{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9c945170",
   "metadata": {},
   "source": [
    "# S3A. Score Consistency\n",
    "\n",
    "Here, we briefly demonstrate the similarities and discrepancies in the pipeline outputs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c122940a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "../../data/liana-outputs/ already exists.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import functools\n",
    "\n",
    "import pandas as pd\n",
    "import scipy\n",
    "\n",
    "import scanpy as sc\n",
    "\n",
    "import cell2cell as c2c\n",
    "from liana.method.sc._rank_aggregate import AggregateClass, _rank_aggregate_meta\n",
    "from liana.method.sc import cellphonedb, natmi, singlecellsignalr\n",
    "\n",
    "data_path = '../../data/'\n",
    "output_folder = os.path.join(data_path, 'liana-outputs/')\n",
    "c2c.io.directories.create_directory(output_folder)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "517e0ca9",
   "metadata": {},
   "source": [
    "## S3-A1. Comparison with R output:\n",
    "\n",
    "There are minor differences with the LIANA implementation in R that lead to outputs not being identical\n",
    "\n",
    "- SingleCellSignalR Magnitude (lrscore): precision - slightly different after 3rd decimal place\n",
    "- LogFC Specificity (lr_logfc): similar relative differences but different exact values\n",
    "- CellPhoneDB Specificity (cellphone_pvals): similar relative differences but different exact values\n",
    "- CellChat: not run by default in R\n",
    "\n",
    "Let's check the consistency in the magnitude aggregate rank score when running the different methods that report magnitude (excluding CellChat, which is not present by default in R). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e7d48386",
   "metadata": {},
   "outputs": [],
   "source": [
    "adata = sc.read_h5ad(os.path.join(data_path, 'processed.h5ad'))\n",
    "sadata = adata[adata.obs['sample']=='C100']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0d512811",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using `.X`!\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "5580 features of mat are empty, they will be removed.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "The following cell identities were excluded: Plasma\n",
      "Using resource `consensus`.\n",
      "0.33 of entities in the resource are missing from the data.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generating ligand-receptor stats for 2548 samples and 19218 features\n",
      "Running CellPhoneDB\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████| 1000/1000 [00:04<00:00, 249.89it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running NATMI\n",
      "Running SingleCellSignalR\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_rank_aggregate.py:144: ImplicitModificationWarning: Trying to modify attribute `._uns` of view, initializing view as actual.\n"
     ]
    }
   ],
   "source": [
    "# make rank_aggregate function that only runs on methods of choice and only for Magnitude\n",
    "rank_aggregate_partial = AggregateClass(_rank_aggregate_meta, methods=[cellphonedb, natmi, singlecellsignalr])\n",
    "rank_aggregate_partial(adata = sadata, \n",
    "                       groupby='celltype', \n",
    "                       use_raw = False, # run on log- and library-normalized counts\n",
    "                       verbose = True, \n",
    "                       inplace = True\n",
    "                      )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "07975809",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>ligand_complex</th>\n",
       "      <th>receptor_complex</th>\n",
       "      <th>magnitude_rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>ACTR2</td>\n",
       "      <td>ADRB2</td>\n",
       "      <td>0.405892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>ADAM17</td>\n",
       "      <td>ITGB1</td>\n",
       "      <td>0.081878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>ADAM17</td>\n",
       "      <td>RHBDF2</td>\n",
       "      <td>0.334024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>ADAM28</td>\n",
       "      <td>ITGA4</td>\n",
       "      <td>0.403503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>APOC2</td>\n",
       "      <td>LRP1</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   source target ligand_complex receptor_complex  magnitude_rank\n",
       "67      B      B          ACTR2            ADRB2        0.405892\n",
       "17      B      B         ADAM17            ITGB1        0.081878\n",
       "54      B      B         ADAM17           RHBDF2        0.334024\n",
       "23      B      B         ADAM28            ITGA4        0.403503\n",
       "8       B      B          APOC2             LRP1        1.000000"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rel_cols = ['source', 'target', 'ligand_complex', 'receptor_complex', 'magnitude_rank']\n",
    "liana_aggregate_partial = sadata.uns['liana_res'].loc[:,rel_cols]\n",
    "liana_aggregate_partial.sort_values(by = ['source', 'target', 'ligand_complex', 'receptor_complex'], inplace = True)\n",
    "liana_aggregate_partial.to_csv(os.path.join(output_folder, 'magnitude_ranks_python.csv'))\n",
    "liana_aggregate_partial.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c41528f",
   "metadata": {},
   "source": [
    "Note, to run the correlation, make sure to have run the [companion Python tutorial](../ccc_R/S3_Score_Consistency.ipynb) up to the point where you save the csv named \"magnitude_ranks_R.csv\". "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8ec472a2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The spearman correlation bewteen R and python aggregate magnitude scores is: 0.98\n"
     ]
    }
   ],
   "source": [
    "# read and format R aggregate rank\n",
    "lap_R = pd.read_csv(os.path.join(output_folder, 'magnitude_ranks_R.csv'), index_col = 0)\n",
    "lap_R.columns = ['source', 'target', 'ligand_complex', 'receptor_complex', 'aggregate_rank']\n",
    "\n",
    "# merge the two scores\n",
    "la = pd.merge(liana_aggregate_partial, lap_R, on = ['source', 'target', 'ligand_complex', 'receptor_complex'], \n",
    "                                                how = 'inner')\n",
    "sr = scipy.stats.spearmanr(la.magnitude_rank, la.aggregate_rank).statistic\n",
    "print('The spearman correlation bewteen R and python aggregate magnitude scores is: {:.2f}'.format(sr))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9c246be",
   "metadata": {},
   "source": [
    "## S3-A2. Sensitivity of Consensus to Scoring Methods\n",
    "\n",
    "We can check the extent to which the consensus score is influenced by individual scoring method types. To do so, we can re-run the scoring pipeline on one sample (sample C100) as in [Tutorial 02](./02-Infer-Communication-Scores.ipynb), but see how the consensus score changes as we exclude each of one scoring method type. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a908323a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from functools import reduce\n",
    "\n",
    "import matplotlib\n",
    "\n",
    "import liana\n",
    "from liana.method import cellphonedb, connectome, logfc, natmi, singlecellsignalr, cellchat\n",
    "from liana.method.sc._rank_aggregate import AggregateClass, _rank_aggregate_meta as aggregate_meta"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "613395cf",
   "metadata": {},
   "source": [
    "First, let's load the sample data from Tutorial 02:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "729346c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "adata = sc.read_h5ad(os.path.join(data_path, 'processed.h5ad'))\n",
    "sample_name = 'C100'\n",
    "sadata = adata[adata.obs['sample']==sample_name]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d03294a9",
   "metadata": {},
   "source": [
    "Now, we can run the rank_aggregate function as in Tutorial 02. However, in this case, we must first specify the subset of methods we want to calculate a consensus score on."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d7b492ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "method_names = ['cellphonedb', 'connectome', 'logfc', 'natmi', 'singlecellsignalr', 'cellchat']\n",
    "all_methods = [cellphonedb, connectome, logfc, natmi, singlecellsignalr, cellchat]\n",
    "all_methods = dict(zip(method_names, all_methods))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7d336912",
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_consensus(method_subset, n_perms = 1000):\n",
    "    # re-generate the rank aggregate instance for only a subset of methods\n",
    "    rank_aggregate_subset = AggregateClass(aggregate_meta, methods=method_subset)\n",
    "    \n",
    "    # get the consensus score as in tutorial 02\n",
    "    liana_res = rank_aggregate_subset(sadata.copy(),\n",
    "                                               groupby='celltype',\n",
    "                                               resource_name = 'consensus',\n",
    "                                               expr_prop=0.1, # must be expressed in expr_prop fraction of cells\n",
    "                                               min_cells = 5,\n",
