behaviour_demography_notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d6c385f6",
   "metadata": {},
   "source": [
    "# Forest elephant rumble vocalisation analysis\n",
    "\n",
    "This Jupyter notebook provides a step-by-step guide to using unsupervised deep learning techniques (UDL) to identify rumble vocalisation sub-types and to identify associations between rumble acoustic features and different contexts (age, sex, behaviour, distress). This builds on the call-types analysis notebook using the main UDL techniques of transfer learning to automatically extract acoustic features and unsupervised ordination of these features into lower dimensional space. \n",
    "\n",
    "## Dataset Description\n",
    "\n",
    "The call-type dataset contains audio files of African forest elephants rumble vocalisations recorded by the Elephant Listening Project in the Dzanga-Bai clearing in the the southwestern Central African Republic between September 2018 and April 2019. It has 361 rows each representing an elephant rumble with the start time, end time, low frequency and high frequency annotated alongside different categorical labels pertaining to the age, sex, behavioural context and distress of the vocaliser.\n",
    "\n",
    "## Steps\n",
    "\n",
    "1. **Set-up** Import the libraries and functions needed to conduct the analysis, load the dataset and understand its structure\n",
    "2. **Audio pre-processing and feature extraction**: Pre-process the data to isolate the vocalisations and automatically extract the acoustic features using a pre-trained Convolutional Neural Network.\n",
    "3. **Dimensionality reduction**: Project the acoustic features into 1D space as the basis for clustering and statistical analysis\n",
    "4. **Unsupervised clustering**: Cluster the UMAP ordination of the acoustic features to identify sub-groups and the most prominent categorical labels in each cluster. This corresponds to research question 2 in the paper.\n",
    "5. **Behavioural and demographic analysis**: Perform statistical tests to assess the signficance of changes in acoustic features and different contexts. This corresponds to research question 3 in the paper."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c167756",
   "metadata": {},
   "source": [
    "### 1. Set-up"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e5347cf",
   "metadata": {},
   "source": [
    "Here we will import a series of pre-defined scripts used to automate the import of the libraries needed for each of the aforementioned steps. These are all located in the \"elephant_scripts\" folder in the main project repository."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "09318be4",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Main libraries successfully imported\n",
      "Info file loaded: /home/alastair/forestdata/alastair/temp/elephant_analysis_repository/data/2.elephant_rumble_behaviour_df.csv\n",
      "Audio file paths added to DataFrame\n",
      "Vocalisation Dataset successfully loaded\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Functions for Audio Pre-processing and Feature Extraction successfully loaded\n",
      "Functions for Dimensionality Reduction successfully loaded\n",
      "Functions for Statistical Analysis successfully loaded\n"
     ]
    },
    {
     "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>filename</th>\n",
       "      <th>start_time</th>\n",
       "      <th>end_time</th>\n",
       "      <th>duration</th>\n",
       "      <th>low_freq</th>\n",
       "      <th>high_freq</th>\n",
       "      <th>call-type</th>\n",
       "      <th>Final_Category</th>\n",
       "      <th>Distress</th>\n",
       "      <th>Age</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age_Sex</th>\n",
       "      <th>audio_filepaths</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>opp_578_ru.wav</td>\n",
       "      <td>2.392645</td>\n",
       "      <td>6.613464</td>\n",
       "      <td>4.220819</td>\n",
       "      <td>0.219727</td>\n",
       "      <td>1417.993164</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Nursing</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Juvenile</td>\n",
       "      <td>Male</td>\n",
       "      <td>Juvenile Male</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>opp_595_ru.wav</td>\n",
       "      <td>1.807122</td>\n",
       "      <td>10.038406</td>\n",
       "      <td>8.231284</td>\n",
       "      <td>1.025391</td>\n",
       "      <td>1397.607422</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Logistics</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>opp_560_ru.wav</td>\n",
       "      <td>2.299110</td>\n",
       "      <td>6.934491</td>\n",
       "      <td>4.635381</td>\n",
       "      <td>1.147461</td>\n",
       "      <td>546.850586</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Separation</td>\n",
       "      <td>No distress</td>\n",
       "      <td>Sub-Adult</td>\n",
       "      <td>Male</td>\n",
       "      <td>Sub-Adult Male</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>opp_420_ru.wav</td>\n",
       "      <td>1.832432</td>\n",
       "      <td>6.438646</td>\n",
       "      <td>4.606213</td>\n",
       "      <td>0.708008</td>\n",
       "      <td>486.108398</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Sexual</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opp_582_ru5.wav</td>\n",
       "      <td>0.982550</td>\n",
       "      <td>5.686105</td>\n",
       "      <td>4.703555</td>\n",
       "      <td>1.391602</td>\n",
       "      <td>481.103516</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Affiliation</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Sub-Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Sub-Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>356</th>\n",
       "      <td>opp_346_ru.wav</td>\n",
       "      <td>0.794323</td>\n",
       "      <td>6.482651</td>\n",
       "      <td>5.688328</td>\n",
       "      <td>0.585938</td>\n",
       "      <td>168.164062</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Separation</td>\n",
       "      <td>Distress</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>357</th>\n",
       "      <td>opp_669_ru1.wav</td>\n",
       "      <td>2.830313</td>\n",
       "      <td>4.420829</td>\n",
       "      <td>1.590516</td>\n",
       "      <td>0.048828</td>\n",
       "      <td>163.940430</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Logistics</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Sub-Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Sub-Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>358</th>\n",
       "      <td>opp_755_ru2.wav</td>\n",
       "      <td>1.031033</td>\n",
       "      <td>1.726944</td>\n",
       "      <td>0.695912</td>\n",
       "      <td>23.608398</td>\n",
       "      <td>140.112305</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Competition</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Juvenile</td>\n",
       "      <td>Male</td>\n",
       "      <td>Juvenile Male</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>359</th>\n",
       "      <td>opp_495_ru.wav</td>\n",
       "      <td>1.136562</td>\n",
       "      <td>3.769334</td>\n",
       "      <td>2.632772</td>\n",
       "      <td>2.612305</td>\n",
       "      <td>136.523438</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Competition</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>360</th>\n",
       "      <td>opp_424_ru.wav</td>\n",
       "      <td>2.302418</td>\n",
       "      <td>7.350745</td>\n",
       "      <td>5.048326</td>\n",
       "      <td>0.927734</td>\n",
       "      <td>96.801758</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Competition</td>\n",
       "      <td>No distress</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>361 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            filename  start_time   end_time  duration   low_freq    high_freq  \\\n",
       "0     opp_578_ru.wav    2.392645   6.613464  4.220819   0.219727  1417.993164   \n",
       "1     opp_595_ru.wav    1.807122  10.038406  8.231284   1.025391  1397.607422   \n",
       "2     opp_560_ru.wav    2.299110   6.934491  4.635381   1.147461   546.850586   \n",
       "3     opp_420_ru.wav    1.832432   6.438646  4.606213   0.708008   486.108398   \n",
       "4    opp_582_ru5.wav    0.982550   5.686105  4.703555   1.391602   481.103516   \n",
       "..               ...         ...        ...       ...        ...          ...   \n",
       "356   opp_346_ru.wav    0.794323   6.482651  5.688328   0.585938   168.164062   \n",
       "357  opp_669_ru1.wav    2.830313   4.420829  1.590516   0.048828   163.940430   \n",
       "358  opp_755_ru2.wav    1.031033   1.726944  0.695912  23.608398   140.112305   \n",
       "359   opp_495_ru.wav    1.136562   3.769334  2.632772   2.612305   136.523438   \n",
       "360   opp_424_ru.wav    2.302418   7.350745  5.048326   0.927734    96.801758   \n",
       "\n",
       "    call-type Final_Category        Distress        Age     Sex  \\\n",
       "0      rumble        Nursing  Not Applicable   Juvenile    Male   \n",
       "1      rumble      Logistics  Not Applicable      Adult  Female   \n",
       "2      rumble     Separation     No distress  Sub-Adult    Male   \n",
       "3      rumble         Sexual  Not Applicable      Adult  Female   \n",
       "4      rumble    Affiliation  Not Applicable  Sub-Adult  Female   \n",
       "..        ...            ...             ...        ...     ...   \n",
       "356    rumble     Separation        Distress      Adult  Female   \n",
       "357    rumble      Logistics  Not Applicable  Sub-Adult  Female   \n",
       "358    rumble    Competition         Unknown   Juvenile    Male   \n",
