgjR45MiABMslGVnIvZcVQZVU4b4X2gr4i0BxCRk4HzgP94x0/A6SIcmUypqvuB+cCZ8TKqU6dO/PWvf2XUqFHk5uYyatQonnnmmYTEu4LE6hIkU5eKYSQDlXVI9kGcpu1KESnAfY4/qupT3vGQAE246Mx3QItIBYZJLDJv3rwyG7VixQry8/PZu3cvGzZsYO/evRQUFLBixQrq1atX5vLKyxtvvEG/fv3YuXMnkydPpkOHDpx00km8+eabgdnx0UcfsXr1aq677jrWrVvHpZdeyuTJk3nooYc444wzArEhmSnP7ysemB1FiacdldXZDgGuA64GPge6AH8RkfWq+ndfvvAmnURIcxlVnwWeBdeN0KdPnzIbtWDBAm677bYj4buvuuoqFixYgKpSnvLKy65du7jiiitIS0ujS5cufPjhh+Tn5/Ovf/0rMDv+97//cfjwYSZNmkTz5s2ZP38+4AbKgrwWyUqyXAOzoyjxtKOydiP8GZioqlNUdbmqvgQ8DIzwjidEYjFZFjUsWbKEw4cP065dO/r06UO7du04fPgwS5YsCcyGvLw81q9fT2ZmJr179yYzM5P169cHHlLdMJKFyupsM3CC4X4KKPw8CZFYHDBgAM2bN6d9+/acc845tG/fnhYtWgQ+zeijjz6iR48ejB8/njp16jB+/Hh69OjBwoULA7Nh3bp1Ry0cSEtLY/369YHZYBjJRGV1tjOA4SIyUERai8ilwO+BaeBEaIBHvTw/FZGOwCQCklgMrcRJ1Iqc3r17k52dzXnnncfIkSM577zzyM7O5swz4zY2eBQpKSkcc8wxXHTRRSxevJiLLrqIevXqBT7fF9xg3cyZMxk3bhwzZ860QTojMRQnB5bMG25w7FHgG2A/sA54AKjtyxO4xGKyBDlMBlnBZJGbTJYov36oQpKCsaAq2UFVk1hU1b2q+jtVPV5V01W1jareo6oHfHlUA5ZYTJYlqjVr1mTDhg3cdddd1KlTh7vuuosNGzYEOsc1NTWV9PR0srKy2Lp1K1lZWdSuXTvwxQTJotFgGJXS2SYrybJEFZyz69mzJ+eee25CorguW7aMxo0bU7duXVq1akXdunVp0qQJn332WaB2LF26lB//+Me8/fbbjBs3jrfffpsLLrjAwuIYgVNZp34lJQMGDODZZ5+lV69eRZaoBj1AFh5RdvTo0Tz77LOBLpUtKChgy5Yt7Nixg4yMDHJzc2nYsGHg0oadOnXi+uuv55133qF///6MGjXqyGo6wwgSc7YxJBSdILQOf+zYsQlZh+9/dK5RowZjx46lV69ezJ49OzCNBr+0YejGkyhpQ1Vl9+7dfPDBB+zevTsploYa1Y9KqY0Qb8qrjZAsJFqjIdpZGEH89u6//35eeuklAJo3b86WLVsAuO666xg1alTc649EVdICMDuOKqNYbYQyNzNEpBbQHEgHtqvq9gpZVw5E5Gvg+AiH/qOqAz0hmtG45bf1gUXAbar6eXBWJo5EPzqrahHVr5DDT4TqV7J0ZxhGVM5WRDKBnwNXAT2BGnhLX0VkC/AW8KyqfhwvQ8PoAfj/sc2AJcC/vP27gTspKq/4joiUKq9YlQjdpRPRavB3qVx88cUJ61JJTU2lefPmnHrqqbRo0YLNmzfTvHnzahmpwUgspc5GEJE7gK+B/we8A1yC0yL4EXAGrgWZhnNmb4nISfEyNoSqblfVb0MbcBHwA/Ca16r9HfAnVX3dm+51PW5u7tXxti0ZyM7O5he/+AXt2rXj1VdfpV27dtx0002BB55MTU090