    "                                               n_perms = n_perms,\n",
    "                                               use_raw = False, # run on log- and library-normalized counts\n",
    "                                               verbose = False,\n",
    "                                               inplace = False\n",
    "                              )\n",
    "\n",
    "    return liana_res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "699ce47a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:246: ImplicitModificationWarning: Setting element `.layers['scaled']` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:256: ImplicitModificationWarning: Setting element `.layers['normcounts']` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:246: ImplicitModificationWarning: Setting element `.layers['scaled']` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:246: ImplicitModificationWarning: Setting element `.layers['scaled']` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:246: ImplicitModificationWarning: Setting element `.layers['scaled']` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:246: ImplicitModificationWarning: Setting element `.layers['scaled']` of view, initializing view as actual.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "../../data/liana-outputs//LOO_consensus_scores.pkl  was correctly saved.\n"
     ]
    }
   ],
   "source": [
    "subset_res = dict()\n",
    "for method_name in method_names: \n",
    "    #leave one method out\n",
    "    method_subset = [method_v for method_k, method_v in all_methods.items() if method_k != method_name]\n",
    "    \n",
    "    liana_res = calculate_consensus(method_subset)\n",
    "    liana_res.rename(columns = {'specificity_rank': '_'.join([method_name, 'specificity']), \n",
    "                               'magnitude_rank': '_'.join([method_name, 'magnitude'])}, inplace = True)\n",
    "    subset_res[method_name] = liana_res\n",
    "    \n",
    "c2c.io.export_variable_with_pickle(subset_res, output_folder + '/LOO_consensus_scores.pkl')\n",
    "\n",
    "subset_res = c2c.io.read_data.load_variable_with_pickle(output_folder + '/LOO_consensus_scores.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c924efc3",
   "metadata": {},
   "source": [
    "Next, let's incorporate the scores from running all methods together. This was done in Tutorial 02, and we can load the results here:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e39becf0",
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_name = 'C100'\n",
    "liana_res = pd.read_csv(os.path.join(output_folder, sample_name + '_aggregate_scores.csv'), index_col = 0)\n",
    "liana_res.rename(columns = {'specificity_rank': 'all_specificity', \n",
    "                           'magnitude_rank': 'all_magnitude'}, inplace = True)\n",
    "subset_res['all'] = liana_res"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96410858",
   "metadata": {},
   "source": [
    "Let's merge and format these scores:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "7cad0a49",
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge scores across each iteration\n",
    "merge_cols = ['source', 'target', 'ligand_complex', 'receptor_complex']\n",
    "subset_res = {k: v[['source', 'target', 'ligand_complex', 'receptor_complex', k + '_specificity', \n",
    "                     k + '_magnitude']] for k,v in subset_res.items()}\n",
    "subset_res_df = functools.reduce(lambda x, y: pd.merge(x, y, how = 'inner', \n",
    "                                             on = merge_cols), \n",
    "                       list(subset_res.values()))\n",
    "\n",
    "# get specificify consensus scores\n",
    "specificity_df = subset_res_df[[col_name for col_name in subset_res_df.columns if col_name.endswith('_specificity')]]\n",
    "specificity_df.columns = [''.join(col_name.split('_specificity')[:-1]) for col_name in specificity_df.columns]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28161fc0",
   "metadata": {},
   "source": [
    "Remember that the Connectome and NATMI magnitude score are the same calculation. So, in this case, we should  be leaving both out. Let's re-run our scoring for this one exception:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2d9107e2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/anndata/_core/anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/pandas/core/indexing.py:1819: ImplicitModificationWarning: Trying to modify attribute `.obs` of view, initializing view as actual.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/_pipe_utils/_pre.py:142: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "/home/hratch/miniconda3/envs/ccc_protocols/lib/python3.10/site-packages/liana/method/sc/_liana_pipe.py:256: ImplicitModificationWarning: Setting element `.layers['normcounts']` of view, initializing view as actual.\n"
     ]
    }
   ],
   "source": [
    "# re-run scoring to leave out both Connectome and NATMI\n",
    "method_subset = [method_v for method_k, method_v in all_methods.items() if method_k != 'connectome' and method_k != 'natmi']\n",
    "liana_res = calculate_consensus(method_subset, \n",
    "                                n_perms = 1 # speedup, permutations not necessary for magnitude calculations\n",
    "                               )\n",
    "\n",
    "# Format\n",
    "magnitude_df = subset_res_df[merge_cols + [col_name for col_name in subset_res_df.columns if col_name.endswith('_magnitude')]]\n",
    "liana_res.rename(columns = {'magnitude_rank': 'expr_prod_magnitude'}, inplace = True)\n",
    "magnitude_df = pd.merge(right = magnitude_df, left = liana_res, how = 'inner', on = merge_cols)\n",
    "magnitude_df = magnitude_df[[col_name for col_name in magnitude_df.columns if col_name.endswith('_magnitude')]]\n",
    "magnitude_df.columns = [''.join(col_name.split('_magnitude')[:-1]) for col_name in magnitude_df.columns]\n",
    "magnitude_df.drop(columns = ['natmi', 'connectome'], inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a77aa338",
   "metadata": {},
   "source": [
    "Now that we have consensus scores leaving each of one scoring method out, we can assess the sensitivity of the conesnsus score to each scoring method type. Since these scores are ranks, we will use the Spearman correlation to assess the effect. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6b293bfb",
   "metadata": {},
   "outputs": [],
   "source": [
    "specificity_corr = specificity_df.corr(method = 'spearman')\n",
    "magnitude_corr = magnitude_df.corr(method = 'spearman')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "fb4d1947",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Interaction space detected as a distance matrix\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "''"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = c2c.plotting.clustermap_cci(magnitude_corr,\n",
    "                                 method='ward',\n",
    "                                 optimal_leaf=True,\n",
    "                                 metadata=None,\n",
    "                                 title='',\n",
    "                                 cbar_title='Spearman', \n",
    "                                 cmap='Blues_r',\n",
    "                                 vmax=1.,\n",
    "#                                  vmin=0.,\n",
    "                                 annot=True, \n",
    "                                 dendrogram_ratio=0.15,\n",
    "                                figsize=(4,5))\n",
    "\n",
    "font = matplotlib.font_manager.FontProperties(weight='bold', size=7)\n",
    "for ax in [cm.ax_heatmap, cm.ax_cbar]:\n",
    "    for tick in ax.get_xticklabels():\n",
    "        tick.set_fontproperties(font)\n",
    "    for tick in ax.get_yticklabels():\n",
    "        tick.set_fontproperties(font)\n",
    "\n",
    "    text = ax.yaxis.label\n",
    "    text.set_font_properties(font)\n",
    "cm.ax_col_dendrogram.set_title('Magnitude Consensus')\n",
    ";"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a850ca0",
   "metadata": {},
   "source": [