       "359    rumble    Competition         Unknown      Adult  Female   \n",
       "360    rumble    Competition     No distress      Adult  Female   \n",
       "\n",
       "              Age_Sex                                    audio_filepaths  \n",
       "0       Juvenile Male  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "1        Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "2      Sub-Adult Male  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "3        Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "4    Sub-Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "..                ...                                                ...  \n",
       "356      Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "357  Sub-Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "358     Juvenile Male  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "359      Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "360      Adult Female  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "\n",
       "[361 rows x 13 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Step 1: load all of the main libraries needed for the notebook set-up\n",
    "%run elephant_scripts/import_libraries_behaviour\n",
    "%run elephant_scripts/load_data_behaviour\n",
    "\n",
    "# Step 2: load the functions needed for audio pre-processing and feature extraction\n",
    "%run elephant_scripts/feature_extraction\n",
    "\n",
    "# Step 3 & 4: load the functions needed for dimensionality reduction and clustering\n",
    "%run elephant_scripts/umap_projections_behaviour\n",
    "\n",
    "# Step 5: load the functions to calculate and visualise the statistical relationship between rumbles and behavioural contexts\n",
    "%run elephant_scripts/behaviour_stats\n",
    "\n",
    "# enable plots to be auto-displayed in the notebook\n",
    "%matplotlib inline \n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "451bdc53",
   "metadata": {},
   "source": [
    "### 2. Audio pre-processing and feature extraction"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f7c9d36",
   "metadata": {},
   "source": [
    "Now that the dataframe is loaded and associated with the audio files, we need to pre-process the audio files to extract the exact frequency ranges and time periods associated with the vocalisations. This helps to minimise unwanted environmental sound that may cause erroneous results. These are the steps take to pre-process the files and extract their audio features:\n",
    "\n",
    "<ol type=\"A\">\n",
    "\n",
    "<li>Read the entire audio file.<li>\n",
    "Apply a bandpass filter to exclude frequencies outside the vocalisation range using the low_frequency and high_frequency information.<li>\n",
    "Extract the audio clip corresponding to the vocalisation using the start_time and end_time information.<li>\n",
    "Zero-pad the vocalisation to a length that is a multiple of 96 (the input size of the VGGish model) and centre the recording within this padding.<li>\n",
    "Normalise the audio clip to have a peak amplitude of 1 to control for elephant distance from the microphone.<li>\n",
    "Reapply the bandpass filter to remove any acoustic artifacts that have been introduced by the pre-processing."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "466636ac",
   "metadata": {},
   "source": [
    "These individual steps are combined in one single function called the wav_cookiecutter which we'll use for the pre-processing the rest of the data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59e99a65",
   "metadata": {},
   "source": [
    "The pre-processed audio files are then passed through the VGGish pre-trained CNN in 0.96s multiples and their acoustic features are automatically extracted to produce embeddings. These embeddings are then averaged to obtain a single embedding per vocalisation. This feature extraction phase involves the following steps:\n",
    "\n",
    "1. Extract the acoustic feature embeddings for each 96-sample chunk of the audio clip using the pretrained CNN VGGish.\n",
    "2. Average the embeddings across the chunks to obtain a single embedding for each vocalisation.\n",
    "3. Add in information about the duration of each vocalisation\n",
    "\n",
    "This resulting embedding encodes the acoustic feature representation of the vocalisation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "28053bf7",
   "metadata": {
    "metadata": {},
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Processing: 100%|██████████| 361/361 [00:22<00:00, 15.76it/s]\n"
     ]
    },
    {
     "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>feature_0</th>\n",
       "      <th>feature_1</th>\n",
       "      <th>feature_2</th>\n",
       "      <th>feature_3</th>\n",
       "      <th>feature_4</th>\n",
       "      <th>feature_5</th>\n",
       "      <th>feature_6</th>\n",
       "      <th>feature_7</th>\n",
       "      <th>feature_8</th>\n",
       "      <th>feature_9</th>\n",
       "      <th>...</th>\n",
       "      <th>feature_119</th>\n",
       "      <th>feature_120</th>\n",
       "      <th>feature_121</th>\n",
       "      <th>feature_122</th>\n",
       "      <th>feature_123</th>\n",
       "      <th>feature_124</th>\n",
       "      <th>feature_125</th>\n",
       "      <th>feature_126</th>\n",
       "      <th>feature_127</th>\n",
       "      <th>duration</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filename</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>opp_330_ru.wav</th>\n",
       "      <td>-0.246848</td>\n",
       "      <td>-0.111789</td>\n",
       "      <td>0.241722</td>\n",
       "      <td>-0.195451</td>\n",
       "      <td>0.101328</td>\n",
       "      <td>-0.366939</td>\n",
       "      <td>-0.180415</td>\n",
       "      <td>0.016249</td>\n",
       "      <td>-0.349659</td>\n",
       "      <td>-0.147815</td>\n",
       "      <td>...</td>\n",
       "      <td>0.047084</td>\n",
       "      <td>-0.12486</td>\n",
       "      <td>-0.135784</td>\n",
       "      <td>0.080366</td>\n",
       "      <td>0.291493</td>\n",
       "      <td>-0.100012</td>\n",
       "      <td>-0.192962</td>\n",
       "      <td>0.228055</td>\n",
       "      <td>-0.119971</td>\n",
       "      <td>2.161813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_332_ru.wav</th>\n",
       "      <td>-0.14961</td>\n",
       "      <td>0.129676</td>\n",
       "      <td>0.085841</td>\n",
       "      <td>-0.098854</td>\n",
       "      <td>0.040766</td>\n",
       "      <td>-0.487508</td>\n",
       "      <td>-0.158237</td>\n",
       "      <td>0.22541</td>\n",
       "      <td>-0.535318</td>\n",
       "      <td>0.109821</td>\n",
       "      <td>...</td>\n",
       "      <td>0.107642</td>\n",
       "      <td>-0.155294</td>\n",
       "      <td>0.056881</td>\n",
       "      <td>-0.230121</td>\n",
       "      <td>-0.001885</td>\n",
       "      <td>-0.100687</td>\n",
       "      <td>-0.297468</td>\n",
       "      <td>-0.061319</td>\n",
       "      <td>0.094915</td>\n",
       "      <td>3.522186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru.wav</th>\n",
       "      <td>-0.161457</td>\n",
       "      <td>0.042482</td>\n",
       "      <td>0.090267</td>\n",
       "      <td>-0.00557</td>\n",
       "      <td>-0.023373</td>\n",
       "      <td>-0.54754</td>\n",
       "      <td>-0.165326</td>\n",
       "      <td>0.148796</td>\n",
       "      <td>-0.87906</td>\n",
       "      <td>-0.020349</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.049731</td>\n",
       "      <td>-0.290999</td>\n",
       "      <td>0.053796</td>\n",
       "      <td>-0.348644</td>\n",
       "      <td>0.07725</td>\n",
       "      <td>-0.015267</td>\n",
       "      <td>-0.368231</td>\n",
       "      <td>-0.094716</td>\n",
       "      <td>0.16168</td>\n",
       "      <td>3.540844</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru1.wav</th>\n",
       "      <td>-0.113422</td>\n",
       "      <td>0.033881</td>\n",
       "      <td>0.179464</td>\n",
       "      <td>-0.152382</td>\n",
       "      <td>0.039331</td>\n",
       "      <td>-0.561756</td>\n",
       "      <td>-0.092054</td>\n",
       "      <td>0.174603</td>\n",
       "      <td>-1.016679</td>\n",
       "      <td>-0.091761</td>\n",
       "      <td>...</td>\n",
       "      <td>0.036599</td>\n",
       "      <td>-0.223549</td>\n",
       "      <td>0.051192</td>\n",
       "      <td>-0.232041</td>\n",
       "      <td>0.077678</td>\n",
       "      <td>0.030441</td>\n",
       "      <td>-0.270696</td>\n",
       "      <td>-0.020462</td>\n",
       "      <td>0.183512</td>\n",
       "      <td>3.441889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru2.wav</th>\n",
       "      <td>-0.242428</td>\n",
       "      <td>-0.14745</td>\n",
       "      <td>0.098017</td>\n",
       "      <td>-0.081086</td>\n",
       "      <td>0.001065</td>\n",
       "      <td>-0.622324</td>\n",
       "      <td>-0.058791</td>\n",
       "      <td>0.168942</td>\n",
       "      <td>-1.242459</td>\n",
       "      <td>-0.102003</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.080686</td>\n",
       "      <td>-0.400064</td>\n",
       "      <td>0.02553</td>\n",
       "      <td>-0.511895</td>\n",
       "      <td>-0.01331</td>\n",
       "      <td>-0.027813</td>\n",
       "      <td>-0.63056</td>\n",
       "      <td>-0.074626</td>\n",
       "      <td>0.185636</td>\n",
       "      <td>3.261884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru2.wav</th>\n",
       "      <td>-0.024894</td>\n",
       "      <td>0.045767</td>\n",
       "      <td>0.110447</td>\n",
       "      <td>-0.072183</td>\n",
       "      <td>0.243159</td>\n",
       "      <td>-0.255672</td>\n",