kc8aNCghGjIdunShe3bt5OTk8OGDRvIyclh27ZtdO7cOXBbjOpNNLf3M4GS5qguBp4XkVuAG4Fzga9iZF+peM71RuAfqporIm2IIK8oIiF5xRK1bKsCp5xyCkOGDKFGjRp07tyZN998k7y8PF599dVEmxY4BQUF5ObmkpeXxwknnMBnn31GXl6erSIzAqdUZ6uql4fei0gnVc0uJt9B4KlIx+LMBTj92r95+2WWV4TYSCyC+3MvXryYr776ipNOOikhc1ynTZtGQUEBDRs2JCcnh3r16vHdd98xbdo0GjRoEKgtIRIloff000+TmZnJ73//e9avX8/AgQN5+OGHefrppxN2LaB6SAqWhepgR5lmI4jIQZy2wH2quiluVpUBEXkNOF5Ve3r7ZwIfAK1UdaMv3wtAM1W9sLQyyzsboaCggMGDB7Nq1aojI98dOnRg+vTpgTrc8847j6VLl1K7dm1q1arFwYMHOXDgAN26deO///1vYHaESORo889//nPeffddNm7ceCRU0XHHHccFF1xwZJZC0FSl0Xez46gyYhapoTNudP9LEXlQRI6pkGUVREQa4/qQn/MlJ0ReEVwMsgULFpCVlUXv3r3Jyspi/vz5zJw5M57VHkWjRo3YvXs3TZo04aqrrqJJkybs3r2bhg0bBmpHMnD55Zdz4MABevbsyYgRI+jZsyf79+/nsssuS7RpRjWjTEOyqroaGOy1Hv8ErBWRB4DHVTURQqVDgYPAFF+aX17xYygir/iHeBrz2muvUatWLQYMGMDChQsZMGAAzz33HFOnTg009peIICKsWrWKnTt3sm3bNkQkIYpbicQ/33fZsmVFlugOHjz4yPtkaFUZVZ9y/ftU9UNVPQc3Q+EG4CsRuS6mlpWCNzD2C2CKfzqXp7zzKAmQV8zPz2fHjh1MnDiRL7/8kokTJ7Jjx47A53SmpKSQlpZG/fr1SU1NpX79+qSlpVU7ZxtSW8rPz2fGjBkAzJgxg/z8/HAVOcOIOxX9980HbgM2AS9U3Jwy0Qc4iaJdCCEewkVueBL4BDcPt1+859gWFBRw+PBhRISWLVse6QMK2tm2adOGgoICmjZtylVXXUXTpk0pKCjghBNOCNSOZCEZpqAZRpm6EUTkVuBk39YIOIybh/tGrI0rCVWdi1tYEemY4rRsxwRoEl988cUR0eytW7dyzDHHsH37dr788ssgzaBGjRq0bt2ab7/9ln/+858cOHCA1q1bWxhxw0ggZW3ZjgbaA58B9wCnA/VU9SRV/Wmsjats1K9fH1UlIyODVq1akZGRgapSv379QO3o0qULO3fupE6dOrRq1Yo6deqwY8cOm8hvGAmkrANkTeJlSFWgd+/ezJ8/n2+//ZYmTZrw7bffoqqBhqPxk8jluoZhFCWa5bpRd/SJ47iKmVR5qVWrFm3atCkSnaBNmzbUrl07UDuWLVtGw4YNjwyIpaSk0KhRo8CFuw3DKCSaboSPROTvInJGcRlEpL6I/ApYiZv3Wi3p3r079erV48Ybb+TEE0/kxhtvJCsri27dugVqR0jpKjs7m/nz55Odnc2WLVtM6cowEkg0zrY98D0wS0S2e2IzL4jI0yIyRUSygW04VbDfqeoT8TQ4hIg0E5HJnk0HRGSliJzrOy4iMkZEtojIfhGZJyKnxNOmfv36sWXLFh555BFyc3N55JFH2Lp1K/369YtntUeRmppKs2bNaNGiBccccwwtWrSgWbNmpnRlGAmkVGerqrtV9Q84XYFfAauBY3B6BPnAZKCrqvZW1bfjaOsRvJVrH+BmIwwEOgC345x+iJDM4u04ScZtOGWyzHjZNWfOHJo3b84rr7zCwIEDeeWVV2jWrBlz5swp/eQYcvLJJ7Nu3Tr27NnD4cOH2bNnD+vWraNDhw6B2mEYRiFRN3VUdT8w1dsSzd3AVlX1L6RYH3oTLrPopV2Pc7hXEyflr6VLl9K/f38GDx58ZIXSokWLWLZsWWDhaACmTnVf0fHHH89xxx3Hxo0b+frrr5k6dSqXX355KWcbhhEPKvRcKSLHA1tUNS9G9kTLYOAtEXkV6Atswal+PenNsT2BMsosxkL1Ky0tjddff53zzz+ftLQ08vPzef3117nhhhsCVTWaP38+NWrU4ODBg+zZs4eDBw9So0YNFixYkDB1peqg6lQWzI6iVAc7KhSDTESWA2eGVmaJSAOgvap+GCP7iqv3gPf2EVx0hi7A48BwVX3Cp/x1vKpu8J33PNBCVfuXVH5FVL9CoWD80XWDDgVzzjnnsGDBArZv307Dhg3ZsWMHjRo14uyzz2b+/PmB2RGiKqk6mR1mRyllxEz1K5y8sCWwu4AgBshSgE9VdYSqLlXVF4DHcEuH/YRfOYmQFjNCoWDGjh1LnTp1GDt2bOCOFuDcc88lNTWVxo0bk5WVRePGjUlNTaVPnz6B2mEYRiEVdbYbROSc0I73CB/EmtCtuGlmflYBrbz3CZNZDJHIO/Vpp51Geno6NWvWJDc3l5o1a5Kenk737t0TZpNhVHcq6mx/DTwhIs+JyE0i8hhOJyHefAC0C0v7EfCN994vswgUkVmMWxdHqBth9OjR5ObmMnr0aC699NLAQ7Dk5+eTk5PDwYMHadKkCQcPHiQnJ8fm2RpGAqmQs/WiNXTHBYJsBXwJXBkDu0rjEeB0EblXRNqKyOXAb3AqXwmTWZw9ezYbN27kwgsvZPHixVx44YVs2LCB2bNnx6vKiDz++OOICJdddhm1atXisssuQ0R44olApkAbhhGBaEOZp6pqxOaZNxPhXxSGEY87qvqxiAwGHgDuAzZ4r/4YaA8B6TgHXB9YRJxlFj/++GO+/PJLVq5cSePGjXn//fdJS0vjk08+CXTq186dOzl8+DBvvvkmjRs35s033+Tw4cPs3LkzMBsMwyhKtFO/9nkrxZb4thWqmrDnUlWdBcwq4XjgMourV68mNzeXzp07M2DAAGbPns1nn33GqlWrgjIBgKysLMBpIrRs2ZIdO3YAkJkZt/UchmGUQrTO9kagK67L4CqgHnBQRFZQ6Hw/VdUlcbGykrBq1SpSU1OPhGMJhaJZvXp1oHbs3esa73l5eaxcuZK8vLwi6YZhBE9Ufbaq+oqq/kFVz1PV+rjBqRuA94C2uEf2RfEzs3LQoEEDRIR7772XOnXqcO+995KSkhJ4yOyQfm4osm6tWrWO2GcYRmIobwyyr4CZOBHxvUAtiuoSVEt+85vfAHDdddcxa9YsrrvOrSa+/fbbA7WjV69eAOTm5lKjRg1yc3MB6NmzZ6B2GIZRSJmcrYhkici1IvIGsB2YgJtu1Q8nVBMInpqXhm3f+o4HrvgFcMkll9C/f/8ierb9+/cPNLIuOD0GcMuOjz32WJo0cZrvixcvDtQOwzAKiXY2wvXA5cCPgc3A68ADqprIroMvcEEfQ/hnS4QUv4Z6+UbhFL/axXM2QmpqKm+88QazZ89m2bJldOnShQEDBgS+gmz37t1kZGTwzDPPsGLFCjp27Mg111zD7t27A7XDMIxCom3ZvgB0Bn4LdFDVuxPsaAHyVfVb37Ydjlb8UtUVwPVAJk7xK66EIrmOHDkyYZFcL774Yg4cOMCYMWPIyclhzJgxHDhwINDpZ4ZhFCUqIRoReQ8n9nIMkAd8TtFpYJ+p6qG4WXm0PWNwrdddwCHc4Nw9qrpORNoAa4Geqvqx75xZwA5Vvb608ssqRBOafVAa8V7Cmyx2hFOVhEbMDrOjlDKKFaKJqhtBVc/zCmoDnAZ0w00DuxRoAOSJyOeqGlT8l0W4LoLVOL2DkcCHXr9sSA8hXAPhO0roV66IxOLcuXOPSuvbt+9R6fGWkfPXd+jQIV5++WVefPFFrrvuOq655pojocwTIWdXHST0ykIQdvzkJz8pdbpfaTfozMxM3nzzTbMjBjZUSGIRQERa4zlgVb2nQoWV34a6wDrgT8BCnHZCK1Xd6MvzAtBMVS8srbzySiyG2VRl7tZVwYbqaEeMWmpWRhnOr3DLtiRU9Wuc+EzCIjio6j4R+Rw4CZjuJTcFNvqyBab4ZRiGEU5FVb+SAk/Rqz1OejEhil+GYRglUSnDrYrIRGAGToCmMU6Epg4wWVVVRB4F7hWR1TglspHEWfHLMAyjJCqlswVaAv8EGuIWVywETlfVkJ5t4IpfhhGiQYMG7Nq1q8Q8pQ0I1a9fn++//z6WZhkJpkzOVkR+qqr/LubYMFV9MDZmlYyqlqiZmwjFL8MIsWvXrpgM5hhVi7L22b4sIn8TkYxQgoi0FJF5wB0xtcwwDKMKUVZn2ws4HVgmIqeJyBBgObAft8IMABE513s9J2IphmEY1YwydSOoaraInIaLiPARLlLtXar6WFjWfiKSB/QHgo+dbRiGkWSUZ+pXZ+BcYA1uqWxPETkSAkBERgM1gP8CaSIyKhaGGoZhVGbKKrF4H66l+gbO6XbHCYkvF5GzAVT1ftx0q1HAV6o6NqYWR7brHk9m8QlfWkJkFg3DMCJR1pbtr4CLVfX3qnpIVb8AzgCmAO/68tVS1T/jWrhxRUROB24CssMOhWQWbwd64MTN3/G3wg3DMIKirM62k6rO8Seoar6qDse3YktVQyHFn664icUjIvWAl3Ex0nb50hMqs2gYhhFOmZytqu4o4ViRgTAROV5E4t2yfRaYqqrvhaWfgNNGOHJjUNX9uC6QM+Nsk2EYxlGUeQWZiKQBPYFWQE3/MVV90bc7E+fY8rzzGgDtVTUm+gQichMu2OS1EQ6XWWaxIhKLxVGd5Pwqgw0QnB3J8vuxMmJbRkXOL5PEooi0x2kSnAAILhRNGs6hHlTVLF/eT/36tt6j/ZJYaN6KSDvgfeBsVV3tpc0DVqjqr0XkTCogs2gSi1XPhiDtSAY5QCsj9mVUVGKxrH22j+IiM9QDcoEOOC3bZcDPwvJu8C9q8JbQ1iQ2nIHTRVghIvkiko+bjnar936nl69p2Hkms2gYRkIoazdCD+BcVc0RkcNAmqp+KiJ3A48DnXx5fw38R0QWAYuBU3G6t7FgOhDe9HwB+Ap4ADf1LCSz+DEUkVn8Q4xsMAzDiJqyOlvBtWjBqW21wEWv3YTrPz2Cqm4SkVDonFNxDjAmkRxUdTewu4hhIjnA997MA0xm0TCMZKKsznYFbjHDOlxrdZiIFODmua4Jz6yqecC/vC1oTGYxIExS0DBKp6zO9o84kW5wLcWZwFxgB3BFDO0qM6raJ2zfZBYDIlkkBc3pG8lMWYVo3va9Xwec7E3p2qUR/m0i0lxVt4hIC1XdXHFzDaN4ksXpG0YkyjPPtilu/mxjfLMZvGkRT3nva+LkGB8Ukb7AbIoOnhmGYVQryhqp4efA33ADZbtwEoshFCe9CG6K1x+AU7x0ay4YRsCMPrcWjKlX8TKMmFDWebZ/xA081VHVpqrazLc1D2VS1X3AYFxAxr8Ru/m1hmEYlZKyOtssYJKq5peWUVUP46ZazQOOE5HmJZ8RPSJym4hki8gP3vaRiAz0HTd5RcMwkoqyLtd9AvhCVR+PMv+HQF/cAoRc4CZVDZdCLDMicglOuPwr3A3jepykYncvmsQw3GyJobh5wKOAs4B20Uz9suW6wddhZVgZyV5GRZfrlnWA7PfAdBE5Hxd7LM9/MIJQuL+v9uLw/OVFVd8IS7pXRH4FnCEiy/HJKwKIyPU4Pdurgb/GwgbDMIyyUFZnezNwIW5ebVuOHiAr4mxV9Qzv7anlNbA0RCQVuByoC3xIMfKKIhKSVzRnaxhG4JTV2d4H3Kmqj8TDmLIgIqfigk7WxvUNX6qqyz3FLyiDvKJXnkksJrgOK8PKSPYygpRY3An0VNW1UeQ9HtjiLdmNOd5c3lbAMTjFsZuAPrhBvHLLK4L12SaiDivDykj2MoLus30BuIaw7oJiiKt4uKoeolCP4RMR6QHcgZueBq4rYaPvFJNXNIwEYPN9HWV1thnAL0SkPy7AYvgA2W98u3lhI/+7gCeACouHF0MKUAtYj8krGoaRZJTV2XYAlnrv24cdC29fbxCRc0KxyVRVvUf/CiMifwJm4VquoSCOfYCBXj2PUg3kFaMRXoGS1/ub8IpRXahoC7uireuyCtH0LUP2eIqHNwX+4b3uwbWyB/iEcqqFvKIJrxiVgfv/d5Ax8w5UrAyRSi/fV2YhmmiJs3j40FKOm7yikTCsj9KIRNycLSRcPDxu2OO7YVQ+KtrCrmjrOq7Otqpij++GYZSVMs2zrS6UNs82Geb8WRlWRjRlVJRYPIEl0/WoTPNsDcOoJEThGJJi4U11oVI6WxEZAfwUaAccBBYCI0KRdb08AozGLcENzUi4TVU/D95iIwhsYMpIZsqqZ5ss9MFFhTgTOA/IB971VqmFuBu4E7gd6IFT/XpHRDKDNdUwDKOStmxVtb9/X0Suxc237Q3M8Fq1v8NkFg3DSBIqpbONQCaulR6aj2UyiwGSLI/vNnneSGaqirP9C7AMJ7kIztFCGWQWyyqxmAxyb8lURixIls9SlcpIhjpiVU8ylBGYxGIyIiIPA1cCZ6nqOi/tTCogs2hTv8rGmD61GdOnYi3TMfMq3ipNluuRLGUkQx2xqicZyqjWU79E5BGco+0bcrQe33qvJrMYAPb4bpRGRef81q9fP0aWJI5K62xF5C84R9tHVVeHHTaZRSOhmHMpxOb7OiqlsxWRJ4FrgcHALhEJ9dHuU9V91Ulm0Ug+zLkYkaiUzha41Xv9b1j6/RQqfVV5mcVkmQVgGEbpVEpnq6qlPqOZzKJhGOFUpHunol07ldLZGoZhlJVEd++Ysy0HyfL4brMAjsYGpoxkxZytUWVIdMvFMErCnK0RE6xFaRglU2lXkInIOcBdQHegOXCDqk7yHS+3xKKtIIstydKiNDvMjnjbUdIKssoqsQhQF1gB/BbYH+G4SSwahpE0VNpuBFX9D/AfABGZ5D9WnSQW7fHdMCoHlbllWxIRJRaBkMRilUBVS91Ky2cRfg0jGCpty7YUqo3EYjLVk+w2gNkRjtlRlHjaUWkHyPyIyD7g16EBsuoisZhM9SS7DWaH2RGEHVV1gKwk/BKLfkxi0TCMhFBVna1fYhEoIrH4YaKMMgyj+lJp+2xFpC7Q1ttNAVqJSBfge1XdEG+JRZsFYBhGWai0zhY4DZjr27