    "We are particularly interested in differences with all scores (top row). We can see that, for the magnitude, CellChat has the largest effect on the consensus score. This is line with results from Tutorial 02 indicated that the CellChat score was the most dissimilar (see the `liana_corr`variable from that notebook). Let's see what this looks like for the specificity consensus scores as well:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "41814537",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Interaction space detected as a distance matrix\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "''"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAHQCAYAAABKlaYxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAACdvUlEQVR4nOzddVRUaR/A8e/Q3d0qEnZ3YK+ta+ta2IWru/qKHavommusrrH2unZgJzZ2CzaiWCDdMff9Ax0dQQUXHdHnc86cI3eevOD85ol7r0ySJAlBEARBeIuaqhsgCIIgfH1EcBAEQRCyEMFBEARByEIEB0EQBCELERwEQRCELERwEARBELIQwUEQBEHIQgQHQRAEIQsRHARBEIQsRHD4jslkMsaPH6907Ny5c1SpUgV9fX1kMhmXL19m/PjxyGSyXJfv5eWFl5dX3jRWEIQvSgSHz+jatWu0bt0aZ2dndHR0sLe3p169esybN0/VTctWWloabdq0ITIyktmzZ7N69WqcnZ3zrPwnT54wfvx4Ll++nGdlvu3y5cv89NNPODo6oq2tjZmZGXXr1mX58uVkZGR8ljoF4VslE/dW+jxOnTpFrVq1cHJyomvXrtjY2PDo0SMCAwO5d+8ed+/eVXUTSU5ORkNDAw0NDQCCg4Px9PRkyZIl9OzZU5EuPT2d9PR0dHR0clV+amoqAFpaWgCcP3+e8uXLs3z5crp165Y3nXhl6dKl9O3bF2trazp37kzhwoWJi4vj0KFD7Nq1i99++42RI0fmaZ2C8C3TUHUDvlWTJ0/G2NiYc+fOYWJiovTeixcvVNOod7z7Yf+6Xe+29+0Akhuvg8LnFhgYSN++falcuTK7d+/G0NBQ8d7PP//M+fPnuX79+hdpiyB8MyThs3B3d5e8vLxylBaQBgwYIK1Zs0Zyc3OTtLW1pTJlykhHjx7Nkvbx48dS9+7dJSsrK0lLS0sqUqSItGzZsizpkpKSpHHjxkmFCxeWtLW1JRsbG6lly5bS3bt3leodN26cJEmS1LVrVwlQetWsWVOSJEkaN26clN2fyurVq6Xy5ctLurq6komJiVS9enVp3759ivdr1qypKOPIkSNZygek5cuXS2PHjpU0NDSkFy9eZKmjV69ekrGxsZSUlPTe8/fDDz9IGhoa0sOHD9+b5m3x8fHS0KFDJQcHB0lLS0tyc3OTpk+fLsnlcqV0r38vW7dulYoWLao433v27FFKFxsbKw0ePFhydnaWtLS0JEtLS6lu3brShQsXlNIFBgZKDRo0kIyMjCRdXV2pRo0a0okTJ5TSvD7Xd+7ckbp27SoZGxtLRkZGUrdu3aSEhASltPv375eqVq0qGRsbS/r6+pKbm5vk6+ureH/58uUSID148EAp3+vfxZEjRxTHbt++Lf3444+StbW1pK2tLdnb20vt2rWToqOjc3ROhW+PGDl8Js7Ozpw+fZrr169TrFixj6Y/evQo69evx8fHB21tbf78809++OEHzp49q8j//PlzKlWqhEwmY+DAgVhaWrJnzx569OhBbGwsP//8MwAZGRk0adKEQ4cO0b59ewYPHkxcXBwHDhzg+vXrFCpUKEv9ffr0wd7enilTpuDj40P58uWxtrZ+b3snTJjA+PHjqVKlChMnTkRLS4szZ85w+PBh6tevnyW9p6cnEydOZOzYsfTu3Zvq1asDUKVKFapVq8bEiRNZv349AwcOVORJTU1l06ZNtGrV6r1TWomJiRw6dIgaNWrg5OT00fMsSRLNmjXjyJEj9OjRg1KlSrFv3z6GDRtGWFgYs2fPVkp/4sQJtmzZQv/+/TE0NGTu3Lm0atWK0NBQzM3NAejbty+bNm1i4MCBFClShJcvX3LixAmCgoIoU6YMAIcPH6Zhw4aULVuWcePGoaamxvLly6lduzbHjx+nQoUKSvW2bduWAgUK4Ofnx8WLF1m6dClWVlZMmzYNgBs3btCkSRNKlCjBxIkT0dbW5u7du5w8efKj5+BdqampNGjQgJSUFAYNGoSNjQ1hYWHs3LmT6OhojI2Nc12m8A1QdXT6Vu3fv19SV1eX1NXVpcqVK0vDhw+X9u3bJ6WmpmZJy6tv0efPn1cce/jwoaSjoyO1bNlScaxHjx6Sra2tFBERoZS/ffv2krGxsZSYmChJkiT9/fffEiDNmjUrS11vfzvmrZGDJL35Rrlx40alPO+OHO7cuSOpqalJLVu2lDIyMt5b/tsjB0mSpHPnzilGC++qXLmyVLFiRaVjW7ZsyfIN911XrlyRAGnw4MHvTfO2bdu2SYD022+/KR1v3bq1JJPJsoystLS0lI69rm/evHmKY8bGxtKAAQPeW6dcLpcKFy4sNWjQQOn8JCYmSgUKFJDq1aunOPb6XHt7eyuV0bJlS8nc3Fzx8+zZsyVACg8Pf2+9OR05XLp0Kdvfu/B9E7uVPpN69epx+vRpmjVrxpUrV/j9999p0KAB9vb27NixI0v6ypUrU7ZsWcXPTk5ONG/enH379pGRkYEkSWzevJmmTZsiSRIRERGKV4MGDYiJieHixYsAbN68GQsLCwYNGpSlnk/Zkvqubdu2IZfLGTt2LGpqyn9Cn1p+ly5dOHPmDPfu3VMcW7t2LY6OjtSsWfO9+WJjYwGU1hk+ZPfu3airq+Pj46N0/JdffkGSJPbs2aN0vG7dukojrRIlSmBkZMT9+/cVx0xMTDhz5gxPnjzJts7Lly9z584dOnbsyMuXLxW/t4SEBOrUqcOxY8eQy+VKefr27av0c/Xq1Xn58qWiv6/XhbZv354lb269Hhns27ePxMTE/1SW8O0QweEzKl++PFu2bCEqKoqzZ8/i6+tLXFwcrVu35ubNm0ppCxcunCW/m5sbiYmJhIeHEx4eTnR0NIsXL8bS0lLp1b17d+DNgvK9e/dwd3f/pEXknLh37x5qamoUKVIkz8ps164d2trarF27FoCYmBh27txJp06dPhhwjIyMAIiLi8tRPQ8fPsTOzi5LMPH09FS8/7bspqpMTU2JiopS/Pz7779z/fp1HB0dqVChAuPHj1cKHnfu3AGga9euWX53S5cuJSUlhZiYmA/Wa2pqCqCot127dlStWpWePXtibW1N+/bt2bBhwycFigIFCjB06FCWLl2KhYUFDRo0YMGCBVnaJHxfxJrDF6ClpUX58uUpX748bm5udO/enY0bNzJu3Lgcl/H6P/1PP/1E165ds01TokSJPGmvKpiamtKkSRPWrl3L2LFj2bRpEykpKfz0008fzOfq6oqGhgbXrl37LO1SV1fP9rj01g7wtm3bUr16dbZu3cr+/fuZPn0606ZNY8uWLTRs2FDxu5s+fTqlSpXKtjwDA4Nc1aurq8uxY8c4cuQIu3btYu/evaxfv57atWuzf/9+1NXV3xtUs7vmY+bMmXTr1o3t27ezf/9+fHx88PPzIzAwEAcHh2zLEb5tIjh8YeXKlQPg6dOnSsdff7t82+3bt9HT08PS0hLInDrJyMigbt26H6yjUKFCnDlzhrS0NDQ1NfOo5crly+Vybt68+d4Pu+x8bMqpS5cuNG/enHPnzrF27VpKly5N0aJFP5hHT0+P2rVrc/jwYR49eoSjo+MH0zs7O3Pw4EHi4uKURg/BwcGK9z+Fra0t/fv3p3///rx48YIyZcowefJkGjZsqJiWMjIy+ujvLjfU1NSoU6cOderUYdasWUyZMoVRo0Zx5MgR6tatqxhtREdHK+V7d3T0WvHixSlevDijR4/m1KlTVK1alUWLFvHbb7/lWZuF/ENMK30mR44cUfp2+dru3bsBcHd3Vzp++vRpxZoBwKNHj9i+fTv169dHXV0ddXV1WrVqxebNm7Pdsx8eHq74d6tWrYiIiGD+/PlZ0mXXptxq0aIFampqTJw4Mcs0xofK19fXB7J+WL3WsGFDLCwsmDZtGkePHv3oqOG1cePGIUkSnTt3Jj4+Psv7Fy5cYOXKlQA0atSIjIyMLOdm9uzZyGQyGjZsmKM6X8vIyMgy/WJlZYWdnR0pKSkAlC1blkKFCjFjxoxs2/f27y6nIiMjsxx7Hahf1/s6KB07dkypvYsXL1bKFxsbS3p6utKx4sWLo6ampihL+P6IkcNnMmjQIBITE2nZsiUeHh6kpqZy6tQp1q9fj4uLi2Kd4LVixYrRoEEDpa2skLll9LWpU6dy5MgRKlasSK9evShSpAiRkZFcvHiRgwcPKj4wunTpwqpVqxg6dChnz56levXqJCQkcPDgQfr370/z5s3/U99cXV0ZNWoUkyZNonr16vz4449oa2tz7tw57Ozs8PPzyzZfoUKFMDExYdGiRRgaGqKvr0/FihUpUKAAAJqamrRv35758+ejrq5Ohw4dctSeKlWqsGDBAvr374+Hh4fSFdIBAQHs2LFD8e23adOm1KpVi1GjRhESEkLJkiXZv38/27dv5+eff852m++HxMXF4eDgQOvWrSlZsiQGBgYcPHiQc+fOMXPmTCDzG/7SpUtp2LAhRYsWpXv37tjb2xMWFsaRI0cwMjLC398/V/VOnDiRY8eO0bhxY5ydnXnx4gV//vknDg4OVKtWDYCiRYtSqVIlfH19iYyMxMzMjH///TdLIDh8+DADBw6kTZs2uLm5kZ6ezurVqxVfSITvlMr2SX0hrVq1koyNjbNs23xbSkqK1LNnT8nAwECysrKSZs+erXgvNDRUql27tqStrS25ublJhw4dylG9e/bskby9vSUPDw/JwMBA0tLSklxdXaVBgwZJz58/V0rLWxfBvb5orXTp0tlu4Xz+/Lk0YMAAydHRUdLU1JRsbGykOnXqSIsXL1ZKl5iYKI0aNUoqUKCAIl3r1q2le/fuKdX7KVtZX/v777+l0qVLS9ra2pKpqalUs2ZN6cCBA4r3393KKkmStH37dqlIkSKShoZGtttaz549KwFS/fr1szutH3ThwgWpY8eOkp2dnaSpqSmZmppKderUkVauXKm05TYuLk4aMmSIIl3hwoU/eBHcu5ydnaWuXbtKkpT5tzNs2DCpZMmSkqGhoaSvry+VLFlS+vPPP7Pku3TpkvTjjz9K5ubmkra2tuTs7Cy1bdtW6W/q9bl+d4vqu9tSDx06JDVv3lyys7OTtLS0JDs7O6lDhw7S7du3lfLdu3dPqlu3rqStrS1ZW1tLI0eOlA4cOKC0lfX+/fuSt7e3VKhQIUlHR0cyMzOTatWqJR08eDDH51749nzzwaFjx45S586dPxgcpk+fLgHS//73P+nHH3+UAOn06dOSJElS48aNJX19femPP/6QihUrJpmZmUnx8fF52sb3fQh9jy5fviwB0qpVq1TdFEH4ruXbNYeUlBRiY2OVXtnNj65du5aOHTt+sKzVq1djY2PD1KlTmT59OgArVqzg5cuX7N69m6ZNm+Lj48OQIUOIjIzM9RSAkHNLlizBwMCAH3/8UdVNEYTv2hcLDsnpefvy8/PD2NhY6fW+ue6PefDgAXZ2dgCKbXsPHjzg4cOHSJKU7XtC3vL392fatGksXryYXr16KRavBUFQjS+2IC3P4zuD+/r6MnToUKVj2tran1ye9Kp9Ujbt/NB7Qt4YNGgQz58/p1GjRkqL8IIgqMaXCw7/7Qr/LPS0tT85GMjlclJTUxW3onZxcVHc+iAsLAzIvGrU2dkZmUyW7Xt5SQQdCAkJUXUTBEF4yxebVpJLUp6+cmr9+vXs27cPQHF3y927d6Orq8vo0aMB6Ny5M8+fP2fEiBEMGzYMyNwOam5uTsOGDdm5cydz585l9uzZmJqa0rRp07w/QYIgCF+RLzZyyJCr5tvx//73P8UVof7+/vj7+7N8+XKlND4+PgQHBzN//nz09PSYMWMGVapUAWDhwoV07dqVYcOG4ezszPr168V8uCAI37wv9pjQlwnpH0+UC+b64vo9QRCEzyXfrjkIgiAIn0++3a0kCIIgfD4iOHwGvr6+JCcnq7oZwjuCgoL+84NxPhdLS0usrKxU3Yz30tHR+eTriIT86QsuSH+pmlQvOTk5y7OIBdUbMmSI+L18oiFDhqi6CcIXJkYOAvB5RzsvXrz4pNtS54SampriKW45acfX8CEnvoUL+cEXXJAWweFrll9HO6oaDfyXYBoaGvrJQUoEFuFL+XLTSmLkIHxDVBVMv4aRj/B9EFtZBUEQhCy+WHAQ9w/6/P7LVMd/mY8XUx2C8O0R00rfEDHVIQhCXhHTSoIgCEIWYiurIAiCkIXYyioIgiBkIdYcBEEQhCw+OThkZGSQkZGh+FlLS+uD6UVsEARByD9y/SS4devW4ejoiJaWFrq6uujq6qKnp/fRfHK5lKcvQRAE4fPJ9chh0KBBpKamUqVKFTQ0cp5dTCsJgiDkH7kODtbW1gwcOJB+/frlKp+qHhMqCIIg5F6ug0PZsmWZNm0asbGxmJqaAiCTyejVq9cH84mRgyAIQv6R6+CwZs0aIPNWDa/lJDiki5GDIAhCvpHr4LB8+fJPqkiMHARBEPKPXAeHrl27flJFYs1BEAQh/8j1Vtbz589Tvnx59PT0UFdXR11dPUe7ltIleZ6+BEEQhM8n1yOHfv368fTpU5KTkylXrhzXrl2jRo0aH80n1hwEQRDyj1yPHG7cuMHPP/+MTCZjypQpTJ48OYcjBylPX4IgCMLnk+vgoK2tjZmZGerq6ixYsIDdu3cTEBDw0XwZcnmevnIjICAAT09PdHR08PLy4uHDh1nShIWF0aRJEwwNDXFxcWHDhg2K97y8vJDJZIpXpUqVclW/IAhCfpPr4ODl5cXLly9p3bo127dv5/DhwzRq1Oij+dLlUp6+ciopKYk2bdogk8mYPn06Fy9epE+fPlnS9enTh/379zN58mSsra3p3LkzYWFhivc9PT1Zt24d69at47fffstx/YIgCPlRrtcctm7dCmTeeK9Lly4ANGjQ4KP5VLWVde/evURERPDbb7/Rp08fzp07x+rVq3n69Cm2traKdEePHqVYsWL4+Pigra1N3759+ffff/nll18AsLKyokmTJhgYGKikH4IgCF9SroNDQkIC//77LyEhIYq7sh4/fpzJkyd/MF9erxOkpKSQkpKidExbWxttbW2lYw8ePADAzs4OAAcHBwBCQkKUgoOlpSUhISEEBwdz6tQpRZrXjh07hqGhIebm5syYMYNu3brlaX8EQRC+JrkODk2bNuXo0aNIb33Yy2SyjweHjLwNDn5+fkyYMEHp2Lhx4xg/fny26V+3V3pPkPLz86NLly54enpiaGgIgI6ODgCtWrWib9++SJLE8OHD6d27Nw0aNFAKLoIgCN+SXAeH8+fP07lzZ7p27YqaWs6XLNLyODj4+voydOhQpWPvjhoAXFxcAHjy5AmAYh3BycmJ5ORkNDQ00NDQoF27dnh5eXH//n1u3LhBr1698PT0BDLvRPvahQsXmDlzJvfu3RPBQRCEb1aug8OwYcM4f/48BgYGGBsb5zhfWh5f55DdFFJ2GjZsiLm5OXPnziU9PZ1t27ZRr1497ty5g4ODA//73/+YOnUq69at4+XLl8jlcvz8/LC1taVDhw4AVK1albZt26Kurs6qVavQ19dXBA5BEIRvUa6Dg5qaGjt37mTnzp2KYzKZjPT09A/my+uRQ07p6uqyYcMG+vfvzy+//EKlSpVYvHix0noCQHx8PBMmTCA2NpayZcuyYMECdHV1AbC3t8fPz4/Y2Fg8PDyYNm0a5ubmKuiNIAjCl5Hr4DBt2jQKFChA5cqVc/ewHxVeIV27dm2Cg4OVjrm4uCitP/Tq1eu9d5Z9+5oHQRCE70Gug0PNmjWpXr06w4cPz1W+vJ5WEgRBED6fXAeH0NBQfH19WbVqFUZGRkDmtNLJkyc/mE9V00qCIAhC7uU6OFy7dg2AmzdvKo7JZLKP5svrrayCIAjC55Or4CCXy7l06RJWVla53sYpppUEQRDyj1zdW0lNTY1atWqxZ8+eXFeUliHl6UsQBEH4fHI9rdS5c2c2bdpE3bp1sbGxURzX0tL6YD4xchAEQcg/ch0c5s2bh0wmo0CBAopjObnOQS6CgyAIQr6R6+BQo0aNHC1Avys9QzzaUxAEIb/IdXDIyYN9siMeEyoIgpB/5Do4pKSksHTpUm7cuEFycjKQOa20bNmyD+YTIwdBEIT8I9fBoWfPnqxduxaZTKa4/UTOgoMYOQiCIOQXuQ4Oe/bsYdCgQcyfP5/169ezZcsWihYt+tF86bl87rMgCIKgOrl+hnRCQgLFixdHkiTCw8Nxd3dn5syZH80nl0t5+hIEQRA+n1yPHAoUKEBMTAyenp6Kh+C4urp+NJ+YVhIEQcg/ch0c/v33X7S1talbty5TpkxBJpMxcuTIj+YT00qCIAj5R66DQ4kSJXj+/DmnT59m9uzZpKenY2pq+tF8YuQgCIKQf+R6zWH//v24urrSunVrgoODadOmDX369PlovvQMeZ6+BEEQhM8n18Hhl19+oVSpUoptrG3atOHo0aMfzZeRIeXpSxAEQfh8ZNLbz8rMAUNDQ0aPHs3IkSM5cOAAt2/fZtiwYcTFxX0wX/WZJ/5TQ991/JdqeVreu5I/fKuozybgVrhK6h234+bHE30mY5t6qqTekJhEldRroaepknqbF7dXSb06uZ68Fr4Guf61eXp6sm7dOgDWr1/P7t27KVas2Efzie2ngiAI+Ueup5VmzJhBaGgokiSxZMkSEhISmD59+kfzZWTI8/QlCIIgfD6fdFfWu3fvcvr0aQAqV66MmZnZR/OJdQJBEIT8I9fBITU1lTVr1igWoR88eEDv3r0/+rAf8W1fEAQh/8h1cOjWrRv//vuv4udt27YRGBjImjVrPphPjBwEQRDyj1yvOezZs4eOHTvy7Nkznjx5Qrt27di9e/dH84k1B0EQhPwj1yOHunXr4uHhgZWVFQBFihQhIyPjo/nEbiVBEIT8I9cjhzt37jBu3DiKFi1KkSJFGDduHHfu3KFKlSpUrVr1vfnS0+V5+sqNgIAAPD090dHRwcvLi4cPH2ZJExYWRpMmTTA0NMTFxYVNGzco3nv06BEN69fBxECHEkXdCThyOFf1C4Ig5De5Dg5Xr15FkiSCgoIIDg5GkiQuX75MYGAggYGB782nqiukk5KSaNOmDTKZjOnTp3Px4sVsb/fRp08f9u/fz+TJk7G2tqZHt86EhYUBMHhgP86dPcNkv9/R0tSiU/s2JCQk5PbUCYIg5Bu5Dg5yuRy5XM7169fZuHEjN27cUBz70PSSPEOep6+UlBRiY2OVXikpKVnq3bt3LxEREQwePJhBgwbRokUL9u3bx9OnT5XSHT16lGLFiuHj44O3tzepqals3PAvL1++ZO+e3TRq0pQBg3wYNHgIkZGR7Nrpn9tTJwiCkG/kODi0b9+eFi1aAJk7lEqUKEHbtm0pXrw4ixYt+mj+jIyMPH35+flhbGys9PLz88tS74MHDwCws7MDwMHBAYCQkBCldJaWloSEhBAcHMypU6cAeBgSQujDh0iShK1tZn77V/kfhjzI6akTBEHId3IcHI4ePUqVKlUAmDVrFpIk0b9/f8qVK5ejK6TzeuTg6+tLTEyM0svX1/e99b++hdT7biXl5+dHQkICnp6ebN26FQAdHZ0c5xcEQfiW5Dg4REVFYWdnR1JSEoGBgdjb2zNv3jz69u3LkydPPpr/9dRTXr20tbUxMjJSemlra2ep18XFBUDRxtfrCE5OTiQnJ5OennmHvXbt2hEaGsqpU6eYNWsWAB4enjg5OyOTyXj6NDP/k1f5nV0K5PTUCYIg5Ds53srq5OTE0qVLOXHiBOnp6TRs2BDI/LDNye0z5Cq6NqFhw4aYm5szd+5c0tPT2bZtG/Xq1ePOnTs4ODjwv//9j6lTp7Ju3TpevnyJXC7Hz88PG1tb2rbvgK6uLg1+aMieXTtZMG8uy5ctwdTUlMZNmqqkP4IgCF9CjkcOEydOJDAwkMWLF2NmZsYvv/wCZD42tFKlSh/Nn9fTSjmlq6vLhg0bkMvl/PLLL5QuXZrFixdnSRcfH8+ECRMYNmwYBQoUYOv2Xejq6gLwx/yFlC1XnpEjhpGSmsLqf9ajr6+f4zYIgiDkN7l6nkNERAQPHjzA09MTAwMD0tPTuXHjBjY2NlhbW38wr3nXdf+5sW97ubJDnpb3rtw+z+HC+XOs+HsZQTevEx4ezuy5C6hdp26u6337eQ7H92zm8LZ1xEZHYu9SiFY9h+BcuEi2+TLS0zmwZTVnj+whJjICKztHmnXuh2cZ5cAd/TKcHasXEnQxkLTUZCxsHOg4cCTLbr4JuK3L2NGpoiPmBlrceRHPzP13ufk0++d1qKvJ6FbZiUbFrbE01Cb0ZSLzA+4TeD9KkaaUozE/VXTEw8YAS0Nthm26zrE7LxXvv36ew4k9WziyfR1x0ZHYuRSiZY+fP9jfg1tWcz5gLzGREVjaOdKkcz88S1fM0t+daxYSfPEMqa/622GAL46uHornOVw5tIOLezeRGBOJhWNBanbqj01Bj/fWe373vwSdPEhCVASmNg5UadMDl+LlFWkCt63m7A7l28mY2jjQecoy4M3zHM7s28ZJ//XEx0Ri7VSIxt0H4eCa/bMtMtLTObb9Hy4f3UdcVATmto7U79ibwqUqKNIc3riCgM2rlPJZ2DniM2slkPvnOeTV37R4nkP+lKtfm4WFBRYWFm8ya2hQsmTJHOVV1bTSl5KUlIi7uzstfmzF0MED/3N5F08cYuvy+bTt8ysubkUI2LmBhROHMmreOgxNsj6ze9c/izl/bD/t+v0Pa3sngi+fZdnvI/l5yiIcCroBkBgfyx8j++FarAx9x8zAwMiE8KeP0TMwBGIAqOtpyeA6hZi29zY3nsTRvrw9f7QrTtvF54hKTMtSb98aLvxQzBq/PbcJeZlIpQKmTPuxKL1WX+b283gAdDXVufMiHv+rT/m9VfbP/rh08hDbV8ynTZ9fcCpchGM7N7J40i+MmPcPhsZZ+7t73RIuHNtP277DsbZ3JvjyGZb/PhKfyQvf6m8c80b1x7VYaXqNno6BkQkRTx+ja2CoKOf22QCOr19M7c6DsC7oweUDW9k+axSdpyxDz8gkS72BW1cQfPowdbr9jKmNIw9vnGfX/Im0GTkbK2dXRToze2da/jpV8bOamrpSOddOHWHv6oU07fkzDq6enN69mVV+/8Nn1koMsunvofV/c+XEAZr3/gULOyfuXjnHuplj6TVxHrYFCivSWTm40HX0jPfWmxt5/Tf9qbp168bKlSsJCgpCR0eHAgUK0K5dO6V7vAl5L9fXOXwqVU0rfSnVqtdk4OAh1KlbL0/KC/D/lyr1mlKpTmNsHAvQts8wtLR1CDy8M9v0547uo16rzhQtWxkLG3uq/dASzzKVObzjzX+gg1vXYmJhRadBI3EuXARzazs8SlXAwubNN8oOFRzYfuUpO68958HLRKbuvUNyupymJWyyrbdhMWtWngrl1L1InkQns+XSU07fi6RjBQdFmtP3I/nrWAhHb7/MtgyAo/7rqVS3KRVqZ/a3dZ9f0dTW4eyhXdmmv3B0H3V/7EyRspUxt7Gj6g8t8SxdmQD/N/09/Kq/HQa+6a/7O/29tG8LxWr8QJHqDTC3d6Z2Fx80tLS5eXxftvUGnzpEucbtcSlRAWMrW0rUaopLifJc2rdZKZ2amjr6xmaKl66hsdL7p3ZtpGztRpTxaoiVgwtNew5BU0ubiwF7sq33yokD1GjRCbfSlTCztqNC/ea4la7IyV0bletVV8fQxEzx0jcyzra8nMjrv2khf/liA76c3H9JyJSelsaje7ep+2NnxTE1NTXcSpQj5NaN9+bR0FTeraWppc2DoKuKn6+fO4lHqQosnz6auzcuY2xuSbUfWlKlXjMANNRkeNgYsvJUqCKPBJwLiaK4vVG29WppqJHyzu1MktPllHTI+YdSeloaj+/dpk7Ln7L29/aH+qt8m3hNbS0eBF1T/Hzj/AncS1Vg5Ywx3HvV3yoNWlD5VX8z0tN48fAO5Rq3V+SRqanhWKQ0T+9l/9jUjPSs9WpoavPkjnI7o5+HsWxIB9Q1tbB19aRKK28MzTPvR5aensbTB7ep0aKjUn8LFS/L49vZ15tdfzW0tAkNvqZ07OWzMKb3a4OGphaOhYtQt0NPTCw+POX7tXj27BkNGjTg7t27aGhoUKNGDf7++29VN+u79cWCgyRuvJdjCXExyOUZGJoo7wIzNDHjRVjW+0IBeJSuQID/vxQqUhILG3tuX73A1cCjyOVvPrhfPn/CyX3b8GrajnqtuhB6N4gty+agoaEJOGOip4mGmozId6aPIhPScDbXy7bewPuZo4TLj2J4HJVEeRdTarlboCaT/ff+Gpu+t7/upSpw1H89hYqUxNzGnjvXLnAt8Ng7/X3KqX3bqdm0LXV+7Myju8Fs/fsPNDQ0KV+rIUlxsUhyeZbpIz0jU6KePsq2XqdiZbm0fzP27sUxtrTlUdAl7l08qVSvTUEP6vX4FVMbBxJiIjmzfQ2bpv5Cp4l/oaWrR2JsDHK5HP13po/0jU0JDwt9t0oAXEuU49Tujbh4lsDU2o771y8SdPa4Ur0Orp607DccC1tH4qIjCdi0kmXjBzNw+t9o62b/+/sSUlJSsty9QFtbO8vWc3V1ddq2bYu1tTUPHz5kypQpTJ06FUE1vlhwkIuRw2fVynsw/y78nSk+nZAhw8LGjoq1G3Hm8JtpGUmS41jIg6Y/Zd5byqGgG09DH3By3zaoMviT6p114B4jG7mxvnd5JCAsKomdV5/R5D3TUHmlpbcPGxb+ztTBPyFDhrmNHRXe09/Gnd7q76P7nNq/nfK1Gn5SvTU69OPwyjmsHtkTZGBsaYdn1frcPPFmGsqlxJvFaQvHgtgU9GD5sM7cOXeMojV++KR6G3UbyPbFM5k7tBsyGZha21Ha6wcuHnkzDeX21mK8jXMhHFw9mTWwA9dPB1C2dqNPqjcv+Pn5MWHCBKVj48aNY/z48UrHkpOT2b59O+fPn1dcbPp6w4vw5X25fQQZWRczhezpGxqjpqZOXHSk0vG46EgMTcyzzWNgbErPEX6kpaaQEBeLsZkF/qsXYm5tp0hjZGKOjYOLUj5rB2euBAYgA6IT00iXS5i92k3zmpm+JpHxqdnWG52UxvDNN9BSl2Gsq0l4fCoDvArwJDr5v/c3JuqD/fV+p7871yzC3Eq5v9YOzsr9tXfmamDmUwx1DY2QqamRGButlCYxNgq9bBaFAfSMTGgyaDzpaakkx8eib2LOqU3LMLZ8/weYtp4BJtYORL948qoMY9TU1EiIiVJKlxATlWX09Jq+kQkdf51EWmoqSfExGJpacOCfJZha2763Xl19A8xtHYh8HvbeNF+Cr68vQ4cOVTqW3QWr8+fP59y5c4waNYqaNWvSpEkTkpNz/nck5K0vtiBNRlrevr5hGpqaOBZy4/bVC4pjcrmc21cv4OJe9IN5NbW0MTG3RJ6RwZXAoxQrX13xXgHP4rx4ojxt8eLJI0xffbClyyWCn8VR3uXNB6MMKO9syrWw2A/Wm5ohER6firqajFoelkpbVT9GQ1MTh0Ju3Lmm3N87Vy/g4pbz/l4NPEqxCtUU77l4FOfFE+XpofCnjzB71V91DU2snAvzKOiS4n1JLudR0GVsC2W/hfZNm7UwMLVAnpHB3QsnKFi68nvTpiYnERP+BH3jzA9+DQ1NbAu4cf/6RaX+3r9+EQe3D9erqaWFkVlmf2+ePYZH2fffJj8lOYmo50/eG2C/lJzezeC1qKgotm/fTmpq9l9IhC/jC44ccnnhQD6TmJBAaOibD96wx48JDgrC2NgYWzu7D+TMnlfT9qydNxknVw+cCnty1H8DqSlJVKzdGIA1f0zC2NySpj/1BSDk9g1iIiOwd3ElJjKCPev/RpLk1Gn5ZtHTq0k75ozsy/5NqyhdtTYP79zk9IEdtOs7HP/MnaysO/uYsU08CHoWx81XW1l1NNXYefUZAOOauBMel8qfRzNvPFjUzhBLA21uv4jHykCbntWdUQNWB745F7qaajiY6ip+tjPRobCVPrHJ6TyPzZyLrtm0HevmTcGx0Kv+7txIakoSFV5Nh/wz9zeMzCxo8qq/DxX9LUxMZDj7NvyNJJdT+61F3ppN2zJ3ZD8Obl5FySq1Cb0bROABf9r0HaZIU7rBjxxYOgNrFzesC7hz+cBW0lOSKVKtPgD7l/yOvqkFVVt7A/DsXjDx0RFYOhYiPjqCM9vXIMklyjZsqyjz+PrFFChVCSNzKxKiXxK4bTUymTpuFb0Uaao0bsPWhVOxK+iOg6sHp3dvJjUlmTI1M6edNi/ww8jMgnodegHw6E4QcVHh2Di7EhsZwZFNK5EkiWrN3iym7129EPeyVTCxsCYuKoLDm1YiU1OjeNXaOfiLyyqv/6Y/ZuDAgRw4cIAVK1YwcOBAjI0/faeV8N+JaaU8cuPGdXp276L4ecbvmXeIbda8JZOm5H5RrUy1OsTHRrN73VJioyNxKOBK3zEzMXo17RAV8RyZ2puBX3paKrv+WcLL50/Q1tGlSJlKdB48Bj39N3v6nQt70uN/U9i55i/2bVyBuZUtLb19KFezPv47MnfJHAwKx0RPk97VXTDX1+L2i3h+3nBNsUhtbaTD23sLtNTV6FvTBTsTXZJSMzh17yXj/YOJT3mzxuRpa8jCTqUUPw+pm3k9wM6rz5i06xYApavWIT4mmr3/Lsu86K+AK71Hz1BMs0RFPEf21iJ3Wloqe9Yt4eXzp2jr6OJZphIdfcag+1Z/nVw96T58MrvWLmb/xpWYWdnSvPsgytaor0jjVsGLpLgYAretIiEmCkvHgjQfMlkxrRQXGa58ntNTOb1lJbHhT9HU0cWleHnq9xyOtp6BIk18VAT7FvmRlBCHrqExdoWL0nb0HKWF7+JVapEYG83hjcuJj47CxrkQnUdMw+BVf2MiXiCTKf9+D61fTtSLJ2jp6FK4VEVaDfBFV/9NvbGREWya9xuJcbHoGxnj5F6c3pPmo5/N9Ro5kdd/0x/j6OjIxYtvRlPTpk1T/HvFihWKf4ubX34ZubpC+r/QrTMlT8tLOjQyT8t7V26vkM4rb18h/SWN25H9Fsov4fUV0l/a6yukvzSLd9Z0vpTcXiGdV8QV0vnTl/u1pX/b00qCIAjfEjGtJAiCIGQhgoMgCIKQhditJAiCIGQhRg6CIAhCFl8sOCRd+ONLVSUIgiD8R1/uCmlBEAQh3xDBQRAEQchCBAdBEAQhCxEcBEEQhCxEcBAEQRCyEMFBEARByEIEB0EQBCELERwEQRCELERwEARBELL4Ys9zEHJGVc+RMC0/UDUVA+tWjlZJvQEPolVSb0Unw48n+gxMtbRUUu8PRS1VUq/w34iRgyAIgpCFCA6CIAhCFiI4CIIgCFmI4CAIgiBk8d0Fh8DAQA4fPqzqZgiCIHzVvqvgcP/+fapUqcKgQYMICAhQdXMEQRC+Wt9NcEhMTKRgwYKMHDmSoKAg/ve//3HkyBFVN0sQBOGr9F0Eh/3791OhQgVu3LjBb7/9xoQJEzh37hwjRowQAUIQBCEb33xwyMjI4MiRI9y8eZMuXboQFBTEmDFjlAKEmGISBEFQ9k0HB7lcjrq6OsOHD2fMmDFcvXqVNm3aZAkQvXv35vjx46puriAIwldDQ9UN+FzkcjlqamqKf3fr1o24uDhWrlxJ69at2bRpE2PGjCExMZG5c+fi4OCg4hYLgiB8Pb7JkYMkSYrAMGfOHMqUKUP79u2JjIykbdu2vHjxgvbt23P9+nX8/Px4+PAhBQoUUHGrBUEQvh7fZHCQyWQAzJgxgwkTJvDjjz9y7tw59u/fz7Nnz2jfvj3Xrl2jV69epKamYmFhoeIWC4IgfF2+yeAA8OzZM7Zs2cKQIUMICgqiQIECmJubc+TIEa5fv86QIUNYuXIlWiq6U6UgCMLX7JsNDjY2NixevJjY2FgOHTrElClT6NmzJ0lJSTx+/Jiff/4ZNzc3VTdTEAThq/TNLkgDFCtWjIoVKyJJEuvWrUMulzN69Gh++uknHB0dVd28PHHh/DlW/L2MoJvXCQ8PZ/bcBdSuU/c/ldmnbQ2GdK2DtbkR126HMXTaRs7fePje9AM7etGrTXUcbUx5GZ3A1oOXGDNvBympmQ+nCN41AWc78yz5Fq0/xpCpGxQ/n9q7lWM7/iUuOhJb50I09x6MY2HPbOvMSE/nyNY1XDi6j9jICCztHGnYqQ/upSsqpYt5Gc6etX9x69IZUlOSsbCxp82AETgU8lCkqVbAhNqu5hhqq/MkNoXNV58TGp2cbb1qMqhX2JzyTsYY62jwIj4V/5vhBL9IyDZ9ncJmNC1ixdF7kWy9/kLpvbP7tnHSfz3xMZHYOBWiYfdBOLi+v7/Ht//DlaP7iI2KwMLWkbode1O4VAVFmiMbV3B08yqlfOZ2jgyatVLp2PE9mzm8bR2x0ZHYuxSiVc8hOBcu8t56D2xZzdkje4iJjMDKzpFmnfvhWaaSUrrol+HsWL2QoIuBpKUmY2HjQMeBI3Fy9ci2XOHr900HB4CqVatSvXp1Dhw4gJqaGrNmzfqmFp+TkhJxd3enxY+tGDr4vz+wp3X9Mkz7pSWDJq/n3PUQBnasxY4/B1CyxUTCo+KzpG/3Qzkm+TSn7/i1nL5yn8LOViyZ2BkJ+N/MLQBU+2k66moyRZ4irnbsXjSILQcuKY5dOXmYnSsX0LL3UJxci3Bi10aWTf6VX/9Yg4GxaZZ69/27lEvHDtCq7zAs7Z24ffksq6aPpv/kBdgXyBwRJsbHsXDMQAoWLYX3yN/RNzIh4tljdPXfPGyntJ0hLYpaseHqcx5GJVGzoBl9Kzsy5dB94lMzstTb2NOSsg5GrL/8jBfxqXhY6eNdwZ4/jj8kLCZFKa2jiQ5VnE0Ii8kaaK6fOsK+1Qtp0vNn7F09Cdy9mTV+/2PgrJXZ9vfw+r+5euIATXv/goWdE/eunGP9zLH0mDgP2wKFFeksHVzoMnqG4mc1NXWlci6eOMTW5fNp2+dXXNyKELBzAwsnDmXUvHUYmmStd9c/izl/bD/t+v0Pa3sngi+fZdnvI/l5yiIcCr4+z7H8MbIfrsXK0HfMDAyMTAh/+hg9A9U81CivhYSEsGLFCqpVq0bduv/ti1d+8s1OK71mZ2fH1q1bOXHiBEFBQbi6uqq6SXmqWvWaDBw8hDp16+VJeT4/1Wb5llOs3hFI8P1nDJr8L0nJqXRtUTnb9JVKFuD05fus33ue0KeRHAoMZsPe85Qr6qxIExEVz/OXcYpXo+rFuBcazvELdxRpju/cQIU6TShfqxHWji607P0Lmlo6nDu8O9t6Lx7bT60ff8KjTCXMre2o3KAFHmUqcdz/zUjk6LZ/MDa3pO0AXxwLe2JmbYtbyfKY29gr0ni5mnH6YQxnQ2N4HpfKxivPSM2QU9HZONt6yzkacfDOS4JeJPAyMY2TIdEEPU+gViEzpXRa6jI6l7Vj/ZVnJKXJs5RzetdGytRuRGmvhlg5uNCk5xA0tbS5FLAn23qvnDhA9RadcCtdCTNrO8rXb07h0hU5tWujUjo1dXUMTcwUL30j5X4E+P9LlXpNqVSnMTaOBWjbZxha2joEHt6Zbb3nju6jXqvOFC1bGQsbe6r90BLPMpU5vONfRZqDW9diYmFFp0EjcS5cBHNrOzxKVcDirfOcn4WEhDBhwgQOHjyo6qZ8Ud98cAAwMTGhdOnS4lqGj9DUUKe0pyOHz9xSHJMkicNnblGhRPajrcArDyhdxFERDFzszWlQtSh7T9x4bx3tG5Vn5fbTimPpaWmE3b9N4RJlFcfU1NRwLVGW0NvZl5ORloaGpvJmAk0tbUKCryl+vnn+JA6FPFgzcywTezTnj2E9OHPQX/G+ugwcjHW4Hf5mSkgCbocn4mKqm229GmpqpGUoP1k3LUNOQXM9pWOtS9hw83k8t8MTs5SRnp7Gkwe3KVhcub8Fi5fl8e2bOe6vhpY2oW/1FyDyWRgz+rVhjk8nNs+bTHTE8zf1pqXx6N5t3EqUU6rXrUQ5Qm5lf57T09LQ0NRWOqappc2DoKuKn6+fO4ljIQ+WTx/NqG5N+P2X7pw6sCPb8j5m+fLleHh4oKOjQ5EiRUhMTGTQoEHY2Nhgbm5Op06diIyMBMDLywuZTMawYcOwsLCgdOnSPH78+KPvXbt2jVq1amFoaIirqyurV69+b/03b96kVq1aAEybNg2ZTEZISAjXrl2jZs2aGBgYULBgQf78808AAgICkMlktGjRgpIlS2JlZcXWrVtp0aIFhoaGDBo0SFHXokWLcHV1xdDQkDp16nD37t1POmefy3cRHIScsTA1QENDnReRcUrHX7yMxcbcKNs86/eeZ9LCXRxaPoTYs38QtHMCxy7cYfrf+7NN36xWCUwMdVnjf0ZxLDEuBrk8I8t0iqGxKXHRkdmW41ayPMd3biDi6WPkcjm3r5zj+pljxEa9VKSJfPGUwP3bsbB1oMfo6VSq35wdf8/lQsBeAPS1NVBXkxGXovzg7riUdIx0sp9xDX4Rj1chMyz0NZEBbpZ6lLA1xEj7zfRNaXtDHEy02XkzPNsyEmNjkOTyLP3VNzYl/j39LVSiHKd3b+Tlq/7eu3qeoLPHldI7uHrSot9wfhoxlSY9fibqxVOWjx9MSlJmgEp4dZ4NTZRHOYYmZsRFvyQ7HqUrEOD/Ly+ePEIulxN8+RxXA48S89Z5fvn8CSf3bcPC1pF+Y2dRrUELtiybw9kjmaOglJQUYmNjlV4pKSlZ6goICMDb2xttbW3mz59PgwYNGDNmDPPnz6dly5b079+ff/75h8GDByvlCwkJoW3btly+fJmFCxd+8L20tDSaN2/OgwcPGDFiBB4eHnTr1o1Lly5lW7+9vT1jxowBoFmzZqxbtw5TU1OaN2/OlStXmDZtGs7OzgwYMIBDhw4p6j116hQ9e/bk5cuXtGrVCjc3Nzw8PJg/fz537tzh0KFD9OvXjyJFiuDr68udO3do27Zttr8DVfnm1xyEz6t62cIM827AYL/1nLv2kEKOFswY1pqnvX5g6pK9WdJ3bVGFfSdv8jQ85j/V27S7D5v/ms6MwZ2RyWSYWdtRrlZDpWkoSS7HvpA7P3TsDYB9ATeePXpA4P7tlPX64ZPq3XLtBe1L2TCyTkEkCV4mpnLmUQwVnTKnb0x0NPixmDV/nn5Eulz6SGk517DbQHYsnsn8od1ABmbWdpT2+oFLR95MQxV+ezHeuRD2rp7MGdiBG6cDKFO70SfV28