       "      <td>-0.051282</td>\n",
       "      <td>0.203329</td>\n",
       "      <td>-0.238551</td>\n",
       "      <td>0.004844</td>\n",
       "      <td>...</td>\n",
       "      <td>0.044813</td>\n",
       "      <td>-0.191217</td>\n",
       "      <td>0.16405</td>\n",
       "      <td>-0.196267</td>\n",
       "      <td>0.083124</td>\n",
       "      <td>-0.06605</td>\n",
       "      <td>-0.152629</td>\n",
       "      <td>-0.011545</td>\n",
       "      <td>0.017836</td>\n",
       "      <td>2.381324</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru3.wav</th>\n",
       "      <td>-0.392591</td>\n",
       "      <td>-0.193636</td>\n",
       "      <td>0.31771</td>\n",
       "      <td>-0.121416</td>\n",
       "      <td>0.190764</td>\n",
       "      <td>-0.330069</td>\n",
       "      <td>-0.027058</td>\n",
       "      <td>0.165353</td>\n",
       "      <td>-0.229536</td>\n",
       "      <td>-0.33113</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.068928</td>\n",
       "      <td>-0.083521</td>\n",
       "      <td>-0.026139</td>\n",
       "      <td>-0.214735</td>\n",
       "      <td>-0.040516</td>\n",
       "      <td>-0.14961</td>\n",
       "      <td>-0.49971</td>\n",
       "      <td>0.198395</td>\n",
       "      <td>-0.227462</td>\n",
       "      <td>1.188828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_878_ru.wav</th>\n",
       "      <td>-0.382297</td>\n",
       "      <td>0.227646</td>\n",
       "      <td>0.053632</td>\n",
       "      <td>0.005211</td>\n",
       "      <td>-0.104698</td>\n",
       "      <td>-0.381395</td>\n",
       "      <td>-0.078889</td>\n",
       "      <td>0.193316</td>\n",
       "      <td>-0.23203</td>\n",
       "      <td>-0.115835</td>\n",
       "      <td>...</td>\n",
       "      <td>0.137279</td>\n",
       "      <td>-0.134785</td>\n",
       "      <td>0.086429</td>\n",
       "      <td>-0.174174</td>\n",
       "      <td>-0.077103</td>\n",
       "      <td>-0.120511</td>\n",
       "      <td>-0.409068</td>\n",
       "      <td>-0.004873</td>\n",
       "      <td>0.045533</td>\n",
       "      <td>4.546284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru.wav</th>\n",
       "      <td>-0.053071</td>\n",
       "      <td>0.05196</td>\n",
       "      <td>0.149965</td>\n",
       "      <td>-0.020638</td>\n",
       "      <td>0.040171</td>\n",
       "      <td>-0.049215</td>\n",
       "      <td>-0.107295</td>\n",
       "      <td>0.059979</td>\n",
       "      <td>-0.295649</td>\n",
       "      <td>-0.002131</td>\n",
       "      <td>...</td>\n",
       "      <td>0.235516</td>\n",
       "      <td>0.07404</td>\n",
       "      <td>0.059339</td>\n",
       "      <td>-0.085721</td>\n",
       "      <td>0.207026</td>\n",
       "      <td>-0.060098</td>\n",
       "      <td>-0.149239</td>\n",
       "      <td>0.07116</td>\n",
       "      <td>0.125631</td>\n",
       "      <td>2.756656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru1.wav</th>\n",
       "      <td>-0.302939</td>\n",
       "      <td>0.225697</td>\n",
       "      <td>0.088803</td>\n",
       "      <td>-0.225169</td>\n",
       "      <td>-0.111368</td>\n",
       "      <td>-0.362085</td>\n",
       "      <td>-0.078202</td>\n",
       "      <td>0.164126</td>\n",
       "      <td>-0.207893</td>\n",
       "      <td>-0.057435</td>\n",
       "      <td>...</td>\n",
       "      <td>0.149569</td>\n",
       "      <td>-0.085267</td>\n",
       "      <td>0.033844</td>\n",
       "      <td>-0.005974</td>\n",
       "      <td>-0.020446</td>\n",
       "      <td>-0.20069</td>\n",
       "      <td>-0.390099</td>\n",
       "      <td>-0.013662</td>\n",
       "      <td>-0.081509</td>\n",
       "      <td>8.172052</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>361 rows × 129 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                feature_0 feature_1 feature_2 feature_3 feature_4 feature_5  \\\n",
       "filename                                                                      \n",
       "opp_330_ru.wav  -0.246848 -0.111789  0.241722 -0.195451  0.101328 -0.366939   \n",
       "opp_332_ru.wav   -0.14961  0.129676  0.085841 -0.098854  0.040766 -0.487508   \n",
       "opp_333_ru.wav  -0.161457  0.042482  0.090267  -0.00557 -0.023373  -0.54754   \n",
       "opp_333_ru1.wav -0.113422  0.033881  0.179464 -0.152382  0.039331 -0.561756   \n",
       "opp_333_ru2.wav -0.242428  -0.14745  0.098017 -0.081086  0.001065 -0.622324   \n",
       "...                   ...       ...       ...       ...       ...       ...   \n",
       "opp_876_ru2.wav -0.024894  0.045767  0.110447 -0.072183  0.243159 -0.255672   \n",
       "opp_876_ru3.wav -0.392591 -0.193636   0.31771 -0.121416  0.190764 -0.330069   \n",
       "opp_878_ru.wav  -0.382297  0.227646  0.053632  0.005211 -0.104698 -0.381395   \n",
       "opp_881_ru.wav  -0.053071   0.05196  0.149965 -0.020638  0.040171 -0.049215   \n",
       "opp_881_ru1.wav -0.302939  0.225697  0.088803 -0.225169 -0.111368 -0.362085   \n",
       "\n",
       "                feature_6 feature_7 feature_8 feature_9  ... feature_119  \\\n",
       "filename                                                 ...               \n",
       "opp_330_ru.wav  -0.180415  0.016249 -0.349659 -0.147815  ...    0.047084   \n",
       "opp_332_ru.wav  -0.158237   0.22541 -0.535318  0.109821  ...    0.107642   \n",
       "opp_333_ru.wav  -0.165326  0.148796  -0.87906 -0.020349  ...   -0.049731   \n",
       "opp_333_ru1.wav -0.092054  0.174603 -1.016679 -0.091761  ...    0.036599   \n",
       "opp_333_ru2.wav -0.058791  0.168942 -1.242459 -0.102003  ...   -0.080686   \n",
       "...                   ...       ...       ...       ...  ...         ...   \n",
       "opp_876_ru2.wav -0.051282  0.203329 -0.238551  0.004844  ...    0.044813   \n",
       "opp_876_ru3.wav -0.027058  0.165353 -0.229536  -0.33113  ...   -0.068928   \n",
       "opp_878_ru.wav  -0.078889  0.193316  -0.23203 -0.115835  ...    0.137279   \n",
       "opp_881_ru.wav  -0.107295  0.059979 -0.295649 -0.002131  ...    0.235516   \n",
       "opp_881_ru1.wav -0.078202  0.164126 -0.207893 -0.057435  ...    0.149569   \n",
       "\n",
       "                feature_120 feature_121 feature_122 feature_123 feature_124  \\\n",
       "filename                                                                      \n",
       "opp_330_ru.wav     -0.12486   -0.135784    0.080366    0.291493   -0.100012   \n",
       "opp_332_ru.wav    -0.155294    0.056881   -0.230121   -0.001885   -0.100687   \n",
       "opp_333_ru.wav    -0.290999    0.053796   -0.348644     0.07725   -0.015267   \n",
       "opp_333_ru1.wav   -0.223549    0.051192   -0.232041    0.077678    0.030441   \n",
       "opp_333_ru2.wav   -0.400064     0.02553   -0.511895    -0.01331   -0.027813   \n",
       "...                     ...         ...         ...         ...         ...   \n",
       "opp_876_ru2.wav   -0.191217     0.16405   -0.196267    0.083124    -0.06605   \n",
       "opp_876_ru3.wav   -0.083521   -0.026139   -0.214735   -0.040516    -0.14961   \n",
       "opp_878_ru.wav    -0.134785    0.086429   -0.174174   -0.077103   -0.120511   \n",
       "opp_881_ru.wav      0.07404    0.059339   -0.085721    0.207026   -0.060098   \n",
       "opp_881_ru1.wav   -0.085267    0.033844   -0.005974   -0.020446    -0.20069   \n",
       "\n",
       "                feature_125 feature_126 feature_127  duration  \n",
       "filename                                                       \n",
       "opp_330_ru.wav    -0.192962    0.228055   -0.119971  2.161813  \n",
       "opp_332_ru.wav    -0.297468   -0.061319    0.094915  3.522186  \n",
       "opp_333_ru.wav    -0.368231   -0.094716     0.16168  3.540844  \n",
       "opp_333_ru1.wav   -0.270696   -0.020462    0.183512  3.441889  \n",
       "opp_333_ru2.wav    -0.63056   -0.074626    0.185636  3.261884  \n",
       "...                     ...         ...         ...       ...  \n",
       "opp_876_ru2.wav   -0.152629   -0.011545    0.017836  2.381324  \n",
       "opp_876_ru3.wav    -0.49971    0.198395   -0.227462  1.188828  \n",
       "opp_878_ru.wav    -0.409068   -0.004873    0.045533  4.546284  \n",
       "opp_881_ru.wav    -0.149239     0.07116    0.125631  2.756656  \n",
       "opp_881_ru1.wav   -0.390099   -0.013662   -0.081509  8.172052  \n",
       "\n",
       "[361 rows x 129 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Set audio parameters used in the functions below\n",
    "samplerate = DEFAULT_SAMPLERATE\n",
    "window_size = DEFAULT_WINDOW_SIZE\n",
    "position = DEFAULT_CLIP_POSITION\n",
    "\n",
    "model = hub.load('https://tfhub.dev/google/vggish/1')\n",
    "\n",
    "# Extract VGGish features\n",
    "results = []\n",
    "for _, annotation in tqdm(df.iterrows(), total=len(df), desc=\"Processing\"):\n",
    "    # Apply the wav_cookiecutter function combining all of the audio\n",
    "    # pre-processing steps\n",
    "    wav = wav_cookiecutter(\n",
    "        annotation,\n",
    "        window_size=window_size,\n",
    "        position=\"middle\",\n",
    "        samplerate=samplerate,\n",
    "    )\n",
    "\n",
    "    embeddings = model(wav)\n",
    "    assert embeddings.shape[1] == 128  # Check the number of features per frame\n",
    "\n",
    "    # Store info of the embeddings of each frame\n",
    "    for embedding in embeddings:\n",
    "        results.append(\n",
    "            {\n",
    "                \"filename\": annotation.filename,\n",
    "                **{f\"feature_{n}\": feat for n, feat in enumerate(embedding)},\n",
    "            }\n",
    "        )\n",
    "results = pd.DataFrame(results)\n",
    "\n",
    "# average vggish annotation feature vectors back into original # of annotations\n",