/f2yYDQ4mjxGI0/TrJ0g9lGCEiNRDC0+w3Gz8qrbNV1XlAsc1Lk1g0jKKYI00sldbZGoZROamuLWxztoZhBEqyONKgnb45W8MwqiVBO/2qOvXLMAwjqajyzlZEbhWR9SJyQESWiMjZibbJMIzqR5V2tiIyBPgL8ADQFbegYbaItEqoYYZhVDuqtLMFfg9MUtXnVHWVqt4ObAV+lWC7qjQiUmSLlFbRRSGVCbseBlQRIZpIiEhNIBe4SlVf86U/CXRU1XPD8vtVv7pPmTIl6rr69u0bVb65c+eWnqkCJIsdyYJdDyNo+vbtW6wQTVWejdAQSCWyzOKPwzOr6rPAs+BUv/r06RN1Rclyw0oWO5IFux5GMlHVuxEAwv9xEiHNMAwjrlRlZ7sDKMBkFg3DSAKqrLNV1UPAEnwyix4XYDKLhmEETFXuswV4GHhJRBbjIjfcggt7/kxCrTIMo9pRpZ2tqr4qIsfitGyb4UKfX6Sq3yTWMsMwqhtV2tkCqOpTwFOJtsMwjOpNlZ1nWxFEZDtQ0dZvQ9wgXaJJBjuSwQYwO8IxO4oSCzuOV9VGkQ6Ys40TIvJJcZObq5sdyWCD2WF2JNqOKjsbwTAMI5kwZ2sYhhEA5mzjx7OJNsAjGexIBhvA7AjH7ChKXO2wPlvDMIwAsJatYRhGAJizNQzDCABztjFERM4RkTdFZLOIqIgMTYANI0TkYxH5QUS2i8gMEemYADtuE5Fsz44fROQjERkYtB0R7LrH+26eCLjeMV69/u3bIG3w2dJMRCZ7v48DIrJSRM4t/cyY2vB1hOuhIjIrYDtSRWScL3TWehEZLyIxX/BV5VeQBUxd3JLgF70tEfTBrZj7GCcnORZ4V0ROVtXvA7RjEzAM+Ap3U78emC4i3VU1O0A7jiAipwM3AQmpH/gC9/2EKAjaABE5BqcT8j4wENgOtAG2BWxKD5zedIhmOOGofwVsxzDgNtzvcznQCZgMHATGxbIic7YxRFX/A/wHQEQmJciG/v59EbkW2AP0BmYEaMcbYUn3isivgDNIgLMTkXrAy8CNwKig6/fIV9WEtGZ93A1sVdXrfGnrgzZCVbf790XkRuAH4LXIZ8SNM4EZqhr6b3wtIm8CvWJdkXUjVH0ycd/zrkQZ4D2qXYlr+SdK3vJZYKqqvpeg+gHaeF1M60Vkioi0SYANg4FFIvKqiGwTkWUi8mtJYBA0r+4bgX+oam7A1b8P9BWR9p4tJwPn4TWaYom1bKs+fwGWAR8FXbGInOrVWxvYB1yqqssTYMdNQFvg2qDr9rEIGAqsxgnYjwQ+FJFTVHVngHa0AW4FHgH+BHQBHveOBdqP7eMC4ATgbwmo+0Fcg2SliBTgfOIfPQGrmGLOtgojIg8DZwFnqWrg/YO4PsouwDHAz4DJItJHVVcEZYCItMOFsj/bE5RPCKo6278vIguBdbi+wocDNCUF+ERVR3j7S0XkJFy/ZaKc7U3Ax6q6LAF1DwGuA64GPsf9Xv8iIutV9e+xrMicbRVFRB4BrgT6quq6RNjgObc13u4nItIDuAP3yBgUZ+DUnFb4npRTgXNE5BagjqoeDNAeAFR1n4h8DpwUcNVbgZVhaauA3wZsBwAi0hi4BOfsE8GfgYmqGgqnvVxEjgdGAOZsjZIRkb/gHG0fVV2daHt8pAC1Aq5zOvBJWNoLuFkSDwAJae2KSG2gPRB0HPUPgHZhaT+i4pKi5WUobuR/Sin54kUGR88KKSAO41nmbGOIiNTF9Q2C+7JaiUgX4HtV3RCQDU/i+iYHA7tEJBTwcp+q7gvCBs+OPwGzgI24PrGrcdOeAp1rq6q7gd1htuXgvpMguzMm4maDbMD12d4H1MFNMwqSR3B9xfcCrwJdgd8A9wRsR2hg7BfAFFXdG3T9HjOA4SKyHteN0BX4PfGYuqmqtsVowzkTjbBNCtCGSPUrMCbgazEJ11o6iJvD+S7QP9HfkWfbPOCJgOucAmzBtaQ3A68DJyfo8w8EPgMOAF/inK0kwI6+3m+zZwJ/C5nAo95vdT+uH/0BoHas6zIhGsMwjACwebaGYRgBYM7WMAwjAMzZGoZhBIA5W8MwjAAwZ2sYhhEA5mwNwzACwJytUaURkZnxkLuMV7lG1cWcrZFQRGSSp9A/Miy9j5feMFG2GUYsMWdrJAMHgLtFpFGiDUkmRCRFRFJLz2lUBszZGsnAXOBrnF5AsXgx3hZ5saK+E5FHRKSm73iG11Le5x0/ar2/iNQUkQdFZJOI5IiL19Y/PF/YOTEpV0QGisgXnv3zReRKr/Xe2js+1KvjIhFZgVva2yHKsk8WkVkistcTBf+nTxfDSALM2RrJwGFgOHCLiJwYKYOItABmA0txYiE3AlcBE3zZJuKEqH8GnO/lOyesqBeAc3HCOKfihGBmiEjnEuyrcLki0gr4N06cpzPwGPBQhLpq44TFbwZOxq3ZL63sZsB8XPy7nsCPcVEx3hQR+48nC4kSgLDNNtUjgjUzvfdzcQpQUCjq09Db/yNOGzfFd+5QnNBNBs65HASu8R2vi1P8muTtn4hz7K3CbJgOPFWMfTEpF3dTWIVP8AWntKVAa9/nUaC7L080ZY8F/ht2vD4JFnmxrehmEotGMnE3sNCTIwynA/CRqh72pb0P1MTJWor3/kj4H3UC3f4wPN28fCvDQm7VAoqLTXZijMptj4tG4Fd+WhShvnxcGKOylN0dJ4YeSULzRGBxpA9mBIs5WyNpUNWPReR1XFyo8DDSgmupRTyV6LrEUry8PYC8sGP7izknmkCI0ZRbkv1+DmrREEbRlJ2C6564K0J530VRpxEA5myNZOMeXNiWC8PSVwJXiEiKr3V7Fm4QaS3O4eQBp+M0SRGROkBH7zi4/l4BmqpqtBES1sSo3FW48C9+ekZRfzRlfwpcAXyjquEO2UgSrPPcSCpUdQ0u7Hh4TKyngObAUyLSQUQG4qLDPqGqueqiUPwdeFBELhCRU4DncfHGQmV/CbwMTBKRy0SkjYicJiJ3ichPi7EnVuU+A5woIhNFpJ2XfnOoiBKuRzRlPwnUA14VkV5enh+LyLMikln81TaCxJytkYyMxfVdHkFVNwMDcDMBluEc3j8pGs7lLtwg2zTvdQVulN7PDbjR/YdwYcVn4mYWlBSDq8Llquo3uNkMP8FFSbgDuN8790AJdUdT9hagN24g7S1ceJcncQN7gQezNCJjkRoMI0GIyG9xN5b6YQN/RhXE+mwNIyBE5DbgY2A7rg/4Ptz0MXO01QBztoYRHG1x3R7HAptw/bhjE2qRERjWjWAYhhEANkBmGIYRAOZsDcMwAsCcrWEYRgCYszUMwwgAc7aGYRgBYM7WMAwjAP4/HOcvjaXx1EcAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 349.92x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# appendix plots\n",
    "datasets = ['cora_ml', 'citeseer']\n",
    "models = ['GCN', 'LP']\n",