p7MP8u/J0pPp2QIcPCxo6KtRtx5vAuRRpJkuNYyIOmP/UBwKGgG09DH3By3zYq1GqIn58fEyZMUCp33LhxjB8/XunYrl2ZZc6bN48aNWoAUK5cOfT09Jg/fz7q6ups2LCBPXuUp97++OMPEhMTWbhwISEhIR9879atWzx48ACA0aNHK9IdOXKEp0+fZqkfoHbt2kyaNAlPT0/FxbMPHjygd+/eDBgwgDJlylClShX27NlDkyZNMs9bq1YMGjSIadOmER0dzdSpUxk/fjznz58nJCSEffv2AeDv74+/f+Zo9tGjR0RGRmJmphy8VUUEB0EhIiqe9PQMrMyUFxKtzI149jI22zzj+jdm3a6zrNiaOU104+4T9HS1WTC6A9OW7kOS3nxAOtmaUruiO+1/XaJUhp6hMWpq6sTHRCkdj4uJyvIt9zUDYxO6Dp9MWmoKiXGxGJlZsGftX5hZ2ynSGJqaY+3gotwXe2euBx4DICElnQy5hKG28n8DQ20NYpOVRxOvJaRmsOxsGBpqMvS11IlJTqdpEUteJqQBmYvQhjoa/FrzTb3qajIKmutSrYApv/rfQs/IGJmaWpb+JsREYfCe/uobmdDh10mkpaaSFB+DoakFB/9Zgqm1bbbpAXT1DTC3dSDyeVhmGa/O87ujsbjoSAxNsu4mAzAwNqXnCD/SUlNIiIvF2MwC/9ULMX/rPBuZmGPzznm2dnDmSmAAAL6+vgwdOlTpfW1t5amqnHp9gevbzMzMSE1NBTJvtJmT95o2bYqPj4/iZxcXF/76668c1/n28ezeNzbO/LKgoaGBgYEBampqirs2vN2OefPm4eGRuaNLLpejp6eXpSxVEdNKgkJaegaXgh5Rq6K74phMJqNWBTfOXn2QbR5dHS3k73xDlsvlr/Iqp+3crDIvIuPYc1x5fltDUxP7gm7cvXZBqYy71y7i5Fb0g23W1NLG2NwSeUYG1wOPUbR8VcV7Lu7FCH8SqpQ+4uljTCytAciQ4HFMMoUt9d/0l8ypopCopA/Wmy6XiElOR00GJWwNuf4scyrudkQiUw/fZ3rAA8UrNCqJC49jmR7wAAnQ0NDEroAbD65fVOrv/esXcXDLfkvpm/5qYWSW2d+bZ4/hXrbqe9OmJCcR+fwJBq8++DU0NXEs5Mbtq8rn+fbVC7i4f/w8m7w6z1cCj1KsfHXFewU8i/PinfP84skjTC1tgMxAYGRkpPTKLjg0btwYAB8fH5YtW8Yvv/xC9erVFesO48aN49atWzRs2PCDbf0Qd3d3ChQoQEBAANeuXePmzZtMmzaNsLCwbOuPi4vDxMQEgNOnT7N+/XpFGevXr+fPP/9k1KhRALlq1+u0//zzD6GhoRw7doxJkyaho6PzyX3La2LkkM8lJiQQGvrmP2bY48cEBwVhbGyMrZ3dB3Jmb+6awyyZ2JkLN0M5/2orq56uNqu2BwKwdFJnnryIYey8zAXH3ceu4/NTLa7ceszZayEUcrRkbL8m7D52TSloyGQyujSvxNqdZ8jIyLp7p3qTtmxY4IdDIQ8cXD04sWsTaSlJlKuV+Z9o/bzJGJlZ0rBT5hRR6J2bxEZGYOviSmxkOAc2rECS5NRs3kFRZrUmbfhz9AAOb1lNicq1eHQ3iDMH/WnV51dFmoC7kXQsY8uj6CRCo5KpWcgULXU1zoRmTnt1KmNLTFI6O4My1w+cTXUw1tEgLCYFYx0NfvCwQCaDw3cyv42npMt5Fpeq1LfUDInE1Ayl45Ubt2HrwqnYFXTH3tWDwN2bSUtJpnTNzOmuLQv8MDKzoG6HXgA8vhNEbFQ4Ns6uxEVGELBpJZIkUbVZe0WZ+1YvxL1sFYwtrImLykyjpqZG8aq1FWm8mrZn7bzJOLl64FTYk6P+G0hNSaJi7cwPxjV/TMLY3JKmP/UFIOT2DWIiI7B3cSUmMoI96/9GkuTUadnxTZlN2jFnZF/2b1pF6aq1eXjnJqcP7KBd3+HZ/5G9h5eXF8uWLWPatGkMGDCAggULcv78edLS0ti4cSNpaWl06NCBP/74I1flvk1TU5Pt27fz888/M27cOLS0tKhUqRIuLi44OztnqX/SpEkUL16cRo0aceDAAU6ePEl6ejrbtm1j0KBBDB8+HCsrKxYsWECdOnVyfPv/OnXqsGjRImbOnEn//v2xt7enXbt2n9yvz0EmvT3uF1TuPbMZ73Xu7Bl6du+S5Xiz5i2ZNGVqjssxLf/mGom+7WowpGtdrM0NuXorjF9+38i565kXwe1bMpiHT17Se9waANTV1fhfjwZ0bFwBOytjIqLi2XXsOuPn+xMT/+bbd51KHuxcOJDizSdyN1T5YrB1KzPnfk/t2cLRVxfB2bm40szbB6dXF2f9NW4wppY2tB3oC8D9G5fZumQWkS+eoqWji0fpijTs1AcjM+X7ZAVdOMXetYuJeBaGqZUN1Zu0pWLdpgAEPIgG3lwEZ6StTlhsCluuPedhVOa1CQOrOhGZmMY/lzLnowuZ69KmpA3mepqkpMsJepGA/83w905DvS4jLCZZcRFcRafMabsze7dyaud64qOjsHEuRMNug3B4ddHf8glDMLG0oWX//wEQcvMKO5fNIerFE7R0dClcqiJ1O/RS6u/GPybxMPgqSXGx6BkZ4+RenDrtvDF7taXU9NWtYo7t3szhbf8QGx2JQwFXfuzxMy6vRmjzxgzEzMqWToMyvw3fvXGJDX/N5OXzJ2jr6FKkTCWadu6H8Tvn+fr5k+xc8xfhTx9jbmWLV7N2VKnXDIAfilq+99wIXy8RHL4yuQ0OeeXt4PClvQ4OX9rr4PClvQ4OX5qpiu4jJoJD/iTWHARBEIQsRHAQBEEQshDBQRAEQchCBAdBEAQhCxEcBEEQhCxEcBAEQRCyEMFBEARByEIEB0EQBCELERwEQRCELERwEARBELIQwUEQBEHIQgQHQRAEIQsRHARBEIQsRHAQBEEQshDBQRAEQchCBAdBEAQhC/GwHwFQ3UOGAExrjVVJvbouHiqpNynonErqLdGqhUrqPeNbUyX1Cv+NGDkIgiAIWYjgIAiCIGQhgoMgCIKQhQgOgiAIQhYiOAiCIAhZiODwBby9ISwlJYWkpCQVtkYQBOHjRHD4AmQyGQALFy6kefPm1KpViylTpqi4VYIgCO8ngsMX8vfff/O///0Pc3Nzzp49y4oVK7h3756qmyUIgpAtERy+gMTERA4ePIiXlxchISEULlyYuXPnMmPGDDZu3Kjq5gmCIGQhgsNn8O5F53p6esTGxrJz506ePHnC3Llzsba25sqVK5w/f560tDQVtVQQBCF7GqpuwLfo9RrD4cOHiYuLo2HDhrRp04YbN24gSRL+/v7cvHmT+/fvs2rVKjQ1NVXcYkEQBGVi5PCZrFy5krp169KyZUu6dOlCwYIF6dGjB9ra2qxbt47Y2FgOHDiAq6urqpsqCIKQhRg5fAaRkZFcv36djh07IpfLWb9+PRkZGYwaNYpevXoRHR2NpaUlZmZmqm6qIAhCtkRwyAOSJCmmkhYsWMDs2bOJiIhgwoQJDB48GBsbGxYsWMCTJ09YuXIl7u7uKm6xIAjCh4ngkAdeB4YjR46wcuVK7O3tuX//PiNHjqR8+fJMnTqVhIQENm/ejJaWlopbKwiC8HEiOOSRNWvWKNYW9u7dy7Fjx+jZsyctWrRgw4YNLFiwgIkTJ2Jtba3qpuaZC+fPseLvZQTdvE54eDiz5y6gdp26n1xen5YVGNKhKtZmBly795yhc3ZxPijsvekHtqlMrxblcbQ25mV0IluP3mDMXwdJSc18OEXwhiE425pmybdoyxmGzN6l+LlH3cIMauSJlbEuNx5F8b9VF7h4/+V76+3bwJ3udQrjYK5HZFwKO849YuKGy6SkyQEw0NFgZKsSNC7niIWRNtceRuG7+gKXHkQq97dNVYZ0ro21uSHX7jxh6PQtnL8R+v7+dqhBr9ZVcbQ24WV0AlsPX2XM/J1v+rtjDM52WacqF204wZDfNyt+bl3Gjk4VHTE30OLOi3hm7r/Lzadx2dapriajW2UnGhW3xtJQm9CXicwPuE/g/ShFmlKOxvxU0REPGwMsDbUZtuk6x+68//wJ+YNYkM4jtra2eHp6cv/+ffbs2YO3tzeLFi0iIiKCbt26kZaW9k0FBoCkpETc3d3xHT3uP5fVunYxpg38gckrAqjccxFX7z5jx8wuWJroZ5u+Xd3iTOpTlynLj1Dqp3n0nbaN1rWLMbH3m+BUrfdfuDT/XfFq9PMKALYcuaFI07KiE791LMPvW69Ta8werodGs2l4LSyMtLOtt1VlZ8a2LcXvW69R6X+78Fl6hhYVnRjTppQizR89KuJVzIa+i05RzXc3R649Y+uI2tia6r7pb71STBvSgslL9lH5p5lcvf2EHfP6YGlqkH1/G5Rh0sAmTFm8j1JtptJ30npa1yvFxAGN3/S3yyxcGoxVvBr1X5jZ30OXFWnqeloyuE4hlp0IoevfF7j7PJ4/2hXHVC/7HXN9a7jQorQtMw/cpf2Sc2y59IRpPxbFzfpNO3U11bnzIp7p++9kW8bn4uXlhUwmIzk5mYCAAGQyGSNGjPhgHplMRqVKlXJVz+XLlxk/fjznz5//L83Nd8TI4RO8vcZw+/Zt0tPTKVasGIsWLWLQoEGMHDkSDQ0N+vXrh6amJhUqVEBXV/cjpeY/1arXpFr1vHnKl0+7Kiz3v8Dq3ZcAGDTDn4aV3ejauAwz1h7Pkr5SMSdOX3/E+oPXAAh9Fs2Gg9coX8RBkSYiOlEpz6+dqnPv8UuOXw5RHOvf0INVAff45/h9AIYuP0u9knZ0qlGIP3bezFJvhcKWnLkTzubTDwF4FJHAltMPKVvIHAAdTXWalnek0+xjnL4VDsC0rddoUNqe7nUKM2XT1cz+dvJi+bbTrPY/m9lfv400rOZJ12YVmbHyUNb+lnTh9JUHrN93MbO/T6PYsO8i5Ys5v9XfBOX+dq3DvUfhHL/w5kr8DhUc2H7lKTuvPQdg6t47VHE1p2kJG1YFPspSb8Ni1qw4Fcqpe5mjni2XnlLBxZSOFRwY7x8MwOn7kZy+H5kl77fi8uXLTJgwARsbG8qVK6fq5nwxYuTwCV4Hhj/++IP69etTqlQp2rRpw927d5k7dy4FCxZkwIABLF26lO7du1O0aFEVt/jrpqmhTmk3Ww6/9SEmSRKHz9+jQlGHbPMEXg+ltJst5TztAXCxNaVBJTf2Bmb/7VVTQ5329Uuw8lXwAdBUV6OkixlHbzx7q144euMZ5V0tsi3n7J1wSrmYUaZgZjBwttSnXkk7Dlx5AoCGugwNdTVS0jKU8iWnplPJzfJNfz0cOHzmtnJ/z96hQglnshN4JYTSno6UK+qU2V97cxpULcLek0Hv72+jsqzccVZxTENNhoeNIWcfvJkSkoBzIVEUtzfKthwtDTVS0uXKfUmXU9LBONv0/9Xy5cvx8PBAR0eHIkWKkJiYiI+PD7a2tpibm+Pt7U18fPxHywkKCqJ+/foYGRlhaWnJli1bFO8lJSXRrl07jI2Nad26NXK5HLlcTsWKFTE0NMTAwAAvLy/u3r1LQEAA3bt3B6Bfv36K//vfAzFy+EQ7duzA19eXLl26oK2tzdq1a5k+fTqrVq1i5syZjBgxgurVq6u6mfmChbEeGhrqvIhU/ub7IioBd2fLbPOsP3gNc2M9Di3ogUwmQ1NDncXbzjJ99bFs0zer7oGJgQ5r3goO5obaaKirER6TrJQ2PDYZN7vsPyw3n36IuaE2u8fURYYMTQ01/j50h9n+maOM+OR0zt4J59cWxbj9JJYXMcm0quxM+cIW3H+e+aFmYaL/qr/K8/wvIuNwd7HKvr/7LmJuos+hpYPe9HfTSaYvP5h9f72KY2Kgyxr/N8HBRE8TDTUZkYnKV+RHJqThbK6XbTmB9yPpWMGBy49ieByVRHkXU2q5W6CWiw/JlJQUUlJSlI5pa2ujra08dRcQEIC3tzclSpRg/vz53Lhxg1GjRjF//nwGDhyIkZERv//+OyYmJsyaNeu99aWlpdGsWTMeP37MmDFj0NfXR03tzffgq1ev0rx5c168eMHmzZs5dOgQderUoUGDBvTq1Yvnz58zbdo0/ve//7Fw4UL69u3LokWL6NGjB3XrfvqaWn4jgkMOyeVypT+wW7duoampibe3NxUqVKBSpUp06tSJS5cu0atXL44fP/5NTiV9LaqXcmFY5xoMnrWTczcfU8jenBmDG/K0a02mrjyaJX3XJmXZd+YuT19mv/CaU1U9rBjStCjDVpzn/L0IClob4vdTWX5tXowZ268D0HfRaeb1qsjNeS1Jz5BzJSSKzacfUsrl069rqV62EMO612Xw1E2cux5KIUcLZvzakqc96jF12YEs6bs2r8i+U8E8jYj95DoBZh24x8hGbqzvXR4JCItKYufVZzQpYZPjMvz8/JgwYYLSsXHjxjF+/HilY7t2ZW4SmDdvHjVq1ACgXLlySJLEvHnzFOkOHsw+IL52+/Zt7t69S/fu3Rk5cmSW9wsXLszEiRNZunQpAQEBhISEEB8fT2BgIJMnT0Yuzxwp3bhxAysrKypWrMiiRYsoV64c7du3z3G/8zsRHHIgJSUFf39/atWqxc6dO6lQoQLp6enExcWxZ88eChUqhK2tLcbGxqSmpgKIwJALETGJpKdnYGWmvPhsZarPs/d8mI/rWYd1+6+wYmfmHPyN+y/Q09VkwbBmTFt1TOn+Vk7WxtQuW5D2o/9VKuNlXArpGXIsjXWUjlsa6fA8Wnk08drI1iXYcPIBq49mToEFPY5BT1uD2d4VmLnjOpIEIS/iaTr5EHra6hjqaPI8JpllA6oSEp45coiITnjVX0Pl/poZ8uxl9h/m4/o2Yt3u86zYfiazv/eeoqerxYJRbZn290Hl/tqYUruCG+2HL1cqIzoxjXS5hNk7i89m+ppExqdmW290UhrDN99AS12Gsa4m4fGpDPAqwJP3nJ/s+Pr6MnToUKVj744aPkRHR4cdO3YopnRykzc7ry8+1dDI/PjLyMhg7dq1HDhwgJ49e9K2bVu8vb1JTs7s4/c0lfQ2sebwEZIkcfjwYfr370+1atXo3r07V69epV+/flSsWJGJEydSv359Ro8ejba2No0bN/54oYKStPQMLt1+Sq2yBRXHZDIZtcoW5OyNx9nm0dXRRC5XvsGhPEN6lVc5bedGZXgRncCe07eVjqdlyLkSEkmNIm92kclkULOoDefuRmRfr5YG71RLxqsDMpQrTkzJ4HlMMsZ6mtQubsuei4/f9Df4MbUquCn3t3xhzl59+P7+vnNDx9ffcLP0t1kFXkTFs+eE8oJ6ulwi+Fkc5V3ebO+VAeWdTbkW9uERRmqGRHh8KupqMmp5WOZqq6q2tjZGRkZKr+w+4F//3/Hx8WHZsmX88ssvVK9eneTkZNauXcvDhw/ZunUr69ev/2B97u7uuLq6sm7dOvz8/Jg7dy7btm3LUVtjYmI4efIkYWFvtlCbmJgAmSOW16Ob74EYOXyETCajQYMGmJqacv/+fUqUKIGZmRkmJib8+++/DB8+nJs3b2Jtbc2hQ4dwcXFRdZO/mMSEBEJD3+zLD3v8mOCgIIyNjbG1s8tVWXPXn2LJyJZcCH7C+aDHDGxTGT1dLVbtzhwZLB31I08iYhn7V+aUwu6Tt/BpV5krd55y9tW00tietdl98pZS0JDJZHRpVJq1ey6TkSHPUu+fe4JZ0Lsylx9EcvH+S/o2cEdPW4N/jmXuXvqzT2WeRiUyacMVAPZdCqN/Qw+uPYxSTCuNbF2CfZfCFB/etYvbIgPuPIuloLUhE9qX5s7TWNa+KhNg7toAlozvyIWbjzh/4yEDO9bM7K9/5shg6YSOPHkRw9gFmR9Gu4/fwKejF1duhXH2+kMKOVowtm9Ddh+7kbW/TSuwdue5bPu77uxjxjbxIOhZHDefxNG+vD06mmrsvJq5KD+uiTvhcan8efQBAEXtDLE00Ob2i3isDLTpWd0ZNWB14Jvfu66mGg5vbdO1M9GhsJU+scnpPI9VXmv4EC8vL5YtW8a0adMYMGAABQsW5Pz580iSxIYNG9i4cSNubm4MHz78g+VoaGiwY8cOfHx88PPzQ1tbm7/++uuDeTp16sSGDRvw9/dHR0eHokWLEheXOWqtXbs2FStWZNu2bVy+fPm7+QIogsN7vL1dNTo6mkKFCqGpqcnVq1dZtmwZVlZWFCxYUPEtJjk5GR0dnQ8V+c25ceM6Pbt3Ufw843c/AJo1b8mkKVNzVdamw9exMNFjbI/aWJsZcPXuM5r/upoXUZmL1I7WxkrfnKeuOookSYzrWQc7SyMiohPYdfIW45cobwOtXa4gTjYmrHwVZN619Uwo5oY6+LYqgZWxDtdDo2gz/QjhsZlTCg7mekr1zth+HQmJka1LYGuqy8vYFPZeDuO3jVcUaYx0NRnTtiR2ZnpEJaTif+4Rv228QnrGm3I2HbiMhakBY/v+gLW5EVdvh9F80F+8iMycenK0MVX60J+67ACSBOP6NcTO0jizv8duMP5P5W+ytSu44WRrxsodZ7Lt78GgcEz0NOld3QVzfS1uv4jn5w3XFIvU1kY6SiMjLXU1+tZ0wc5El6TUDE7de8l4/2DiU97sxvK0NWRhp1KKn4fUzbyZ5M6rz5i061a27Xgfb29vvL29lY7NmTOHOXPmZEkbEBCg+LeXl5fS1JqnpycHDmRdi3k7Tbdu3ejWrZvi5yNHjmTbJkNDQwIDA3PYg2+HTHr34QOCklmzZnHx4kXmzJlDQkICM2fOZP78+VSpUoUnT55gb2/PsWOZO2Ty89xkcrrq6jatNVYl9eq6eKik3qSgcyqpt0SrFiqp94xv3lwLI3xZYs3hA1auXMmIESNQU1Pj+fPnODs7M2bMGHx8fDh9+jQZGRnMmTMHmUyWrwODIAjCu8S00nvExcWxefNmPD09mTx5Mo6OjgAYGBgwZ84c2rZti5OTEw4O2V+kJQiCkJ+JkcN7aGtr8/LlSyIiIjA0zNxyuGfPHgYMGEBiYiJVqlQRgUEQhG+WCA4oL1I9evSIu3fvIpPJaNWqFREREXTo0AFfX1/mzJnD9u3biYqK+kBpgiAI+Z8IDrxZSJ41axaNGzemXLlydOjQASsrK/r160dgYCBz584lLCyMgIAA7O3tVdxiQRCEz0sEh1e2bNnCpEmT+OGHH4iNjeXu3buUL1+e/v37c/78eY4ePUpAQADFixdXdVMFQRA+O7Eg/cqdO3ewtbXl0KFDFCxYkClTprBixQoMDAwYNWqUqpsnCILwRX2XI4fsLu1ITk4mODiYJ0+eMHXqVAoUKMDJkyc5f/48CQkJ2ZQiCILw7fouRw6v1xjWrFlDSEgIlStXpm/fvpw6dYoDBw7g5+eHpqYm9+/f5+jRo+jrZ/80MkEQhG/VdxUc3r4lxqFDh+jSpQsymQxTU1NGjBjBnDlzmDhxIlevXsXGxoa///4bT09PFbdaEAThy/uugsPrwPDnn39y4MABunXrhpeXFyNGjGDixIno6+uzbt064Pu8V5IgCMJr392aw7Vr1/Dx8WH79u1YWVnRpUsXNm7ciLGxMQMGDFA8VEQEBkEQvmfffHB4d/G5ePHiLFiwADMzM2bMmMHevXupWrUqa9aswc3NjXr16qmopYIgCF+Pb35a6fVU0ooVK7h69SrVq1enT58+GBgY0LNnT1q2bMm///5L8+bNuXTpkniCmyAIAt9wcHh78fno0aP06dMHgIULFzJq1ChGjx6NmpoanTp1Ujw4XE8v+4esC4IgfG++2eDwOjDMnz8ff39/OnToQPPmzRk5ciRjx45FTU2NkSNHoqGhQfHixb/77ao6KvxLiDoyUSX19lh3WSX1thlWWyX1aqp987PIQh76ZoODJEncvn0bX19fEhIS6Nu3Ly1btsTd3Z02bdowevRoNDU1GTZsmKqbKgiC8NX5pr5KvL34LJPJcHd3Z8WKFdjZ2bFo0SLWrFlDkSJFWLduHWXLlv1ungUrCIKQW9/UyOH1VNKOHTu4fPkyurq69OnTh9WrV9O6dWu6dOlCcnIyPXv25OTJk2hpaam4xYIgCF+nb+4Z0vPnz2fEiBHo6+sTHh5O4cKFOXXqFBcvXuSHH35AR0eHp0+fYmhoKB7t+ZVQ1fOrVbbmUNJGJfWqas2hcTErldQr/Df5flrp7dh269Ytfv/9d9q3b8+RI0cYOnQod+7coU+fPtSrV4/169dz6dIljIyMvqvAkJaWpvRzRkaGiloiCEJ+ka+Dw9vbVYOCgjAxMSExMREdHR2KFCnCjBkzKFu2LA8fPgSgdevWuLm5qbLJX9yTJ084fPgwAKtXryYhIQF1dXUVt0oQhK9dvl5zeB0YZs6cyYgRI/j555+xs7Pjzz//pFixYhQrVgwNDQ1F0NDV1f2uRgypqakcPXqUX3/9FRcXF06fPo0kSXTp0kUpsAqCILwrXwcHgCNHjjBmzBh8fX3x9vamdu3adO/enf79+2NmZoa2tjbLli37Li9w09LSombNmgCcP3+emjVrUrRoUeBNYBVBQhCE7OT74HDjxg2Sk5OpWLEiLi4uuLi4MHjwYGxsbEhISKBx48YUKFBA1c38ot7+wJckCTs7OzQ0NDh+/DgbN27E0tISfX19jIyM0NTUVHFrBUH4GuX74GBpaQnApUuXqFSpEo8fP2bz5s3079+fgQMHqrh1X55cLkft1a6U0NBQ1NTU2LJlC1FRUYwePZqZM2dy//59wsLCcHJy4p9//hEjB0EQssj3waFevXrUqVOHsWPHsmzZMmQyGTExMVSsWFHVTfviJElSBIY5c+bwxx9/oKOjQ506dZg/fz5z585l6NChbNq0CQsLC/78808RGARByFa+3q0EYGZmxooVK/D19cXJyYnq1atz8uRJxdz69+TtBfqJEyfSuHFjHjx4wJ9//kmXLl1wdnZm8+bN7Ny5k7Nnz1KyZEkVt1gQhK/VN3URXGpqKpIkoa2treqmqIRcLufp06c0atSIVq1acfr0aYKDgzE3N+fixYs0adKEbdu2KUYXXwtxEdyXIS6CE3Lj6/qU+I+0tLS+y8Agl8sBUFNTw97enn/++YeIiAgOHjzIpEmT6NWrF/r6+ly4cIEnT56ouLWCIOQH31Rw+N48ePAAyAwKK1asoGfPnuzevRtPT0/q1q1LRkYGq1evxt/fn65du3L69GkcHBxU3GpBEPKDfL8g/b06cOAADRo0YOrUqZQpU4bBgwcTFxfHunXrGDt2LGXKlKFly5bs2rULPT09Zs+ejZOTk6qbnScunD/Hir+XEXTzOuHh4cyeu4Dadep+cnn13CxoXNQKY10NQqOSWHk2jPsvE7NNqy6DZsWsqV7IDFM9TZ7GpPDvpSdcfRKnSFPHzZy6bhZY6mfe2PFxTDJbrz7jyltpAE7t3cqxHf8SFx2JrXMhmnsPxrGwZ7b1ZqSnc2TrGi4c3UdsZASWdo407NQH99LKGy9iXoazZ+1f3Lp0htSUZCxs7GkzYAQOhTwUaU7s2cKR7euIi47EzqUQLXv8jHPhIu+t9+CW1ZwP2EvMq3qbdO6H5zv1Rr8MZ+eahQRfPENqajIWNg50GOCLo6tHtuUKXz8xcsinDA0NAfjtt9+YPn06Xbt2ZeXKlZQsWZIpU6Zw/fp1Ro0axblz57hy5QqFCxdWcYvzTlJSIu7u7viOHvefy6rkbEKncnZsufqM0btuERqVxIg6BTF6z9OP2pSypbabOSvPPmb4jmAO3YlgSM0COJu+ebxsZGIa/158wqjdtxi9+zY3nsUx1KsA9sY6ijRXTh5m58oF1GnTFZ9pS7B1LsSyyb8SHxOVbb37/l3KmQP+NPcezNDZK6lYrxmrpo8m7MFtRZrE+DgWjhmImro63iN/55fZq2jcdQC6+oaKNJdOHmL7ivk0aNuNodOXYufsyuJJvxD3nnp3r1vC6QM7aNnjZ/43ZzVV6jdn+e8jeXxfud55o/qjrq5Br9HT+d+c1TTvOgBdA8Nsy/yQkJAQZDIZ7du3z3Xe15YsWYKlpSUymYylS5d+cjnfOxEc8hm5XI5cLqdSpUqcOXOGtLQ0Dhw4QGxsLJ07d2bu3LkULVqUX375haNHj1K8eHEcHR1V3ew8Va16TQYOHkKduvX+c1kNi1hy5M5Ljt2LJCwmhb8DH5OSIadmIbPs6y5oxo5rL7jyJI7w+FQO3X7J5bBYGhWxVKS59DiWK0/ieB6XyrO4FDZefkZyuhxXyzdX6R/fuYEKdZpQvlYjrB1daNn7FzS1dDh3eHe29V48tp9aP/6ER5lKmFvbUblBCzzKVOK4/wZFmqPb/sHY3JK2A3xxLOyJmbUtbiXLY25j/yaN/3oq1W1KhdqNsXEsQOs+v6KprcPZQ7uyrffC0X3U/bEzRcpWxtzGjqo/tMSzdGUC/P9VpDm8dS0mFlZ0GDgS58JFMLe2w71UBSzeqvdLmjVrFgkJCaxcuZLatVXz1L1vgQgO+cjrC9zU1NS4cuUK5cuX58yZM+jo6LBq1Sr++OMPypUrx8yZM6lTpw4NGzZUdZO/aupqMgqY6XH9WbzimARcfxpPYcvsHxuroS4j9dUGgNdSM+S4Wxlkm14mg0ouJmhrqHE3PAGA9LQ0wu7fpnCJsop0ampquJYoS+jtG9mWk5GWhoam8vNHNLW0CQm+pvj55vmTOBTyYM3MsUzs0Zw/hvXgzEF/xfvpaWk8vncbt3fqdStRjpD31JueXb3aWjwIelPvjfMncCzkzsoZYxjbvSkzf/Xm9IEd2ZaXG6GhoTRp0gRjY2McHR0ZN26c4i7MU6dOxczMjNKlS9OuXTtkMhkBAQF4eXkRHBxMUlISXbt2JTQ0lKCgIOrXr4+RkRGWlpZs2bLlP7fteyDWHPKR11tQp0+fzrx58xg5ciR9+/bl8OHD1KpVi19//ZXExER8fX3x9/dHR0fnIyV+3wy11VFXkxGTpHxL89jkNOyMs9/1du1JHI08LQl+Hs+LuFSK2hpQ3skEtXeuJXQ00WH8D4XRVFcjOV3O7IAHhMWkAJAYF4NcnoGBsalye4xNCQ8LzbZet5LlOb5zAwWLlMTM2o671y5w/cwxxU41gMgXTwncv53qTdpQ68efeHw3mB1/z0VDQ5OyXj+Q8KpeQxPlUZGhsSkvwh5mW697qQoc9V9PoSIlMbex5861C1wLVK735fOnnNq3nZpN21Lnx848uhvM1r//QENDk/K1GpKSkkJKSopSudra2h/dWdixY0fOnj3LlClTOHPmDBMnTsTJyYmyZcvi6+tL6dKl6dmzJyNHjlTkGTt2LO3atSMtLY1Fixbh6elJtWrVePz4MWPGjEFfX/+r28r9tRJnKR94+1KUffv2MXHiRNq3b0+dOnUAqFSpEqdOnSIjI4Np06YRGRkpAsNnsurcY57FpTKjmScrO5Wka3kHjt17ybtXCz2JTWHkrluM3XObQ7cj6FvVGfv3BJycaNrdBwtbB2YM7syoDnXZvuwPytVqqHSFuySXY1egMD907I19ATcq1mtGhbpNCNy//ZPrbentg6WtA1MH/8TwdrXZsnQ2FWo3QvZWNJQkOQ4F3WjcqQ8OBd2oXL8Zleo25dSrev38/DA2NlZ6+fn5fbDetLQ0Tp48qfjSM3fuXAD27NnD0aNHARg+fDj9+/enXbt2iny1a9dGX18fHR0d2rdvT0REBHfv3qVDhw6MHDmSwYMH06JFi08+H98TMXL4ygUEBHDr1i26deuGtrY29+7dIyUlBS8vL8Ui89GjR6latSqXL19GS0sLM7Ps58sFZXEpGWTIJYx1lW8+aKSjSUxS9lfmxaVkMDvgAZpqMgy0NYhKSqN9aVtexCt/M86QSzyPSwUgJDKJguZ6NPCw5O8zj9EzNEZNTT3L4nNcTFSWb/WvGRib0HX4ZNJSU0iMi8XIzII9a//CzNpOkcbQ1BxrBxelfFb2zlwPPAaA/qt646Ijs6nX/D31muI9wo+01BQS4mIxNrNg55pFmFu9qdfIxBxrB2elfNb2zlwNzPwQ9/X1ZejQoUrv5/R6pNfBL7vbvIgRwOclgsNXLCIigvbt25OYmIiamhrdu3dHQ0OD9PR0Dh8+TIkSJYiIiGDGjBkEBwfTp08fVTc5X8mQSzyITKSojQEXHsUAIAOK2Riw/1bEB/OmySWiktJQl0F5JxPOPIz+YHqZDDTVMz/MNDQ1sS/oxt1rFyhaoTqQuZ5099pFqvzQ8oPlaGppY2xuSUZ6OtcDj1GiipfiPRf3YoQ/UZ6Winj6GBNLa0W9DoXcuHPtAsUr1lDUe+fqBao1/PGj9Zq8qvdq4FFKVan1pl6P4rx48kgpffjTR5hZZl4JnpMppCz1aWpSrVo1jhw5wsyZMzlz5gwADRs2pEyZMkDm9GpERATr169/bznu7u64urqybt06ChcujL6+Pk5OTmL0kAMi9H6lnj17hoWFBX/99RdmZmaMGzeONWvW0KZNG5o1a8asWbOoUaMGP/30E0ePHlU8t+F7kJiQQHBQEMFBQQCEPX5McFAQTz/h6u89N8OpVdic6gVNsTPSpntFB7Q11Dh6L/Pbdd8qTrQrbatIX8hCj3KOxlgaaOFupc/wOoVQk8HOGy8UadqVtsXDSh8LfS0cTXRoV9oWT2sDTj548429epO2nD20iwsBe3n+OIStS2aRlpJEuVqZmwjWz5vMnrWLFelD79zk+pljvHz+hAdBV1g2eRiSJKdm8w6KNNWatCH0zk0Ob1lNxNPHXDp+gDMH/ZUCTs2m7Qg8uJNzR/bw/HEImxbPJDUliQq1GwHwz9zf2LlmkSL9w9s3uBp4lJfPnnD/5hUW//YLklxO7RYd3yqzLQ9v3+Dg5lWEP33MheMHCDzgT9WPBLqPWbNmDfXq1WPixImcOnWKsWPH4u3tTenSpfHz8+P+/fusWLGC6tUzA6yJiUmWMjQ0NNixYwfVqlXDz8+PSZMmKa2XCO8nRg5foSNHjlCnTh1WrFhBly5d0NDQoHfv3vj6+pKWlsasWbNwcnLi7NmzODk5sWnTJjw8vp+LjW7cuE7P7l0UP8/4PXP+ulnzlkyaMjVXZQU+jMZQR4PWJW0x1tXgYVQS0w7fJ/bVDZ/M9bV4ezlBU01G21K2WBpqkZIm53JYLAtPPiQx7c1zuY10NOhb1RkTXQ0S0zJ4FJXMtEP3uP70za6oklVrkxAbzf71f7+6GM0V71HTFdNK0REvkMnefHdLT01l37qlRL54ipaOLh6lK9J+0CilaxgcXT3pMuw39q5dzKFNqzC1sqFpt4GUrv5my2/pqnWIj4lm77/LiI2OxL6AK71Hz1DUGxXxXGkKJy0tlT3rlvDy+VO0dXTxLFOJjj5jlOp1cvWk+/DJ7Fq7mP0bV2JmZUvz7oMoW6N+rn4XAC4uLkprbDt37sw2nbGxMevWrSMqKkrxBMjX/wdCQkKU0np6enLgwIFct+V7903deO9b8c8//9CjRw9SUlL4559/aN++Pbt27aJXr17IZDL8/Pzo0iXzwzEtLS3fP7BH3Hjvy/iWbrzXoUMH/P39UVdXp0yZMornxQt5R4wcviKvn+DWsWNH1NTUGDx4MB07Zg7f27dvz5IlS+jfvz89evRAkiS6du2Khob4FQrfn3Xr1qm6Cd88sebwlZDL5UrD+fbt27N48WJMTU3p2LEj//77L40bN2bOnDkULlyYKlWqANnv4hAEQfivxNfOr8DbT3DbsWMHt2/fxtzcnDZt2rB27Vq6dOlCx44dSU5Oplu3btSvXx99/eyv4BUEQcgLIjh8BV5/+589ezaTJ0/GyMiIkJAQAgMD+euvv1i+fDlNmzZlwIAB/PjjjxgZGam4xYIgfOvEtJIKvb2l7vbt2/z5558MGDAAU1NTihYtSvPmzVmyZAkNGzZk+/btXLp0SQQGQRC+CDFyUJHXN9EDCAoKwtTUlKioKObMmYONjQ1z585V3FnS3d2dpk2bqrjFgiB8T8TIQUVeB4aZM2dSokQJZs6ciaOjIwkJCTRu3Bi5XM7KlSt5+fIlrq6uKm6tIAjfG3GdgwodOXKExo0b8+uvv+Lt7c21a9fo06cPz549w9jYGH19fXbv3k2JEiVU3dTPSlzn8GV8S9c5CJ+fmFZSoRs3bpCcnEzFihVxcXHBxcWF69evY2NjQ0pKCj/88AMuLi6qbqYgCN8hMa2kQpaWmU8Pu3TpEi9fvuTKlSts3rwZSZLo27evCAyCIKiMGDmoUL169ahTpw5jx45l2bJlyGQyYmJiqFix4sczC4IgfEZi5KBCZmZmrFixAl9fX5ycnKhevTonT56kaNGiqm6aIAjfOTFyUDF7e3smT55MamoqkiTl+r73giAIn4MIDl8JLS2tjycSBEH4QsS0kiAIgpCFuM5B+G6p6vqKH+afUkm9bcrbfTzRZzCgqotK6hX+GzFyEARBELIQwUEQBEHIQgQHQRAEIQsRHARBEIQsRHAQBEEQshDBQfgsJEni9Ua458+fExcXp+IWCYKQGyI4CJ+FJEnIZDK2b99O48aN2bx5M/Hx8apuliAIOSSCg5CnoqKiiIuLQ01NjfPnz9O9e3fc3NwoXrw4BgYGqm6eIAg5JIKDkGeuXr1KzZo1uXz5MgBXrlwhOTmZLl26ULZsWQCuX7/OixcvVNhKQRByQgQHIU/ExsbSo0cPNDQ0ePjwIcePH+fZs2ckJydz6dIl0tPTefDgAQsWLODw4cPI5XJVN1kQhA8QwUHIEzExMWRkZHDlyhW6dOnCnTt36N69Ox4eHkyePJn69evTr18//vrrL2xsbBTP0BYE4esk/ocKecLR0ZHq1asjSRIaGhrExcVhZ2fHunXrKFWqFDdv3iQiIoKdO3fi5eWl6uYKgvAR4pbdwn/yeldSTEwMaWlpNG7cmD179jB8+HDs7e1p3bo1J06cIDw8HA0NDUxNTRVbXGUymYpbLwjC+4iRg/CfyGQyDh8+TJMmTejYsSP+/v4sX74cuVxO9+7dWbVqFZD5vGxTU1NFHhEYBOHrJoKD8Elef/s/c+YMS5Ys4eTJk4wcOZKLFy/SuXNnli1bRkJCAt27d+fZs2eIO8MLQv4ippWETyKTyTh+/DgtWrSgRYsWVK5cmbNnz+Lt7c3SpUvp0qULaWlp2NnZYWNjo+rm5okL58+x4u9lBN28Tnh4OLPnLqB2nbqfXF6LEja0L2eHmZ4WdyMSmHvkAcHPs79QUF1NRqfy9jTwtMLSQIvQqCQWn3jI2YfRijQdy9tTo5A5Tma6pKTLufE0lr9OPORRVHKW8q4c2sHFvZtIjInEwrEgNTv1x6agR7Z1Z6Snc373vwSdPEhCVASmNg5UadMDl+LlFWkCt63m7I41SvlMbRzoPGXZJ5wZ4WsgRg5CrkmSRHp6Ojt27CAmJoaffvqJQ4cO8b///Y+rV68yaNAg7t69S48ePWjYsKHSrTTys6SkRNzd3fEdPe4/l1XLzZz+NVxYEfiYXv9c4V54AtNbFsFEVzPb9D2qONG0uDVzA+7TddUldlx7xqSm7rha6ivSlLI3YtvVp/T/9yq/brmBupoa01sWRUdD+b/57bMBHF+/mIrNOtF+3AIsHAuyfdYoEmOjs607cOsKrgfsxqtTf376bQnFajVm1/yJvHh4Vymdmb0zPWavU7xa+876bycphwICAhg/fjx37979eOJ3yGQyKlWq9Blalf+J4CDkmkwmQ0NDA1dXV+RyOefPn0dHR4dq1arh7OzMpUuXGDx4sFL6b2GNoVr1mgwcPIQ6dev957LalLFj1/Xn7L35goeRScw6dJ/k9AwaFbXKNn19D0vWng3jTEg0T2NT2HH1OYEPomlX5s3T3YZvC2LvzXBCIpO4F5HI1P13sDHSxs1a+cr0S/u2UKzGDxSp3gBze2dqd/FBQ0ubm8f3ZVt38KlDlGvcHpcSFTC2sqVEraa4lCjPpX2bldKpqamjb2ymeOkaGv/Hs5QzAQEBTJgw4ZOCw7p16/jtt98+Q6vyPxEchI96+5v/tWvX+Oeffzhw4ACenp40a9aMsWPH0qhRI/78808KFy5MyZIlefjwobiX0ntoqMlwtzLgwqMYxTEJuBAaQxFbw2zzaKrLSM1QvnAwNV1Ocfvs0wMYaGXOGse99TzUjPQ0Xjy8g2ORMopjMjU1HIuU5um9m9mWk5GehoamlnIfNLV5cueG0rHo52EsG9KBFcO7sm/xVOJe5v5K+JCQEGQyGV5eXtSuXRsjIyN8fHx49uwZJUuWRF9fH2NjY5o2bUp4eDgrVqxgwoQJADRs2BAXFxdFGbVq1aJatWqYmJiwbNky+vTpg76+Pq1atVJchNmhQwdGjx6d63Z+D0RwED7q9Tf/Xbt2Ubt2bXr37k2DBg2YOXMm9evXp1OnTgQGBnL27Fm6d+9OgQIFcHZ2RkNDLGllx1hXA3U1GZGJqUrHoxLTMNPPflrp3MNo2pSxw95EBxlQ1smY6q5mmOlpZZteBgys6cK1sFgevExUHE+Ki0WSy9EzMlFKr2dkSmJMVLZlORUry6X9m4l+HoYklxN64wL3Lp4kISZSkcamoAf1evxK86GTqdVlEDHhz9g09RdSkxJJSUkhNjZW6ZWSkvLBc3Tq1CkaNWqEi4sL8+bN49GjR7Rt25Y//vgDHx8fdu/