    "results = results.groupby(\"filename\").mean()\n",
    "\n",
    "# Add in the missing duration information as the 129th feature\n",
    "duration = df[[\"filename\", \"duration\"]].copy()\n",
    "results = results.join(duration.set_index(\"filename\"))\n",
    "\n",
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "0b03fc61",
   "metadata": {
    "metadata": {}
   },
   "outputs": [],
   "source": [
    "# Save the extracted acoustic features to a csv file\n",
    "results.to_csv(os.path.join(os.path.sep, DATA, 'rumble_acoustic_features.csv'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a08e09a",
   "metadata": {},
   "source": [
    "### 3. Dimensionality reduction"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "534b634a",
   "metadata": {},
   "source": [
    "Now that we have the 129D embeddings for each of the 361 recordings we need to reduce this high-dimensional data into lower dimensional space to make it interpretable to the human brain and usable in the statistical tests. This involves 4 steps:\n",
    "1. Specify the parameters for the dimensionality reduction, including the number of components (1) and distance metric we want to use (cosine)\n",
    "2. Normalise the 129D embeddings so that their mean = 0 and variance = 1. This ensures equal weighting of the features\n",
    "3. Carry out the dimensionality reduction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "731b28af",
   "metadata": {
    "metadata": {}
   },
   "outputs": [],
   "source": [
    "# Load the saved acoustic features\n",
    "results = pd.read_csv(os.path.join(os.path.sep, DATA, 'rumble_acoustic_features.csv'))\n",
    "results = results.set_index('filename')\n",
    "\n",
    "# Specify the UMAP parameters\n",
    "results_df = results # name of the dataframe where the embeddings are contained\n",
    "metadata_df = df # name of the dataframe where the call-type labels are contained\n",
    "N_COMP = 1 # select number of dimensions\n",
    "metric = 'cosine' \n",
    "min_dist = 0\n",
    "random_state = 2204\n",
    "n_neighbors = 14\n",
    "damping = 0.95\n",
    "preference=-50\n",
    "max_iter=1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "5e1ba905",
   "metadata": {
    "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>feature_0</th>\n",
       "      <th>feature_1</th>\n",
       "      <th>feature_2</th>\n",
       "      <th>feature_3</th>\n",
       "      <th>feature_4</th>\n",
       "      <th>feature_5</th>\n",
       "      <th>feature_6</th>\n",
       "      <th>feature_7</th>\n",
       "      <th>feature_8</th>\n",
       "      <th>feature_9</th>\n",
       "      <th>...</th>\n",
       "      <th>feature_119</th>\n",
       "      <th>feature_120</th>\n",
       "      <th>feature_121</th>\n",
       "      <th>feature_122</th>\n",
       "      <th>feature_123</th>\n",
       "      <th>feature_124</th>\n",
       "      <th>feature_125</th>\n",
       "      <th>feature_126</th>\n",
       "      <th>feature_127</th>\n",
       "      <th>duration</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filename</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>opp_330_ru.wav</th>\n",
       "      <td>0.367740</td>\n",
       "      <td>-0.223345</td>\n",
       "      <td>0.832650</td>\n",
       "      <td>-0.150125</td>\n",
       "      <td>1.173903</td>\n",
       "      <td>0.017512</td>\n",
       "      <td>-0.658524</td>\n",
       "      <td>0.080850</td>\n",
       "      <td>0.016878</td>\n",
       "      <td>0.084146</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.760345</td>\n",
       "      <td>-0.752949</td>\n",
       "      <td>-0.552275</td>\n",
       "      <td>0.989991</td>\n",
       "      <td>1.264996</td>\n",
       "      <td>0.102087</td>\n",
       "      <td>0.633005</td>\n",
       "      <td>0.931091</td>\n",
       "      <td>-0.583415</td>\n",
       "      <td>-0.615028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_332_ru.wav</th>\n",
       "      <td>0.909980</td>\n",
       "      <td>1.053873</td>\n",
       "      <td>-0.661153</td>\n",
       "      <td>0.305009</td>\n",
       "      <td>0.626951</td>\n",
       "      <td>-0.664588</td>\n",
       "      <td>-0.524754</td>\n",
       "      <td>1.411508</td>\n",
       "      <td>-0.707733</td>\n",
       "      <td>1.634552</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.255880</td>\n",
       "      <td>-1.012490</td>\n",
       "      <td>1.133998</td>\n",
       "      <td>-1.171262</td>\n",
       "      <td>-0.609187</td>\n",
       "      <td>0.097003</td>\n",
       "      <td>0.067107</td>\n",
       "      <td>-1.002129</td>\n",
       "      <td>0.921309</td>\n",
       "      <td>0.163114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru.wav</th>\n",
       "      <td>0.843913</td>\n",
       "      <td>0.592664</td>\n",
       "      <td>-0.618734</td>\n",
       "      <td>0.744534</td>\n",
       "      <td>0.047700</td>\n",
       "      <td>-1.004212</td>\n",
       "      <td>-0.567517</td>\n",
       "      <td>0.924098</td>\n",
       "      <td>-2.049332</td>\n",
       "      <td>0.851215</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.566839</td>\n",
       "      <td>-2.169782</td>\n",
       "      <td>1.106990</td>\n",
       "      <td>-1.996287</td>\n",
       "      <td>-0.103649</td>\n",
       "      <td>0.739693</td>\n",
       "      <td>-0.316070</td>\n",
       "      <td>-1.225245</td>\n",
       "      <td>1.388820</td>\n",
       "      <td>0.173786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru1.wav</th>\n",
       "      <td>1.111779</td>\n",
       "      <td>0.547173</td>\n",
       "      <td>0.236033</td>\n",
       "      <td>0.052802</td>\n",
       "      <td>0.613999</td>\n",
       "      <td>-1.084641</td>\n",
       "      <td>-0.125571</td>\n",
       "      <td>1.088283</td>\n",
       "      <td>-2.586449</td>\n",
       "      <td>0.421471</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.847688</td>\n",
       "      <td>-1.594563</td>\n",
       "      <td>1.084204</td>\n",
       "      <td>-1.184629</td>\n",
       "      <td>-0.100915</td>\n",
       "      <td>1.083589</td>\n",
       "      <td>0.212078</td>\n",
       "      <td>-0.729177</td>\n",
       "      <td>1.541697</td>\n",
       "      <td>0.117184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru2.wav</th>\n",
       "      <td>0.392385</td>\n",
       "      <td>-0.411974</td>\n",
       "      <td>-0.544465</td>\n",
       "      <td>0.388729</td>\n",
       "      <td>0.268408</td>\n",
       "      <td>-1.427294</td>\n",
       "      <td>0.075050</td>\n",
       "      <td>1.052266</td>\n",
       "      <td>-3.467648</td>\n",
       "      <td>0.359834</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.824702</td>\n",
       "      <td>-3.099885</td>\n",
       "      <td>0.859605</td>\n",
       "      <td>-3.132650</td>\n",
       "      <td>-0.682167</td>\n",
       "      <td>0.645300</td>\n",
       "      <td>-1.736565</td>\n",
       "      <td>-1.091032</td>\n",
       "      <td>1.556570</td>\n",
       "      <td>0.014220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru2.wav</th>\n",
       "      <td>1.605449</td>\n",
       "      <td>0.610043</td>\n",
       "      <td>-0.425349</td>\n",
       "      <td>0.430673</td>\n",
       "      <td>2.454805</td>\n",
       "      <td>0.646995</td>\n",
       "      <td>0.120340</td>\n",
       "      <td>1.271033</td>\n",
       "      <td>0.450523</td>\n",
       "      <td>1.002822</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.779263</td>\n",
       "      <td>-1.318837</td>\n",
       "      <td>2.071971</td>\n",
       "      <td>-0.935612</td>\n",
       "      <td>-0.066124</td>\n",
       "      <td>0.357612</td>\n",
       "      <td>0.851405</td>\n",
       "      <td>-0.669603</td>\n",
       "      <td>0.381563</td>\n",
       "      <td>-0.489466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru3.wav</th>\n",
       "      <td>-0.444989</td>\n",
       "      <td>-0.656271</td>\n",
       "      <td>1.560830</td>\n",
       "      <td>0.198704</td>\n",
       "      <td>1.981614</td>\n",
       "      <td>0.226100</td>\n",
       "      <td>0.266449</td>\n",
       "      <td>1.029434</td>\n",
       "      <td>0.485706</td>\n",
       "      <td>-1.019013</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.726758</td>\n",
       "      <td>-0.400408</td>\n",
       "      <td>0.407377</td>\n",
       "      <td>-1.064165</td>\n",
       "      <td>-0.855973</td>\n",
       "      <td>-0.271081</td>\n",
       "      <td>-1.028018</td>\n",
       "      <td>0.732939</td>\n",
       "      <td>-1.336113</td>\n",
       "      <td>-1.171581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_878_ru.wav</th>\n",
       "      <td>-0.387583</td>\n",
       "      <td>1.572086</td>\n",
       "      <td>-0.969803</td>\n",
       "      <td>0.795332</td>\n",
       "      <td>-0.686758</td>\n",
       "      <td>-0.064268</td>\n",
       "      <td>-0.046166</td>\n",
       "      <td>1.207333</td>\n",
       "      <td>0.475972</td>\n",
       "      <td>0.276595</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008997</td>\n",
       "      <td>-0.837589</td>\n",
       "      <td>1.392606</td>\n",
       "      <td>-0.781824</td>\n",
       "      <td>-1.089698</td>\n",
       "      <td>-0.052146</td>\n",
       "      <td>-0.537201</td>\n",
       "      <td>-0.625027</td>\n",
       "      <td>0.575515</td>\n",
       "      <td>0.748905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru.wav</th>\n",
       "      <td>1.448320</td>\n",
       "      <td>0.642797</td>\n",
       "      <td>-0.046650</td>\n",
       "      <td>0.673538</td>\n",
       "      <td>0.621578</td>\n",