    "types = ['min', 'max']\n",
    "for dataset in datasets:\n",
    "    for model in models:\n",
    "        for plot_type in types:\n",
    "            filter_dict = {'model': model, 'projection': True}\n",
    "            fig = plot_all(dataset, max_degree=8, type=plot_type, plot_class=None, filter_dict=filter_dict)\n",
    "            fig.savefig(f'images_overrobust/{dataset}/{model}_{plot_type}_boxplot.pdf', bbox_inches='tight', format='pdf') \n",
    "\n",
    "\n",
    "# main plot\n",
    "filter_dict = {'model': 'GCN', 'projection': True}\n",
    "fig = plot_all('cora_ml', max_degree=8, type='max', plot_class=None, filter_dict=filter_dict, main=True)\n",
    "fig.savefig('images_overrobust/cora_ml/boxplot_main.pdf', bbox_inches='tight', format='pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3cbd74fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1f0a77174b4a4c20a0ad66d697319d4d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "050c8f5519d84af8832756fd601feb70",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=63.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 349.92x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# main plot\n",
    "filter_dict = {'model': 'GCN', 'projection': True}\n",
    "fig = plot_all('cora_ml', max_degree=8, type='max', plot_class=None, filter_dict=filter_dict, main=True)\n",
    "fig.savefig('images_overrobust/cora_ml/boxplot_main.pdf', bbox_inches='tight', format='pdf') "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.8.12 ('robust-gnn')",
   "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": "3.8.12"
  },
  "vscode": {
   "interpreter": {
    "hash": "0d6f2ee3047258fa0cf687e68f0d64b0f2d521f396e0ac9911dbad55d897a773"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}