ezdSpU6lZsyatWrUCYMSIEcybN09RxpkzZ2jbti0ymYxevXqRlpZGjRo12LJlC4cOHfpg/YIIDsIH3Lt3T7EV9fnz5/Tr14969erxzz//KLat3r9/n7lz5xIYGMiUKVO4cOECu3fvZtq0aejo6Ki4B9+OeUcfEBadxKoupTnoU5nBXgXZc/MFEtmv5fxcuyAFLPSYuOf2f667Rod+mFjbs3pkT+b3bkzAmj/xrFpfaarQpUR5CpevgYVjQZyLlaP5kN9ISYznzrlj+Pn5YWxsrPTy8/P7YJ1eXl78+uuvNGvWDIDg4GC2b99O7969+e2335DL5dy4cYMCBQpQrFgxAGrWrEnTpk0VZdSsWRMfHx+KFi2KJEn4+fnRpk0bIHOEInyY+GonZCsxMZHatWvz5MkT5HI5Xl5eJCUlYWVlRbNmzWjWrBkXL17k9OnT6Ojo4ObmRnR0NDt27ODEiROK/7BCVjFJ6WTIpSzf+k31NIlMSHtvntH+t9BSl2Gko0lEQiq9qznzJCbrN/DBXgWoXMAUn43XCY9XHp3oGhohU1PLsvicGBuFnrFptnXrGZnQZNB40tNSSY6PRd/EnFOblmFs+f5daNp6BphYOxD94gmzp45i6NChyu9ra783L4CZmRmAYvQ5f/58zp07x6hRo6hZsyZNmjQhOTlzF9b71rOMjY2VyjA2NlbctiUjI+OD9Qti5CC8h56eHnPmzMHQ0JARI0awYMECMjIy2LNnDxs3buTevXuYmJhgYGBAbGwskiRRoUIF1qxZQ4kSJVTd/K9aulzi1ot4yji+WbCVAWUdjbn59MMPRUrNkIhISEVdTUZNVzNO3otUen+wVwGquZoxZPMNnsVmDRzqGppYORfmUdAlxTFJLudR0GVsCxX5YN0amloYmFogz8jg7oUTFCxd+f3tTE4iJvwJ+sZmaGtrY2RkpPT6WHB4n6ioKLZv305q6pugZ2JiAsD27dsJCAj4pHKFrMTIQXivli1boqWlRdu2bdm/fz9ly5bl0KFD/PTTTzg5OfHkyRM2b96MiYmJYsFaV1dXxa3+fBITEggNDVX8HPb4McFBQRgbG2NrZ/eBnFltvPgE3/qFufU8nqBn8bQuY4uOpjp7bmYu4vrWdyUiIZUlJzPr87QxwEJfi7vhCVgYaNGtkiMymYx/L4Qpyvy5VkHqelgwakcwSakZmOllrl/Ep2QoLWaXbvAjB5bOwNrFDesC7lw+sJX0lGSKVKsPwP4lv6NvakHV1t4APLsXTHx0BJaOhYiPjuDM9jVIcomyDdsqyjy+fjEFSlXCyNyKhOiXBG5bjUymjltFr1ydl/cZOHAgs2fPZsWKFQwcOFAxKoDMv9PFixfz119/cefOHZYuXZondX7vZNK3sAFd+Kz8/f1p3749bm5uuLq6YmJigqmpKT/88AO1a9dW3F8pv3lrE0+OnDt7hp7du2Q53qx5SyZNmZrjcn6YfwqAliVtaFfWHjM9Te5GJDAv4AFBzzJ3eM1pXZRnsSlM3Z+5PbOkvRFDahfEzliHpLQMAh9EsfjkQ16+NQ0V8HOVbOubuv8Oe2+G06b8mwB25dB2Lu7dREJMFJaOBanZsT82hTIvgts8bRhGFtbU6/ErAI9vXeXIqnnEhj9FU0cXl+LlqdK6Bwam5ory9iyawpNb10hKiEPX0Bi7wkWp/GM3TKzsGFDVJcfnRvh6iOAg5Ii/vz+dO3dGV1eXcePG0bdvX4B8fRO93AaHvPI6OHxpbweHL0kEh/xJrDkIOdK0aVNWrFhBVFQUhQoVUhz/Vi5wEwRBmVhzEHKsRYsWPHr0CEtLy3w7lSQIQs6IkYOQK5aWlqpugiAIX4AIDsInEaMGQfi2ieAgCIIgZCGCgyAIgpCFCA6CIAhCFiI4CIIgCFmI4CAIgiBkIYKDIAiCkIUIDoIgCEIWIjgIgiAIWYjgIAiCIGQhgoMgCIKQhQgOgiAIQhbieQ6C8IWp6jkS26+FfTzRZ9CutL1K6hX+GzFyEARBELIQwUEQBEHIQgQHQRAEIQsRHARBEIQsRHAQvkkXLlwgMTFR1c0QhHxLBAfhm3PgwAHKly/P4cOHSU9X0dYgQcjnRHAQvjkVK1Zk4cKFHDp0iG7duokRhCB8Ag1VN0AQ8ookSWRkZGBkZETXrl0pUqQIISEh6OnpMWfOHPT09FTdREHIN8TIQfgmSJKETCZDQ0OD2NhYdHR0uHLlCqVLl2bp0qX4+PiIEYQg5IIIDsI3QSaTAbBo0SIaNGjA2LFjSU9P59ixY5QsWZK///4bb29vkpKSVNxSQcgfxO0zhHxNLpejppb5HefSpUu0aNECuVxOWFgYLVq0YMGCBRgbG1OkSBFCQ0MJCwvD1tZWpW0Wt88Q8gOx5iDkW5IkKQLDwYMHuXr1KqVLl2bgwIH4+/szf/581NXVmTNnDjdv3uTZs2cqDwyCkF+I4CDkS2+PGNavX0+HDh0A0NbWZvbs2UyZMgW5XM6CBQvQ0dFhxYoVFCxYUJVNFoR8RQQHIV96HRhWrVrF8uXL8fX1JT4+nnnz5lGtWjWOHTvGlClT0NHRwdvbG3V1dRW3WBDyFxEchHwlOTkZHR0dACIiIujZsyfp6ekUL16cuXPnYmNjw6hRoyhatChBQUH8/vvvisVqQRByTgQHId94+vQpW7ZsoVGjRqxfvx5DQ0OOHDlC69atWbZsGWXKlMHX15fExERmzpxJenr6NxEYLpw/x4q/lxF08zrh4eHMnruA2nXq/qcyz+zbxkn/9cTHRGLtVIjG3Qfh4OqZbdqM9HSObf+Hy0f3ERcVgbmtI/U79qZwqQqKNIc3riBg8yqlfBZ2jvjMWvmf2imojtjKKuQLcrmcBw8esGDBAsqVK8fIkSOJiYmhatWqrFq1Ci0tLX799VcWLVrEpEmTePToEYULF1Z1s/NEUlIi7u7u+I4elyflXTt1hL2rF+LVugt9/f7CxrkQq/z+R3xMVLbpD63/m/MH/WncfRADZyynfN2mrJs5lqcP7iils3JwYdiiTYpXj/Fz86S97+Pl5YVMJiM5OZmAgABkMhkjRoz4aL5u3bohk8kIDg7+rO3Ljo2NDS4uLgCsWLECmUzGokWLvng7ckIEByFfUFNTo0qVKjg7OxMVFYWdnR0lSpQAoF69emzZsoXIyEh+++03YmJiMDc3V3GL80616jUZOHgIderWy5PyTu3aSNnajSjj1RArBxea9hyCppY2FwP2ZJv+yokD1GjRCbfSlTCztqNC/ea4la7IyV0bldKpqatjaGKmeOkbGedJewXVEMFB+KrJ5XLFvzMyMihVqhRNmzYlJiaG0aNHc+LECYKDgzE0NOTChQscPnwYY2PxofQ+6elpPH1wm0LFyyqOqampUah4WR7fvpl9nrQ0NDS1lI5paGkTGnxN6djLZ2FM79eG2T6d2DRvMtERz3PdvuXLl+Ph4YGOjg5FihQhMTERHx8fbG1tMTc3x9vbm/j4+I+W83pUMWLECCwtLalYsSIhISFKaZYtW4a9vT2urq5cvXoVgNDQUJo0aYKxsTGOjo6MGzeO15eCyWQyihcvzo8//oixsTGtW7dW/H1u2bKFYsWKoa+vT8WKFTl37hwAkZGRNGjQAH19ffr37092l5WdO3cOV1dXrK2tWbny65mGE8FB+Gq9vV315MmTXLx4kZYtW7Ju3TomT57MgwcP+Omnn2jTpg3e3t4ULFgQNzc3Fbf665YYG4NcLkff2FTpuL6xKXHRkdnmcS1RjlO7N/Ly6WPkcjl3r54n6OxxpfQOrp607DecziOm0qTHz0S9eMqy8YNJScr5LUsCAgLw9vZGW1ub+fPn06BBA0aNGsX8+fNp06YN/fr1Y82aNYwdOzbHZYaGhjJo0CDOnj3Lzz//rPTehQsX8Pb25t69e0ybNg2Ajh07sn//fsaMGUOlSpWYOHEif//9tyLP9evXKVasGGXKlGHz5s0cOnSI4OBg2rVrh6mpKaNHjyYxMZHmzZuTnJzMxIkT2b9/Pz169EBDQ4MXL15kaeORI0cYPnw4enp69OnTh2fPnuW4f5+TWJAWvlqvA8P06dPx8/MjOTmZokWLMmXKFHx8fIDMb3/h4eHs27dPjBg+k0bdBrJ98UzmDu2GTAam1naU9vqBi0feTEO5la6o+LeNcyEcXD2ZNbAD108HkFLWiZSUFKUytbW10dbWVjq2a9cuAObNm0eNGjUAKFeuHJIkMW/ePEW6gwcP5rjts2fPxtramnXr1nH06FGl93777TcqVKjAb7/9RkhICHFxcZw8eZL69evz66+/8vTpUzZt2sSePXvo0aMHAIULF2bixIksXbqUgIAAQkJCuHXrFunp6Zw4cYITJ04oyr958yZHjx5V3PgRMv9e39WnTx969+5NVFQUI0aM4MyZMzRv3jzHffxcRHAQvmp//PEHU6ZMYfTo0UyfPp1Lly7Rq1cvlixZgo+PD926dSMlJQVLS0tVNzVf0DMyRk1NjYR3Fp8TYqIwNDHLNo++kQkdf51EWmoqSfExGJpacOCfJZhav/9qc119A8xtHYh8Hoafnx8TJkxQen/cuHGMHz8+R23W0dFhx44dip1n7waVD3k9jZPddI6ZmRkaGpkfgRkZGYrjr+vJbqebmVnmOcou3+jRo6lZsyaQOeotUKDAf26jKongIHxV3p5KSk9PJzExkR49enD48GH09PT49ddf+f3332nfvj3Lly+nWbNmKm5x/qKhoYltATfuX7+IZ/lqQOY5v3/9IhUatPhgXk0tLTTNLMlIT+fm2WMUreT13rQpyUlEPX+CYfV6+P7Sm6FDhyq9n90HfOPGjZkxYwY+Pj4MGjSImzdvUr16dS5cuMDatWupXr06Fy9eRCaTUb169Rz1d+jQoXh6enLr1i1atPhw/wwNDalWrRpHjhxh5syZnDlzBoCGDRt+MF/dunXR0NBg06ZNODs78+LFC1atWkVwcDA1a9bk8uXLiimt7O4MvHjxYszMzFi8eDHa2tpUrFgxSxpVEMFB+Gq8fa+kDRs2cOnSJRo2bEhoaChLliyhX79+lCtXjjJlyvD06VOKFi2q4hZ/GYkJCYSGhip+Dnv8mOCgIIyNjbG1s8t1eVUat2HrwqnYFXTHwdWD07s3k5qSTJmaPwCweYEfRmYW1OvQC4BHd4KIiwrHxtmV2MgIjmxaiSRJVGvWXlHm3tULcS9bBRMLa+KiIji8aSUyNTWKV62d7RRSdry8vFi2bBnTpk1jwIABFCxYkPPnzyNJEhs2bGDjxo24ubkxfPjwHPe1QIEC/PHHH5QvX57Zs2d/NP2aNWsYMGAAEydOxNDQkLFjx+Lt7f3BPB4eHmzYsIGxY8cyaNAgLCwsqFs38zqUsWPHcvPmTVatWkXr1q2zHeHWqlWLOXPmkJCQwF9//YWNjU2O+/c5ibuyCl+Ft0cMmzZt4qeffiI1NRVjY2MGDx7Mn3/+iaamJlZWVtSoUYMRI0bk25vo5faurOfOnqFn9y5Zjjdr3pJJU6bmuJy378p6Zu9WTuxcT3x0FDbOhWjUbRCOhTMvgvt7whBMLG34sf//AHhw8wo7l80h6sUTtHR0KVyqIvU69MLIzEJR3oY/JvEw+CqJcbHoGxnj5F6cuu28MbOxV8ldWb28vDh69ChJSUmKK+qF3BHBQfiqrF27lmXLllG+fHkcHBz466+/uHXrFgULFkRXV5e4uDi2b99OsWLFVN3UTyZu2f35ieDw34lpJeGr8fDhQzp37gxA0aJFGTRoEC4uLkyePJmzZ8+yatUqGjdujKmp6UdKEr53AQEBqm5CvieucxC+Gs7OzuzcuRMzMzNWrVrFqlWraNq0KcOGDaNs2bJUqFBBBAZB+ELEtJLw1dm9ezft2rXD0NCQ8ePH07t3bxISEtDX11d10/KEmFYS8gMxchC+Oo0aNWLDhg08e/aM6dOnExMT880EBkHIL8Sag/BVatiwIXv37sXJyUlc+SwIKiCCg/DVql+/vqqbIAjfLTGtJAiCIGQhgoMgCIKQhQgOgiAIQhYiOAiCIAhZiOAgCIIgZCGCgyAIgpCFCA6CIAhCFiI4CIIgCFmI4CAIgiBkIYKDIAiCkIW4K6sgfCdUdTdYHXGTnnxJjBwEQRCELERwEARBELIQwUEQBEHIQgQHQRAEIQsRHAQhj8jlcqWfxV4PIT8TwUEQ/qPU1FQA1NTUuHXrFidOnCAmJgaZTKbilgnCpxPBQRD+g9u3bzNs2DAePXrEoUOHKFu2LDVq1KB06dKsX7+exMREVTdRED6JuM5BED6RXC5nwIAB/PXXX/Ts2ZO4uDh0dXUxNDRkz549JCYmsnXrVsqXL48kSSofSYjrHITcEMFBED5BWFgYZmZmxMXFMXz4cFatWoW+vj4zZsygT58+bNu2jR9//JEOHTqwdu1aVTcXEMFByB0xrSQIufT48WOcnJxo0KABBgYG/P7773h7e5OQkMCJEycIDQ3F1dUVKysr0tPTSUlJUXWTBSHXREwXhFwyNTWlYsWKnDhxgtatW7Nx40YmT55MSkoKa9eu5dSpU5QsWRItLS28vb3R1tZWdZMFIdfEyEEQciE1NRV9fX0WLlxI8eLF2bt3L+3bt8fQ0JBZs2bRs2dPHjx4QFxcHDt37qRBgwZiS6uQL4mRgyDkgpaWFrt27cLX1xd3d3fi4+PZtWsX7dq1Y/369UycOJHExEQ6duxIiRIlAFS+EC0In0IsSAvC/9u796io6v3/48+BQRgQxxsKZHgD44hXyBCVqEzEK14Qb60UQ1MSNA/mjQj7peDJKxqa2RHNUPNSYImOHsQELUFFSEFBVCS8fU2BUIEZ5vdHB46GdfRojsD7sRbLxbj3zHsWa+3Xfn8++7P3Q9Lr9Wi1WoYNG0ZaWhoajQZHR0c8PDxISkrCw8ODXbt2YWpqiomJyTNxhdK9ZEJaPAoZVhLiISkUCkxMTHjuuef4+eefOX/+PAqFgsmTJ2NhYcHBgwfJzMzExMSkanshairpHIT4E5Vn/xcuXODOnTtYW1uTnJzMiBEjqKioIDg4mJycHCwsLHjrrbfo2bOnoUv+Q4/aORxLTSH6n5+Tefonrl+/zrLIT3it9+uP/LnSOdRM0jkI8ScUCgVxcXG4u7vj5eVFs2bNUKvVLFq0CGtra5YuXUpsbCzjxo2jZ8+e6PX6WjMBfefObV544QXmhHxgsBqio6NRKBSsWbPmsd6nVatWWFtbP6Gq/uP39T3M54wfPx6FQkFWVtYTr+dJvr9kuhB/4syZM0ycOBFfX1/+9a9/oVAouHv3Lv7+/owdO5bs7GzMzc1r5eRzL3cPerl7GLQGDw8PNm/eTLdu3Qxax5M0ZcoUvLy8eO655576Z+t0OoyNjR9qW+kchHiAyrP/vLw8jI2NSUlJIS8vj40bN3Lq1Cnc3NzQ6/V07969KhhqS8dgKFevXuXVV1/FwsKChg0b0qtXLw4ePMjo0aPZt28f8NuZuZWVFRMnTqRhw4a8+uqrFBcXA7Bx40asra2xt7fH398fhUJBdHR0tc+5e/cuQUFB2NjY0KRJEyZMmMCvv/4KQGZmJp6enjRo0AArKyt27twJwM6dO+nQoQMWFha4urqSkpLyX7/PoUOHcHJywszMDBsbG/7+978DsHr1akaPHs3PP/+MXq9n2rRpqNVqXn31VTw9PauGMS9cuIBCoeCVV17htddeo0GDBgQFBQFw5coVOnfujIWFBWq1mkGDBnH9+vVqNdz7Hj179sTV1fWh/x4SDkLco/IAX3n77SZNmnD9+nWOHj3Khx9+iJOTE7m5uZSXl1NUVHTfvrWpa3iSSktLKSoquu/nQavGY2JiSExMJCQkhEWLFtG6desHvt///d//YWJiQp8+fUhMTGTLli1cuXKFSZMmYWZmxsyZM9FoNH9Yz4IFC1i1ahUjRoxgypQpbNq0idDQUMrLyxk8eDCHDh1i9uzZhISEYGRkRFZWFiNHjqRRo0aEhIRw+/ZtvL29uXv37p9+78WLF1NQUMDq1at57733qF+/frVtYmNjiYyMpFu3bgwcOJDExMRq2xw+fJj+/fvTqlUrVq5cSXZ2NsbGxvj6+rJixQqCgoLYvXs3ERERf1jLoUOH6Nu3L+++++6f1nwvGVYS4h4KhYKEhAQ+//xzPD09cXV1JTg4mMjISBYtWsT69evJyclh69attGnTxtDl1gjh4eHMnz//vtc++OADwsLC7nutbdu2AGg0Gnr27ElgYCCnT5+u9n4qlYrIyEgSExPZvn07Fy5c4Mcff6S0tJTJkyfz9ttv88svvzB37twH1hMfH49er2flypVVr+3fv5+zZ8+Sk5ODn5/fffuuWrUKrVZLUlISSUlJVa8/qLbff5/du3cTHx/Piy++iL+/f7VtDh48CMBHH31E9+7d0Wg01YLtlVdeITg4mFu3bpGRkcGFCxdwdHQkNjaW1NTUqhOaU6dO/WEtvXr1IjQ09E/r/T0JByH4z1VJaWlpTJo0ifPnz7N582beffddevfuTZs2bfjiiy+wtbVlyZIleHp6PnPrGJ5Vc+bMYcaMGfe99qBbigwePJjk5GT27dvHt99+S3h4OFFRUdW2a9CgAUqlEqXyt8OXTqer+j8jo4cbDDEzMyMuLq7q7/cwtzgJCQnBw+O3OZiKigpat25Nenr6H27/8ccf4+7uzpEjR4iMjCQyMpL8/PwHbvtndTdu3Bjgvu+7atUqUlJSmDdvHh4eHgwcOPBPOxkbG5v/+v1+T8JBCH7rGDQaDQsWLKB79+4sXLiQ7du3s2LFCnQ6HdOnT8ff35/y8nLq1asn8wuPwNTU9KEOvjt27ODkyZO0bduWF154gdTUVK5cufJQn+Hq6oqpqSlr166lYcOGf3p1U79+/Th27Bhffvkl7u7uHD9+HIVCwfLly7G3t2fz5s04ODhgYWGBnZ0dr7/+Okqlku3bt9OyZUuuXbvGxo0b/+vVQBEREZiamtKxY0esra1JT09Hq73/emIPDw+WL19OaGgonp6eHDhw4KG+b6WbN28SGxtb9cCpJ0nCQdR5Op2OiooKAgICyM3NxdXVFV9fX15++WWMjY2JjIykpKSElStXYmZmBtSN+YXbJSXk5eVV/f5zfj5ZmZmo1WpsbG2f+OepVCq++uorLly4gIWFBZMnT6ZFixYPta+1tTVr165l5syZLF++nJ49e5KXl0fDhg2rbTtv3jyKi4v56quv2LZtG+3ateO9995DqVQSFxdHUFAQ4eHhmJqa8umnn+Lo6MhXX31FaGgogYGBNG3alNdff7j1HsuXL+fatWtYW1uzcuXKqrP/St7e3gQFBbFhwwa0Wm3VDR3VajWFhYV/+L5Tp05l3759REdHM3XqVNRq9UPV8yhkEZyosyqHhSr/vXnzJn379iU1NZWAgACWLFnCtWvXePfdd5kwYQL9+/c3dMmP5VEXwaUc/RF/vzervT7Yeyj/b+EfT37+3tNaBPfFF1/QpEkTysvLmTVrFvn5+WRnZ/9PQypP05IlS+jcuTN5eXkEBgbi6OjIsWPHDF2WhIOomyoD4YcffmDXrl3Uq1ePIUOG0K5dO9zc3EhPT8fPz481a9ZQUVGBqalpjZ9jqO33Vpo5cyZr165Fp9PRvn17FixYQJ8+fZ7Ohz+GXr16cezYMczNzenRowfLli3D3t7e0GVJOIi6a8+ePfj4+NCgQQOuXLlC48aN2bhxIy+//DJOTk5cunSJjIwMnJycDF3qE1Hbw0E8WRIOok4qKiqib9++WFlZER4ezvnz5/Hx8aFDhw6kpKRw48YNTp06VXV1Sm0g4SAehSyCE3WSiYkJ169f59q1a7Rt25aBAwfi5+dHXl4e2dnZNG3aFA8Pj1p1ryQhHoWEg6gTKg/wV69epaioCJVKhbOzM0ePHmX27NkkJiZy+fJlmjVrhkqlqtpPoVDU6HkGIf5XMqwkarXs7GyysrIYNGgQ8fHx+Pn50aJFC/z8/HjjjTcYOHAgycnJ1K9fH2NjYzZu3MigQYMMXfZfQoaVxKOQP5uotYqKihgxYgTp6emsX7+e+Ph47O3t+emnnwgMDKS0tJSEhARiYmIoKyujS5cudOvWrcZflSTEkyCdg6jVPv30Uz7++GPy8/NxcHBgxYoVtGjRAjc3N27evElYWFi1e87U1nCQzkE8CplzELVS5f123n77bebNm0erVq3Iysri559/pl27dnz//fcYGxsTFhbG2bNn79u3NgaDEI9KMl3UOnq9HmNjY65evYqVlRV+fn6o1WqmTp3K5MmTMTU1xdfXl7S0NAoKCmjXrp2hSxbimSPhIGqNyuEghUJBfHw8AQEBqFQq3N3dWb16NUqlksDAQEaNGoVOp2P06NE4OTlVXckkHYMQ/yHhIGoNnU6HUqnkzJkzfPnll7Rt25aCggI+++wzCgoKiIuLo7CwkFmzZtGoUaOq/SQUhKhOJqRFjVdYWIhKpaJevXrs27ePvn370rVrV1avXo2TkxNjx44lLi6OrVu3MmLECPLz82nRokWtnXj+IzIhLR6FTEiLGu3cuXP07NmTqKgotFotly5dwsTEhFOnTnHlyhUsLCyYMmUKCoWi6kErlbeBrkvBIMSjkkwXNVpubi45OTksW7aMhg0bMmHCBJRKJVOmTGHYsGFMmDABrVaLqakpzs7Ohi5XiBpDOgdRI2VlZbFz50769OnD119/jVarZdasWURHR/Pmm2+yevVqLC0t+fzzzzEyMiIzMxMPDw8qKioMXboQNYLMOYgap7i4mJdeeokzZ86wZcsWfH19+e6775g0aRJarZZFixYxfvx41q9fz4wZM7h79y6xsbHy3GchHoF0DqLGsbS0ZPbs2dja2jJq1Ci2bNnCgAEDWLt2LUqlkpCQED777DP8/PxYvHgxpaWljBkzhpKSEkOXLkSNIZ2DqFEqKiowMvrtnCYmJoZp06Zx48YNYmJiGDVqFPHx8YwcORITExNOnz5N8+bNWb9+PW5ubjg6Ohq4eiFqDgkHUWNUDgnduHGDJk2aALB3717Gjh3LL7/8UjXE9O2336JSqejdu7eBKxai5pJwEDVCZTDs3r2b8PBwysrKeP7554mOjkaj0TB16lSuXLnCpk2bGDNmTNU+IJesCvG/kDkHUSMoFAqSk5MZM2YM7du3JyUlhe+//54TJ04wbNgwoqKiUKvVWFpa3rdPXQmG35/jyTmfeFwSDuKZdu9BLjExkZYtW3L58mUsLCyIjIwkPT2d999/nyFDhpCZmcmgQYPq1IHx3u6osoOq/F2IxyGL4MQzq3IoKT4+nnPnzmFqakpGRga5ubls3boVOzs7wsLCMDIyorCwEGtra6BuHhjT09NZuHAh9vb2FBQUYGtra+iSRA0n4SCeWQqFgv379zNkyBBeeeUVXn/9dWxtbSkoKGDjxo2UlJSQkJDAjh07UKvVhi73qbp58yaNGjVCoVBw6NAhevfujVqtZtasWdja2sp6DvHYZFhJPLOKi4tZsWIF7du355NPPmHmzJlERUXh4uLCjz/+yIULF9i6dSteXl51aigpLy+PNm3aMH/+fAAaN26Mra0tN27cID4+nrKyMhQKhawGF49FOgfxzFIoFFy8eBG9Xl81TGJnZ4eHhwfz58+nqKgIGxubOhUM8Nv9pJo2bcqHH36ImZkZs2bNYs+ePVUT802aNGHmzJmYm5sbulRRk+mFeEZVVFTop0+frlcoFPr3339fv2nTJv3EiRP15ubm+kuXLukrKioMXaLBaDQa/YsvvqhXKBT6iIgIvV6v1586dUrv5OSkVygU+rlz5+p1Op2BqxQ1mXQO4pmlUCiYMWMGmZmZfPTRRygUCtRqNZs2baq67XZdor9nHqFPnz6Ympoyffp05syZA8CsWbPYvHkzQ4cOpVevXlUryYX4X8giOPHMu3XrFsnJydy6dYtOnTrRsWPHOrfArTIYjhw5QlJSEnZ2dnh7e5ORkcHbb7/NyZMn+eCDDwgNDaWkpAQLCwuZlBaPRcJBiBrim2++Ydy4cZiZmVFUVMSIESOIiorizJkzjB49mpycHHJycmjTpo2hSxW1gISDEDXA4cOH8fT0ZNasWRw+fJj9+/ej0+l44403+OSTT8jKyqK0tJRevXoZulRRS8icgxA1gJmZGQEBAVy4cIEDBw7wxRdfEBERwaZNm7h79y6bN2/G2NgYQIaTxBMh4SDEM6jyAH/x4kWUSiXOzs40a9aMsWPH4uDggEqlolOnTuj1evz8/KqCAerOPIz4a8nlDEI8gypXh7u5ufHSSy+xatWqqkA4deoUU6ZM4fjx42zfvp1+/frVubUe4q8nnYMQz5DKjiE9PZ1//OMftG7dmry8PGbMmEF5eTleXl6Ym5tz7NgxAgMDcXBwAKRbEE+eTEgL8Yw5ePAgffv2pXnz5sTExGBhYUFgYCDHjx8nJCSEgIAAzM3NMTExkfkF8ZeRYSUhngH3nqNZWlrSqFEjLl26hEajoUuXLixdupR27doxb948rl+/jomJCSAdg/jrSOcgxDNi7969nDt3joCAAE6ePMnQoUO5dOkSCxYsIDg4mNTUVHQ6HW5uboYuVdQBMucghAFVDgtlZGQQFBREdnY2KpUKPz8/duzYga+vL7Nnz0ar1TJ37txq+wnxV5FhJSEMqPIJbhMnTsTKygqAt956i08//ZSuXbuyefNmbG1t6dq1a7X9hPgrybCSEE/Z1atXadq0KcbGxty4cYPevXuj1WqJiYkhJyeH6dOnk5+fz7p165gwYQKFhYWo1WrpFsRTJZ2DEE/RuXPnaNGiBfHx8QA0atQIS0tLioqKsLOzY9iwYQwfPhwAf39/NBpN1VPuJBjE0yThIMRTVFBQwMKFCzl79iwhISFotVoGDx5Mfn4+b775Jhs2bKC8vJyhQ4cCsHv3bgNXLOoqmZAW4imoHBJyd3fHwcGBIUOGcPToUdRqNcOGDSM3N5f169eze/du6tevz7Jly8jMzKRx48aGLl3UUTLnIMRTUBkOubm52Nracvr0aaZNm0ZycjKLFy9mzJgx5Obm8tNPP3H79m1u3brFmjVrSEpKwt7e3tDlizpIhpWE+ItVBkNcXBze3t7MmTMHZ2dnwsPDcXV1JTg4mMjISHr06MGkSZMoKSlh3bp1xMfHSzAIg5HOQYinICEhgQEDBhAYGIiLiwvdunXD0tKS4uJihg4dSkZGBqdPn8bR0RG9Xs+VK1ewsbExdNmiDpPOQYi/QOVzrytt27YNa2trbGxsOHToEG5ubjRv3hwTExPWrl1LcnIyjo6OaLVaFAqFBIMwOAkHIZ6wsrIyxo4dS2hoKO+//z4ADg4OXLx4kcWLF2NkZESfPn1QqVQkJCTg6upadUuMe5/LIIQhydVKQjxBN27coHHjxqxbt45Ro0axYMECVCoVc+fOxc7ODhsbG8rKyti7dy8AHTp0uG9/WcsgnhXSOQjxhOTn59OxY0fee+89unbtyvbt22nZsiUhISF8+OGH+Pj4YGVlxcqVK1mzZg0bNmzAxcXF0GUL8UDSOQjxhFROIq9evRpzc3PCwsKIjY3F29ubsLAw9Ho9H3zwAcOHDyc4OJgePXpU3apbOgbxrJGrlYR4gn744QfmzJnD4cOHmT17NmFhYWRkZODj40NOTg6hoaGEhYVVbS/3SxLPKhlWEuIx3Xt+1b17d5YtW4abmxsRERGEhYXRsWNHtm7dynPPPSd3VxU1hnQOQjyGyjP/5ORkjh8/TmlpKePGjSM7O5vQ0FB++OEHpk6dSnh4OL/++iuWlpbSLYgaQcJBiMf03XffMWrUKBwcHEhLS2PEiBGsWrWK/Px8/P39SUtLIysrCwcHB0OXKsRDk3AQ4jEUFhYyePBgWrVqxY0bNzh69CjLly+nrKyM8ePHc/z4ccrKyujevbuhSxXikcjVSkI8ggddXXTx4kVSU1MxMTEhKiqKVq1aMXjwYHQ6HW+99dZ9+8pwkqgpZEJaiEekUChISkpi27ZtVFRU0L9/f+7cuUP//v1RKpVs376devXqyQI3UaNJ5yDEI1AoFOzdu5d+/foxceJEHB0dGTlyJDk5OWzZsoWvv/4apVLJhg0bcHV1NXS5QvzPZM5BiEdw8+ZNPD09ady4MatWraqaZM7MzCQjI4Pi4mK6dOmCi4uLDCOJGk06ByEeQuWBXqfTcfHiRRo2bEjbtm0BSEtLIyMjg5EjR1KvXj0DVyrEkyHhIMSfqAyF8vJy6tWrR9OmTenYsSPJycksWbIEZ2dnoqOjycrKwtPTk+bNm1ftK12DqMlkWEmIP1AZDPv27WPt2rUYGRnh4uJC3759GTt2LKdPn8bc3BylUsnnn3/O8OHDDV2yEE+MhIMQD6DValEqlRw4cIBBgwbh5eXFzp07ef7554mLi8Pa2poNGzagUql46aWXcHV1lTkGUatIOAhxj4KCApo1a4ZSqeTOnTu88847FBQUYGlpiUajYd26dVy+fBkrKytGjx59374SDqI2kXUOQvxbfn4+LVu2xNPTk/LyclQqFWVlZWg0GjQaDTExMdjb27N3715SU1O5ffv2fftLMIjaRMJBiH8zNzenY8eOJCYm4uPjg06nw8vLC6VSiampKSdOnGDx4sV8//33DBgwAHNzc0OXLMRfRoaVRJ2n1+vR6XQolUp++eUXfH19SUhIwNfXly1bthAREcGiRYuoqKigadOmLF26FG9vbxlGErWahIMQ/5aens7zzz+PQqHAx8eHhIQERo8ezaZNm7h48SKXL1+madOmODg4yBPcRK0n4SDqtMqz/4yMDDp37sz48eNZunQpQFVAeHl5ERsbi4mJiYGrFeLpkTkHUacpFAo0Gg0REREAxMTEEBwcDMCOHTvo3r07e/bs4eTJk4YsU4inTjoHUaedPXsWZ2dnRo4cSadOnUhKSmLnzp1MmjSJhQsXVm0jN9ETdY3cPkPUSZXDSQUFBdy+fZtGjRoxbdo0BgwYQFJSEp999hkqlYqlS5dWLXADmWMQdYcMK4k6pfIgX1JSAkCnTp2ws7Pj22+/Zc+ePdy9e5cePXrg5ubG8uXLOX78OPBbKEgwiLpEwkHUKZXPY3jttdcIDAwkIyODhQsXUlJSQv/+/RkwYAAnTpzAxcUFtVqNVqs1dMlCGIQMK4k6QafTYWxsTGZmJgsWLODSpUukpqaSlpZGYGAg27ZtIyoqiuLiYoYMGcKZM2do1qwZrVq1MnTpQhiEdA6iVqu8xYWxsTHx8fH07dsXCwsLvvzySxYvXsyPP/7IihUrKCkpYePGjcyePZuTJ08SFRXFli1baNasmYG/gRCGIeEgaq2MjAx8fHzIyckBYOfOneTn53Ps2DEcHR2ZMWMGERERHDlyhODgYAoLC3FxcaFz586kpKTQtWtXA38DIQxHhpVEraTVatm1axetWrUiLS2NO3fuVF2BtGrVKry8vNi1axczZsxAp9PRoUMH1Go1AOPGjTNw9UIYnqxzELVWRUUFJ06cwN3dnb/97W9ER0fTsWNHAgICWLNmDa1bt+bAgQPY2dkByOWqQtxDhpVErVNRUQFAUlISsbGxzJ8/n+zsbN555x0yMjKIiopi3LhxnD9/nosXL1btJ5erCvEfEg6i1jEyMuKbb77B29ub/fv30759e2bPns2JEycICgoiLS2N9evXk56ejru7O9I8C1GdDCuJWuenn37C3d2dyZMn06VLF0xMTNBoNLRo0YLQ0FBcXV3Zt28f9evXB+QJbkI8iExIi1rnzp076HQ6EhISyMrKoqCggJSUFLy9vfnoo49wdnauCgaQOQYhHkTCQdQ67dq1w8fHh8TERIYPH06TJk2wsLCgvLwcf39/mjVrJt2CEP+FDCuJWqm0tBStVsvRo0fZunUr//znP4mNjaVfv36GLk2IGkE6B1ErmZqaotfrSU1N5ciRI2zdupV+/fpJxyDEQ5LOQdRqv/76K7du3aJFixayjkGIRyDhIIQQohpZ5yCEEKIaCQchhBDVSDgIIYSoRsJBCCFENRIOQgghqpFwEEIIUY2EgxBCiGokHIQQQlQj4SCEEKKa/w+zvBI/blWmCgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 400x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = c2c.plotting.clustermap_cci(specificity_corr,\n",
    "                                 method='ward',\n",
    "                                 optimal_leaf=True,\n",
    "                                 metadata=None,\n",
    "                                 title='',\n",
    "                                 cbar_title='Spearman', \n",
    "                                 cmap='Blues_r',\n",
    "                                 vmax=1.,\n",
    "#                                  vmin=0.,\n",
    "                                 annot=True, \n",
    "                                 dendrogram_ratio=0.15,\n",
    "                                figsize=(4,5))\n",
    "\n",
    "font = matplotlib.font_manager.FontProperties(weight='bold', size=7)\n",
    "for ax in [cm.ax_heatmap, cm.ax_cbar]:\n",
    "    for tick in ax.get_xticklabels():\n",
    "        tick.set_fontproperties(font)\n",
    "    for tick in ax.get_yticklabels():\n",
    "        tick.set_fontproperties(font)\n",
    "\n",
    "    text = ax.yaxis.label\n",
    "    text.set_font_properties(font)\n",
    "cm.ax_col_dendrogram.set_title('Specificity Consensus')\n",
    ";"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "589a9b3e",
   "metadata": {},
   "source": [
    "Specificity consensus scores are even less sensitive to individual scoring methods than magnitude consensus scores. While the consensus score leaving CellChat out still has the lowest consistency as compared to including all scoring methods, it still has a large Spearman correlation (0.91). Overall, these results indicate that aggregating across scoring methods yields robust results that are not sensitive to any individual method.  "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c71f40dc",
   "metadata": {},
   "source": [
    "Given the high overall similarity between score types, we see that Tensor-cell2cell's decomposition tends to capture consistent patterns across samples, smoothing over some of the inconsistencies in communication scores that we saw at the individual sample level in [Tutorial 02](./02-Infer-Communication-Scores.ipynb) "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:ccc_protocols]",
   "language": "python",
   "name": "conda-env-ccc_protocols-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  },
  "vscode": {
   "interpreter": {
    "hash": "a89d9df9e41c144bbb86b791904f32fb0efeb7b488a88d676a8bce57017c9696"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}