       "      <td>1.814995</td>\n",
       "      <td>-0.217499</td>\n",
       "      <td>0.359056</td>\n",
       "      <td>0.227674</td>\n",
       "      <td>0.960846</td>\n",
       "      <td>...</td>\n",
       "      <td>0.809349</td>\n",
       "      <td>0.943266</td>\n",
       "      <td>1.155511</td>\n",
       "      <td>-0.166117</td>\n",
       "      <td>0.725396</td>\n",
       "      <td>0.402391</td>\n",
       "      <td>0.869758</td>\n",
       "      <td>-0.117078</td>\n",
       "      <td>1.136392</td>\n",
       "      <td>-0.274774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru1.wav</th>\n",
       "      <td>0.054951</td>\n",
       "      <td>1.561774</td>\n",
       "      <td>-0.632768</td>\n",
       "      <td>-0.290145</td>\n",
       "      <td>-0.746994</td>\n",
       "      <td>0.044976</td>\n",
       "      <td>-0.042025</td>\n",
       "      <td>1.021629</td>\n",
       "      <td>0.570176</td>\n",
       "      <td>0.628036</td>\n",
       "      <td>...</td>\n",
       "      <td>0.093384</td>\n",
       "      <td>-0.415299</td>\n",
       "      <td>0.932364</td>\n",
       "      <td>0.388994</td>\n",
       "      <td>-0.727754</td>\n",
       "      <td>-0.655401</td>\n",
       "      <td>-0.434483</td>\n",
       "      <td>-0.683749</td>\n",
       "      <td>-0.314089</td>\n",
       "      <td>2.822867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>361 rows × 129 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 feature_0  feature_1  feature_2  feature_3  feature_4  \\\n",
       "filename                                                                 \n",
       "opp_330_ru.wav    0.367740  -0.223345   0.832650  -0.150125   1.173903   \n",
       "opp_332_ru.wav    0.909980   1.053873  -0.661153   0.305009   0.626951   \n",
       "opp_333_ru.wav    0.843913   0.592664  -0.618734   0.744534   0.047700   \n",
       "opp_333_ru1.wav   1.111779   0.547173   0.236033   0.052802   0.613999   \n",
       "opp_333_ru2.wav   0.392385  -0.411974  -0.544465   0.388729   0.268408   \n",
       "...                    ...        ...        ...        ...        ...   \n",
       "opp_876_ru2.wav   1.605449   0.610043  -0.425349   0.430673   2.454805   \n",
       "opp_876_ru3.wav  -0.444989  -0.656271   1.560830   0.198704   1.981614   \n",
       "opp_878_ru.wav   -0.387583   1.572086  -0.969803   0.795332  -0.686758   \n",
       "opp_881_ru.wav    1.448320   0.642797  -0.046650   0.673538   0.621578   \n",
       "opp_881_ru1.wav   0.054951   1.561774  -0.632768  -0.290145  -0.746994   \n",
       "\n",
       "                 feature_5  feature_6  feature_7  feature_8  feature_9  ...  \\\n",
       "filename                                                                ...   \n",
       "opp_330_ru.wav    0.017512  -0.658524   0.080850   0.016878   0.084146  ...   \n",
       "opp_332_ru.wav   -0.664588  -0.524754   1.411508  -0.707733   1.634552  ...   \n",
       "opp_333_ru.wav   -1.004212  -0.567517   0.924098  -2.049332   0.851215  ...   \n",
       "opp_333_ru1.wav  -1.084641  -0.125571   1.088283  -2.586449   0.421471  ...   \n",
       "opp_333_ru2.wav  -1.427294   0.075050   1.052266  -3.467648   0.359834  ...   \n",
       "...                    ...        ...        ...        ...        ...  ...   \n",
       "opp_876_ru2.wav   0.646995   0.120340   1.271033   0.450523   1.002822  ...   \n",
       "opp_876_ru3.wav   0.226100   0.266449   1.029434   0.485706  -1.019013  ...   \n",
       "opp_878_ru.wav   -0.064268  -0.046166   1.207333   0.475972   0.276595  ...   \n",
       "opp_881_ru.wav    1.814995  -0.217499   0.359056   0.227674   0.960846  ...   \n",
       "opp_881_ru1.wav   0.044976  -0.042025   1.021629   0.570176   0.628036  ...   \n",
       "\n",
       "                 feature_119  feature_120  feature_121  feature_122  \\\n",
       "filename                                                              \n",
       "opp_330_ru.wav     -0.760345    -0.752949    -0.552275     0.989991   \n",
       "opp_332_ru.wav     -0.255880    -1.012490     1.133998    -1.171262   \n",
       "opp_333_ru.wav     -1.566839    -2.169782     1.106990    -1.996287   \n",
       "opp_333_ru1.wav    -0.847688    -1.594563     1.084204    -1.184629   \n",
       "opp_333_ru2.wav    -1.824702    -3.099885     0.859605    -3.132650   \n",
       "...                      ...          ...          ...          ...   \n",
       "opp_876_ru2.wav    -0.779263    -1.318837     2.071971    -0.935612   \n",
       "opp_876_ru3.wav    -1.726758    -0.400408     0.407377    -1.064165   \n",
       "opp_878_ru.wav     -0.008997    -0.837589     1.392606    -0.781824   \n",
       "opp_881_ru.wav      0.809349     0.943266     1.155511    -0.166117   \n",
       "opp_881_ru1.wav     0.093384    -0.415299     0.932364     0.388994   \n",
       "\n",
       "                 feature_123  feature_124  feature_125  feature_126  \\\n",
       "filename                                                              \n",
       "opp_330_ru.wav      1.264996     0.102087     0.633005     0.931091   \n",
       "opp_332_ru.wav     -0.609187     0.097003     0.067107    -1.002129   \n",
       "opp_333_ru.wav     -0.103649     0.739693    -0.316070    -1.225245   \n",
       "opp_333_ru1.wav    -0.100915     1.083589     0.212078    -0.729177   \n",
       "opp_333_ru2.wav    -0.682167     0.645300    -1.736565    -1.091032   \n",
       "...                      ...          ...          ...          ...   \n",
       "opp_876_ru2.wav    -0.066124     0.357612     0.851405    -0.669603   \n",
       "opp_876_ru3.wav    -0.855973    -0.271081    -1.028018     0.732939   \n",
       "opp_878_ru.wav     -1.089698    -0.052146    -0.537201    -0.625027   \n",
       "opp_881_ru.wav      0.725396     0.402391     0.869758    -0.117078   \n",
       "opp_881_ru1.wav    -0.727754    -0.655401    -0.434483    -0.683749   \n",
       "\n",
       "                 feature_127  duration  \n",
       "filename                                \n",
       "opp_330_ru.wav     -0.583415 -0.615028  \n",
       "opp_332_ru.wav      0.921309  0.163114  \n",
       "opp_333_ru.wav      1.388820  0.173786  \n",
       "opp_333_ru1.wav     1.541697  0.117184  \n",
       "opp_333_ru2.wav     1.556570  0.014220  \n",
       "...                      ...       ...  \n",
       "opp_876_ru2.wav     0.381563 -0.489466  \n",
       "opp_876_ru3.wav    -1.336113 -1.171581  \n",
       "opp_878_ru.wav      0.575515  0.748905  \n",
       "opp_881_ru.wav      1.136392 -0.274774  \n",
       "opp_881_ru1.wav    -0.314089  2.822867  \n",
       "\n",
       "[361 rows x 129 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "normalised = scaler.fit(results_df)\n",
    "normalised_df = pd.DataFrame(scaler.transform(results_df), columns = results_df.columns, index = results_df.index)\n",
    "normalised_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "4d992990",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.0288175],\n",
       "       [14.946022 ],\n",
       "       [15.191775 ],\n",
       "       [15.021379 ],\n",
       "       [17.173191 ],\n",
       "       [ 3.1475248],\n",
       "       [18.79924  ],\n",
       "       [11.420613 ],\n",
       "       [17.78729  ],\n",
       "       [16.249632 ],\n",
       "       [ 3.4709253],\n",
       "       [ 6.964533 ],\n",
       "       [ 1.6078218],\n",
       "       [13.129976 ],\n",
       "       [ 2.7225595],\n",
       "       [ 2.3866155],\n",
       "       [17.483595 ],\n",
       "       [ 1.1246002],\n",
       "       [ 9.020732 ],\n",
       "       [18.564669 ],\n",
       "       [ 3.8435721],\n",
       "       [ 2.0074365],\n",
       "       [ 3.8003943],\n",
       "       [ 6.3379855],\n",
       "       [13.537108 ],\n",
       "       [17.345644 ],\n",
       "       [14.614272 ],\n",
       "       [ 3.2351794],\n",
       "       [15.976224 ],\n",
       "       [ 4.885398 ],\n",
       "       [ 5.134397 ],\n",
       "       [ 3.33689  ],\n",
       "       [ 1.1537197],\n",
       "       [ 1.9993926],\n",
       "       [14.41291  ],\n",
       "       [ 8.844041 ],\n",
       "       [ 4.506319 ],\n",
       "       [13.765831 ],\n",
       "       [13.279771 ],\n",
       "       [16.39357  ],\n",
       "       [ 3.4386396],\n",
       "       [ 9.045164 ],\n",
       "       [12.832753 ],\n",
       "       [ 4.5001917],\n",
       "       [ 3.6883044],\n",
       "       [ 4.592966 ],\n",
       "       [ 7.444811 ],\n",
       "       [11.674347 ],\n",
       "       [ 2.6956627],\n",
       "       [19.122408 ],\n",
       "       [14.233299 ],\n",
       "       [14.068602 ],\n",
       "       [12.173565 ],\n",
       "       [ 6.2221675],\n",
       "       [14.311236 ],\n",
       "       [ 4.3693886],\n",
       "       [16.22996  ],\n",
       "       [ 6.4646745],\n",
       "       [16.745245 ],\n",
       "       [ 2.28768  ],\n",
       "       [14.315932 ],\n",
       "       [ 1.224709 ],\n",
       "       [ 8.436944 ],\n",
       "       [14.252363 ],\n",
       "       [18.095352 ],\n",
       "       [ 6.2119665],\n",
       "       [ 9.087214 ],\n",
       "       [ 9.070134 ],\n",
       "       [16.434826 ],\n",
       "       [11.5866585],\n",
       "       [ 5.1892643],\n",
       "       [14.651229 ],\n",
       "       [18.219921 ],\n",
       "       [ 5.8630147],\n",
       "       [ 3.3365324],\n",
       "       [13.105484 ],\n",
       "       [16.948528 ],\n",
       "       [15.29529  ],\n",
       "       [ 3.9384663],\n",
       "       [13.222961 ],\n",
       "       [12.572471 ],\n",
       "       [15.924253 ],\n",
       "       [ 4.246686 ],\n",
       "       [ 4.286002 ],\n",
       "       [13.545618 ],\n",
       "       [ 5.36128  ],\n",
       "       [ 1.0178692],\n",
       "       [16.786757 ],\n",
       "       [18.274008 ],\n",
       "       [ 4.0195875],\n",
       "       [17.98662  ],\n",
       "       [14.485663 ],\n",
       "       [13.403964 ],\n",
       "       [ 3.0395987],\n",
       "       [ 2.9263268],\n",
       "       [18.648777 ],\n",
       "       [ 4.4064593],\n",
       "       [ 4.610498 ],\n",
       "       [ 5.805245 ],\n",
       "       [14.861315 ],\n",
       "       [18.64726  ],\n",
       "       [ 2.436003 ],\n",
       "       [ 2.3827603],\n",
       "       [12.556837 ],\n",
       "       [ 7.083992 ],\n",
       "       [15.360732 ],\n",
       "       [ 7.3999667],\n",
       "       [ 7.315985 ],\n",
       "       [ 7.425512 ],\n",
       "       [15.068827 ],\n",
       "       [ 8.368025 ],\n",
       "       [ 8.789578 ],\n",
       "       [ 4.3116455],\n",
       "       [15.781034 ],\n",
       "       [ 3.743081 ],\n",
       "       [ 6.5782595],\n",
       "       [ 8.796332 ],\n",
       "       [ 9.028795 ],\n",
       "       [13.239668 ],\n",
       "       [ 8.88291  ],\n",
       "       [11.964361 ],\n",
       "       [13.450098 ],\n",
       "       [11.189086 ],\n",
       "       [ 7.918652 ],\n",
       "       [ 7.378984 ],\n",
       "       [18.295729 ],\n",
       "       [16.25183  ],\n",
       "       [18.493938 ],\n",
       "       [ 9.007003 ],\n",
       "       [ 2.027731 ],\n",
       "       [ 6.2257943],\n",
       "       [15.835518 ],\n",
       "       [17.104143 ],\n",
       "       [ 7.5434813],\n",
       "       [ 6.1982374],\n",
       "       [ 6.9633546],\n",
       "       [ 9.048421 ],\n",
       "       [ 3.6622975],\n",
       "       [ 3.9577856],\n",
       "       [13.387288 ],\n",
       "       [ 3.891381 ],\n",
       "       [15.625132 ],\n",
       "       [ 8.031063 ],\n",
       "       [13.652497 ],\n",
       "       [ 8.643686 ],\n",
       "       [13.870045 ],\n",
       "       [14.180215 ],\n",
       "       [16.302988 ],\n",
       "       [16.540117 ],\n",
       "       [14.485217 ],\n",
       "       [ 6.6950455],\n",
       "       [17.492126 ],\n",
       "       [ 9.074532 ],\n",
       "       [18.508345 ],\n",
       "       [14.903531 ],\n",
       "       [15.889104 ],\n",
       "       [ 5.2100005],\n",
       "       [ 2.599582 ],\n",
       "       [ 7.537816 ],\n",
       "       [15.410997 ],\n",
       "       [ 6.816221 ],\n",
       "       [15.869378 ],\n",
       "       [ 5.657658 ],\n",
       "       [ 6.9668727],\n",
       "       [ 2.7617998],\n",
       "       [14.134973 ],\n",
       "       [ 1.1844656],\n",
       "       [ 6.668668 ],\n",
       "       [ 2.8276029],\n",
       "       [ 2.9590888],\n",
       "       [ 4.0855017],\n",
       "       [13.938861 ],\n",
       "       [11.306075 ],\n",
       "       [14.067231 ],\n",
       "       [ 4.300688 ],\n",
       "       [ 4.0732574],\n",
       "       [ 1.8178668],\n",
       "       [ 2.3538008],\n",
       "       [ 7.634607 ],\n",
       "       [ 1.8832052],\n",
       "       [ 5.3850994],\n",
       "       [15.762024 ],\n",
       "       [ 2.2895741],\n",
       "       [17.724874 ],\n",
       "       [ 7.580415 ],\n",
       "       [ 4.761821 ],\n",
       "       [ 4.8698125],\n",
       "       [ 4.026668 ],\n",
       "       [ 5.2330794],\n",
       "       [ 3.2842052],\n",
       "       [ 3.2325628],\n",
       "       [ 4.872143 ],\n",
       "       [14.635117 ],\n",
       "       [ 5.356002 ],\n",
       "       [ 2.5929232],\n",
       "       [ 3.7856965],\n",
       "       [ 4.1394453],\n",
       "       [18.006842 ],\n",
       "       [15.514627 ],\n",
       "       [17.431053 ],\n",
       "       [17.88976  ],\n",
       "       [ 0.8833438],\n",
       "       [16.412025 ],\n",
       "       [16.68175  ],\n",
       "       [13.677341 ],\n",
       "       [15.824139 ],\n",
       "       [ 2.1055715],\n",
       "       [16.71301  ],\n",
       "       [ 1.3466902],\n",
       "       [ 1.2020721],\n",
       "       [14.530116 ],\n",
       "       [16.79292  ],\n",
       "       [ 1.4041501],\n",
       "       [11.317366 ],\n",
       "       [11.889676 ],\n",
       "       [ 2.193349 ],\n",
       "       [ 1.8402903],\n",
       "       [ 1.9771032],\n",
       "       [ 8.4078245],\n",
       "       [ 1.4513066],\n",
       "       [14.6373   ],\n",
       "       [ 2.0334814],\n",
       "       [12.047364 ],\n",
       "       [ 2.3901303],\n",
       "       [17.534868 ],\n",
       "       [ 7.463096 ],\n",
       "       [ 7.084468 ],\n",
       "       [ 2.3517597],\n",
       "       [13.830664 ],\n",
       "       [ 9.575135 ],\n",
       "       [ 1.4432815],\n",
       "       [ 2.5266926],\n",
       "       [ 1.6613679],\n",
       "       [ 4.169117 ],\n",
       "       [ 2.7977197],\n",
       "       [ 2.3341308],\n",
       "       [13.184363 ],\n",
       "       [17.390068 ],\n",
       "       [18.453869 ],\n",
       "       [17.13461  ],\n",
       "       [18.158903 ],\n",
       "       [17.92133  ],\n",
       "       [ 4.8944287],\n",
       "       [ 0.9538778],\n",
       "       [ 6.168793 ],\n",
       "       [ 8.722029 ],\n",
       "       [ 8.946    ],\n",
       "       [17.224836 ],\n",
       "       [ 4.111992 ],\n",
       "       [ 7.6964893],\n",
       "       [ 7.3392315],\n",
       "       [ 7.1314545],\n",
       "       [ 1.73699  ],\n",
       "       [ 1.0793216],\n",
       "       [ 0.7527614],\n",
       "       [12.060542 ],\n",
       "       [13.257414 ],\n",
       "       [ 3.2929013],\n",
       "       [ 1.1446939],\n",
       "       [14.611122 ],\n",
       "       [18.893318 ],\n",
       "       [ 1.5073241],\n",
       "       [ 2.381714 ],\n",
       "       [15.875916 ],\n",
       "       [ 8.069477 ],\n",
       "       [ 6.3322606],\n",
       "       [ 4.2017612],\n",
       "       [16.250257 ],\n",
       "       [ 1.8190522],\n",
       "       [ 5.897268 ],\n",
       "       [12.664611 ],\n",
       "       [17.273323 ],\n",
       "       [17.325243 ],\n",
       "       [17.099236 ],\n",
       "       [13.854858 ],\n",
       "       [13.895797 ],\n",
       "       [17.256529 ],\n",
       "       [13.827303 ],\n",
       "       [15.694302 ],\n",
       "       [ 3.780742 ],\n",
       "       [12.990944 ],\n",
       "       [ 2.9187553],\n",
       "       [15.061332 ],\n",
       "       [18.16054  ],\n",
       "       [17.855984 ],\n",
       "       [ 3.513676 ],\n",
       "       [15.303933 ],\n",
       "       [18.55122  ],\n",
       "       [18.594929 ],\n",
       "       [18.496597 ],\n",
       "       [18.224379 ],\n",
       "       [ 8.974836 ],\n",
       "       [11.881752 ],\n",
       "       [ 8.170256 ],\n",
       "       [ 3.1662002],\n",
       "       [ 1.2028409],\n",
       "       [ 7.032901 ],\n",
       "       [17.295982 ],\n",
       "       [18.536386 ],\n",
       "       [17.001873 ],\n",
       "       [11.642115 ],\n",
       "       [12.297052 ],\n",
       "       [11.629154 ],\n",
       "       [16.599895 ],\n",
       "       [17.680588 ],\n",
       "       [17.890812 ],\n",
       "       [ 6.9301825],\n",
       "       [ 3.4615197],\n",
       "       [15.9007225],\n",
       "       [ 3.6555145],\n",
       "       [ 3.0719619],\n",
       "       [15.856273 ],\n",
       "       [13.198028 ],\n",
       "       [11.74396  ],\n",
       "       [ 9.100224 ],\n",
       "       [ 6.2017436],\n",
       "       [ 9.12566  ],\n",
       "       [12.367805 ],\n",
       "       [ 1.0369242],\n",
       "       [ 5.9473586],\n",
       "       [ 7.358275 ],\n",
       "       [16.793232 ],\n",
       "       [ 7.3347983],\n",
       "       [15.680108 ],\n",
       "       [ 9.39491  ],\n",
       "       [17.38747  ],\n",
       "       [17.609205 ],\n",
       "       [ 2.3670428],\n",
       "       [19.119455 ],\n",
       "       [ 2.1272314],\n",
       "       [18.381212 ],\n",
       "       [ 4.9445314],\n",
       "       [ 4.0733347],\n",
       "       [ 3.9493043],\n",
       "       [16.47981  ],\n",
       "       [ 6.3340454],\n",
       "       [ 6.656316 ],\n",
       "       [12.998222 ],\n",
       "       [ 4.4229994],\n",
       "       [ 4.543585 ],\n",
       "       [14.124364 ],\n",
       "       [ 1.36676  ],\n",
       "       [19.1391   ],\n",
       "       [16.403057 ],\n",
       "       [ 4.2140346],\n",
       "       [ 4.0970826],\n",
       "       [ 2.60716  ],\n",
       "       [18.67245  ],\n",
       "       [ 7.080992 ],\n",
       "       [13.366004 ],\n",
       "       [13.750017 ],\n",
       "       [ 8.698614 ],\n",
       "       [18.240885 ],\n",
       "       [17.02606  ],\n",
       "       [ 5.960624 ],\n",
       "       [14.27097  ],\n",
       "       [14.218194 ],\n",
       "       [13.152391 ],\n",
       "       [19.041962 ],\n",
       "       [14.154239 ],\n",
       "       [19.145327 ]], dtype=float32)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Function to fit UMAP and return the embedding\n",
    "\n",
    "def fit_umap(transformed_df, N_COMP, metric, min_dist, n_neighbors, random_state):\n",
    "    reducer = umap.UMAP(n_components=N_COMP, metric=metric, min_dist=min_dist, n_neighbors = n_neighbors, random_state=random_state)\n",
    "    return reducer.fit_transform(transformed_df)\n",
    "\n",
    "# Fit UMAP and obtain embeddings\n",
    "embeddings = fit_umap(normalised_df, N_COMP, metric, min_dist, n_neighbors, random_state)\n",
    "embeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7420cd3a",
   "metadata": {},
   "source": [
    "### 4. Unsupervised Clustering: Rumble Sub-type identification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "4707db7c",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 1, 1, 1, 0, 3, 5, 7, 5, 0, 3, 2, 4, 7, 4, 4, 5, 4, 6, 5, 3, 4,\n",
       "       3, 2, 1, 5, 1, 3, 0, 3, 3, 3, 4, 4, 1, 6, 3, 1, 1, 0, 3, 6, 7, 3,\n",
       "       3, 3, 2, 7, 4, 5, 1, 1, 7, 2, 1, 3, 0, 2, 0, 4, 1, 4, 6, 1, 5, 2,\n",
       "       6, 6, 0, 7, 3, 1, 5, 2, 3, 7, 0, 0, 3, 7, 7, 0, 3, 3, 1, 3, 4, 0,\n",
       "       5, 3, 5, 1, 1, 4, 4, 5, 3, 3, 2, 1, 5, 4, 4, 7, 2, 0, 2, 2, 2, 1,\n",
       "       6, 6, 3, 0, 3, 2, 6, 6, 1, 6, 7, 1, 7, 6, 2, 5, 0, 5, 6, 4, 2, 0,\n",
       "       0, 2, 2, 2, 6, 3, 3, 1, 3, 0, 6, 1, 6, 1, 1, 0, 0, 1, 2, 5, 6, 5,\n",
       "       1, 0, 3, 4, 2, 0, 2, 0, 2, 2, 4, 1, 4, 2, 4, 4, 3, 1, 7, 1, 3, 3,\n",
       "       4, 4, 2, 4, 3, 0, 4, 5, 2, 3, 3, 3, 3, 3, 3, 3, 1, 3, 4, 3, 3, 5,\n",
       "       0, 5, 5, 4, 0, 0, 1, 0, 4, 0, 4, 4, 1, 0, 4, 7, 7, 4, 4, 4, 6, 4,\n",
       "       1, 4, 7, 4, 5, 2, 2, 4, 1, 6, 4, 4, 4, 3, 4, 4, 7, 5, 5, 0, 5, 5,\n",
       "       3, 4, 2, 6, 6, 5, 3, 2, 2, 2, 4, 4, 4, 7, 1, 3, 4, 1, 5, 4, 4, 0,\n",
       "       6, 2, 3, 0, 4, 2, 7, 5, 5, 0, 1, 1, 5, 1, 0, 3, 7, 4, 1, 5, 5, 3,\n",
       "       0, 5, 5, 5, 5, 6, 7, 6, 3, 4, 2, 5, 5, 0, 7, 7, 7, 0, 5, 5, 2, 3,\n",
       "       0, 3, 3, 0, 7, 7, 6, 2, 6, 7, 4, 2, 2, 0, 2, 0, 6, 5, 5, 4, 5, 4,\n",
       "       5, 3, 3, 3, 0, 2, 2, 7, 3, 3, 1, 4, 5, 0, 3, 3, 4, 5, 2, 1, 1, 6,\n",
       "       5, 0, 2, 1, 1, 7, 5, 1, 5])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def affinity_propagation_clustering(features_standardized, damping=0.5, preference= None, random_state = None):\n",
    "    # Perform affinity propagation clustering\n",
    "    affinity_propagation = AffinityPropagation(damping=damping, preference=preference, random_state = random_state)\n",
    "    cluster_labels = affinity_propagation.fit_predict(embeddings)\n",
    "    return cluster_labels\n",
    "\n",
    "# Perform affinity propagation clustering \n",
    "cluster_labels = None\n",
    "cluster_labels = affinity_propagation_clustering(embeddings, damping=damping, preference=preference, random_state=random_state)\n",
    "cluster_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "f6a36be7",
   "metadata": {
    "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>UMAP1</th>\n",
       "      <th>Cluster</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filename</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>opp_330_ru.wav</th>\n",
       "      <td>1.028818</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_332_ru.wav</th>\n",
       "      <td>14.946022</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru.wav</th>\n",
       "      <td>15.191775</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru1.wav</th>\n",
       "      <td>15.021379</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru2.wav</th>\n",
       "      <td>17.173191</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru2.wav</th>\n",
       "      <td>14.218194</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru3.wav</th>\n",
       "      <td>13.152391</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_878_ru.wav</th>\n",
       "      <td>19.041962</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru.wav</th>\n",
       "      <td>14.154239</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru1.wav</th>\n",
       "      <td>19.145327</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>361 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     UMAP1  Cluster\n",
       "filename                           \n",
       "opp_330_ru.wav    1.028818        4\n",
       "opp_332_ru.wav   14.946022        1\n",
       "opp_333_ru.wav   15.191775        1\n",
       "opp_333_ru1.wav  15.021379        1\n",
       "opp_333_ru2.wav  17.173191        0\n",
       "...                    ...      ...\n",
       "opp_876_ru2.wav  14.218194        1\n",
       "opp_876_ru3.wav  13.152391        7\n",
       "opp_878_ru.wav   19.041962        5\n",
       "opp_881_ru.wav   14.154239        1\n",
       "opp_881_ru1.wav  19.145327        5\n",
       "\n",
       "[361 rows x 2 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a DataFrame with UMAP coordinates and add the 'Cluster' column\n",
    "umap_cluster_results = pd.DataFrame(\n",
    "    {\n",
    "        \"filename\": normalised_df.index,\n",
    "        **{f\"UMAP{i + 1}\": embeddings[:, i] for i in range(N_COMP)},\n",
    "        \"Cluster\": cluster_labels  # Add the 'Cluster' column\n",
    "    }\n",
    ")\n",
    "\n",
    "# Set 'filename' as the index\n",
    "umap_cluster_results = umap_cluster_results.set_index(\"filename\")\n",
    "umap_cluster_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "60ffdba0",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set the style of the visualisation\n",
    "sns.set(style=\"whitegrid\")\n",
    "\n",
    "# Create a boxplot\n",
    "plt.figure(figsize=(10, 6))\n",
    "sns.boxplot(x='Cluster', y='UMAP1', data=umap_cluster_results, hue = 'Cluster')\n",
    "\n",
    "# Set the title and labels\n",
    "plt.title('Boxplot of UMAP1 by Cluster')\n",
    "plt.xlabel('Cluster')\n",
    "plt.ylabel('UMAP1')\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "bcc8036c",
   "metadata": {
    "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>UMAP1</th>\n",
       "      <th>Cluster</th>\n",
       "      <th>start_time</th>\n",
       "      <th>end_time</th>\n",
       "      <th>duration</th>\n",
       "      <th>low_freq</th>\n",
       "      <th>high_freq</th>\n",
       "      <th>call-type</th>\n",
       "      <th>Final_Category</th>\n",
       "      <th>Distress</th>\n",
       "      <th>Age</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age_Sex</th>\n",
       "      <th>audio_filepaths</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filename</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>opp_330_ru.wav</th>\n",
       "      <td>1.028818</td>\n",
       "      <td>4</td>\n",
       "      <td>1.628424</td>\n",
       "      <td>3.790237</td>\n",
       "      <td>2.161813</td>\n",
       "      <td>1.220703</td>\n",
       "      <td>326.464844</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Nursing</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Juvenile</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_332_ru.wav</th>\n",
       "      <td>14.946022</td>\n",
       "      <td>1</td>\n",
       "      <td>2.229882</td>\n",
       "      <td>5.752067</td>\n",
       "      <td>3.522186</td>\n",
       "      <td>1.635742</td>\n",
       "      <td>197.631836</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Logistics</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru.wav</th>\n",
       "      <td>15.191775</td>\n",
       "      <td>1</td>\n",
       "      <td>3.098901</td>\n",
       "      <td>6.639745</td>\n",
       "      <td>3.540844</td>\n",
       "      <td>0.756836</td>\n",
       "      <td>197.167969</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru1.wav</th>\n",
       "      <td>15.021379</td>\n",
       "      <td>1</td>\n",
       "      <td>1.130586</td>\n",
       "      <td>4.572474</td>\n",
       "      <td>3.441889</td>\n",
       "      <td>0.048828</td>\n",
       "      <td>299.731445</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_333_ru2.wav</th>\n",
       "      <td>17.173191</td>\n",
       "      <td>0</td>\n",
       "      <td>2.029501</td>\n",
       "      <td>5.291385</td>\n",
       "      <td>3.261884</td>\n",
       "      <td>1.196289</td>\n",
       "      <td>324.829102</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru2.wav</th>\n",
       "      <td>14.218194</td>\n",
       "      <td>1</td>\n",
       "      <td>1.628046</td>\n",
       "      <td>4.009370</td>\n",
       "      <td>2.381324</td>\n",
       "      <td>0.463867</td>\n",
       "      <td>287.939453</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Separation</td>\n",
       "      <td>Distress</td>\n",
       "      <td>Juvenile</td>\n",
       "      <td>Male</td>\n",
       "      <td>Juvenile Male</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_876_ru3.wav</th>\n",
       "      <td>13.152391</td>\n",
       "      <td>7</td>\n",
       "      <td>2.145454</td>\n",
       "      <td>3.334282</td>\n",
       "      <td>1.188828</td>\n",
       "      <td>10.253906</td>\n",
       "      <td>343.530273</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Separation</td>\n",
       "      <td>Distress</td>\n",
       "      <td>Juvenile</td>\n",
       "      <td>Male</td>\n",
       "      <td>Juvenile Male</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_878_ru.wav</th>\n",
       "      <td>19.041962</td>\n",
       "      <td>5</td>\n",
       "      <td>3.575788</td>\n",
       "      <td>8.122073</td>\n",
       "      <td>4.546284</td>\n",
       "      <td>1.391602</td>\n",
       "      <td>244.506836</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Affiliation</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru.wav</th>\n",
       "      <td>14.154239</td>\n",
       "      <td>1</td>\n",
       "      <td>2.392540</td>\n",
       "      <td>5.149196</td>\n",
       "      <td>2.756656</td>\n",
       "      <td>0.463867</td>\n",
       "      <td>248.608398</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Affiliation</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Sub-Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Sub-Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opp_881_ru1.wav</th>\n",
       "      <td>19.145327</td>\n",
       "      <td>5</td>\n",
       "      <td>0.719855</td>\n",
       "      <td>8.891908</td>\n",
       "      <td>8.172052</td>\n",
       "      <td>0.585938</td>\n",
       "      <td>291.552734</td>\n",
       "      <td>rumble</td>\n",
       "      <td>Affiliation</td>\n",
       "      <td>Not Applicable</td>\n",
       "      <td>Sub-Adult</td>\n",
       "      <td>Female</td>\n",
       "      <td>Sub-Adult Female</td>\n",
       "      <td>/home/alastair/forestdata/alastair/temp/elepha...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>361 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     UMAP1  Cluster  start_time  end_time  duration  \\\n",
       "filename                                                              \n",
       "opp_330_ru.wav    1.028818        4    1.628424  3.790237  2.161813   \n",
       "opp_332_ru.wav   14.946022        1    2.229882  5.752067  3.522186   \n",
       "opp_333_ru.wav   15.191775        1    3.098901  6.639745  3.540844   \n",
       "opp_333_ru1.wav  15.021379        1    1.130586  4.572474  3.441889   \n",
       "opp_333_ru2.wav  17.173191        0    2.029501  5.291385  3.261884   \n",
       "...                    ...      ...         ...       ...       ...   \n",
       "opp_876_ru2.wav  14.218194        1    1.628046  4.009370  2.381324   \n",
       "opp_876_ru3.wav  13.152391        7    2.145454  3.334282  1.188828   \n",
       "opp_878_ru.wav   19.041962        5    3.575788  8.122073  4.546284   \n",
       "opp_881_ru.wav   14.154239        1    2.392540  5.149196  2.756656   \n",
       "opp_881_ru1.wav  19.145327        5    0.719855  8.891908  8.172052   \n",
       "\n",
       "                  low_freq   high_freq call-type Final_Category  \\\n",
       "filename                                                          \n",
       "opp_330_ru.wav    1.220703  326.464844    rumble        Nursing   \n",
       "opp_332_ru.wav    1.635742  197.631836    rumble      Logistics   \n",
       "opp_333_ru.wav    0.756836  197.167969    rumble        Unknown   \n",
       "opp_333_ru1.wav   0.048828  299.731445    rumble        Unknown   \n",
       "opp_333_ru2.wav   1.196289  324.829102    rumble        Unknown   \n",
       "...                    ...         ...       ...            ...   \n",
       "opp_876_ru2.wav   0.463867  287.939453    rumble     Separation   \n",
       "opp_876_ru3.wav  10.253906  343.530273    rumble     Separation   \n",
       "opp_878_ru.wav    1.391602  244.506836    rumble    Affiliation   \n",
       "opp_881_ru.wav    0.463867  248.608398    rumble    Affiliation   \n",
       "opp_881_ru1.wav   0.585938  291.552734    rumble    Affiliation   \n",
       "\n",
       "                       Distress        Age      Sex           Age_Sex  \\\n",
       "filename                                                                \n",
       "opp_330_ru.wav   Not Applicable   Juvenile  Unknown           Unknown   \n",
       "opp_332_ru.wav   Not Applicable      Adult   Female      Adult Female   \n",
       "opp_333_ru.wav   Not Applicable      Adult   Female      Adult Female   \n",
       "opp_333_ru1.wav  Not Applicable      Adult   Female      Adult Female   \n",
       "opp_333_ru2.wav  Not Applicable      Adult   Female      Adult Female   \n",
       "...                         ...        ...      ...               ...   \n",
       "opp_876_ru2.wav        Distress   Juvenile     Male     Juvenile Male   \n",
       "opp_876_ru3.wav        Distress   Juvenile     Male     Juvenile Male   \n",
       "opp_878_ru.wav   Not Applicable      Adult   Female      Adult Female   \n",
       "opp_881_ru.wav   Not Applicable  Sub-Adult   Female  Sub-Adult Female   \n",
       "opp_881_ru1.wav  Not Applicable  Sub-Adult   Female  Sub-Adult Female   \n",
       "\n",
       "                                                   audio_filepaths  \n",
       "filename                                                            \n",
       "opp_330_ru.wav   /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_332_ru.wav   /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_333_ru.wav   /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_333_ru1.wav  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_333_ru2.wav  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "...                                                            ...  \n",
       "opp_876_ru2.wav  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_876_ru3.wav  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_878_ru.wav   /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_881_ru.wav   /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "opp_881_ru1.wav  /home/alastair/forestdata/alastair/temp/elepha...  \n",
       "\n",
       "[361 rows x 14 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cluster_df = pd.merge(umap_cluster_results, metadata_df, how='left', left_index=True, right_on='filename')\n",
    "cluster_df = cluster_df.set_index(\"filename\")\n",
    "cluster_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f0aa0b8",
   "metadata": {},
   "source": [
    "We can visually assess the level of separation achieved by the unsupervised methods but this needs to be accompanied by a statistical analysis of its performance. We will do this by calculating the overall silhouette score for the data and the silhouette score per call-type.\n",
    "\n",
    "The silhouete score is a mathematical representation of the tightness and separation of each cluster (Rousseeuw, 1987). It is calculated as the mean Euclidean distance between data points within a cluster (a) and the mean Euclidean distance to points in the nearest cluster (Rousseeuw, 1987).\n",
    "\n",
    "Typically, silhouette scores >=0.5 show evidence for clustering, while those >=0.7 show strong evidence for clustering."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "6b0b4f2c",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average Silhouette Score: 0.585316\n"
     ]
    },
    {
     "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>Cluster</th>\n",
       "      <th>Silhouette_Score</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0.627086</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.547502</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0.535080</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.573070</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.595831</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>7</td>\n",
       "      <td>0.498397</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2</td>\n",
       "      <td>0.581006</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>6</td>\n",
       "      <td>0.744789</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Cluster  Silhouette_Score  Count\n",
       "0        4          0.627086     59\n",
       "1        1          0.547502     47\n",
       "2        0          0.535080     44\n",
       "3        3          0.573070     61\n",
       "4        5          0.595831     49\n",
       "5        7          0.498397     29\n",
       "6        2          0.581006     44\n",
       "7        6          0.744789     28"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Define data and embeddings\n",
    "embeddings = cluster_df[['UMAP1']].values\n",
    "labels = cluster_df['Cluster'].values\n",
    "\n",
    "# Calculate silhouette scores\n",
    "silhouette_avg = silhouette_score(embeddings, labels)\n",
    "print(\"Average Silhouette Score:\", silhouette_avg)\n",
    "\n",
    "# Calculate silhouette scores per sample\n",
    "sample_silhouette_values = silhouette_samples(embeddings, labels)\n",
    "\n",
    "# Add silhouette scores to 'cluster_df'\n",
    "cluster_df['Cluster_Silhouette_Score'] = sample_silhouette_values\n",
    "\n",
    "# Calculate count per cluster\n",
    "cluster_counts = cluster_df['Cluster'].value_counts()\n",
    "\n",
    "# Create a list to store silhouette scores and counts per cluster\n",
    "cluster_results = []\n",
    "\n",
    "# Iterate over each unique cluster label\n",
    "unique_clusters = cluster_df['Cluster'].unique()\n",
    "for cluster_label in unique_clusters:\n",
    "    # Filter samples belonging to the current cluster\n",
    "    cluster_mask = cluster_df['Cluster'] == cluster_label\n",
    "    cluster_count = cluster_counts[cluster_label]\n",
    "    \n",
    "    # Retrieve silhouette scores for samples in the current cluster\n",
    "    cluster_silhouette_values = sample_silhouette_values[cluster_mask]\n",
    "    \n",
    "    # Calculate average silhouette score for the current cluster\n",
    "    cluster_silhouette_avg = cluster_silhouette_values.mean()\n",
    "    \n",
    "    # Store cluster label, silhouette score, and count in the list\n",
    "    cluster_results.append({\n",
    "        'Cluster': cluster_label, \n",
    "        'Silhouette_Score': cluster_silhouette_avg, \n",
    "        'Count': cluster_count\n",
    "    })\n",
    "\n",
    "# Convert list to DataFrame\n",
    "cluster_results_df = pd.DataFrame(cluster_results)\n",
    "cluster_results_df\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81ac755b",
   "metadata": {},
   "source": [
    "Plot column charts of the unsupervised clusters split by their Age, Sex, Distress and Behaviour labels to identify patterns in the cluster composition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "f9c164b6",
   "metadata": {
    "metadata": {},
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJwAAAMQCAYAAABi+3/xAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3hT5f//8Ve6W1po2ZtCmZW9kWlBBESl4ARB9kZZIojIkL1kyZK9ZMtwfBC3KIIsAdnQIrLLKthCV35/+Gu+1rSlaZMmaZ+P6+oFvc/JOa+kLbz7zn3uYzAajUYBAAAAAAAAVuJi7wAAAAAAAADIWmg4AQAAAAAAwKpoOAEAAAAAAMCqaDgBAAAAAADAqmg4AQAAAAAAwKpoOAEAAAAAAMCqaDgBAAAAAADAqmg4AQAAAAAAwKpoOAEAAAAAAMCqaDgBdjZkyBCVK1dOy5Yts3cUAAAAuzpz5owGDRqk+vXrq2LFimrQoIEGDhyoU6dO2TsaAMBCBqPRaLR3CCC7un//vho0aKDixYsrJiZG//vf/2QwGOwdCwAAINOdPXtWL7/8sqpWraqXX35ZefLk0bVr17RmzRqdOnVKq1atUtWqVe0dEwCQRsxwAuzos88+kySNHDlS4eHh+vXXX+2cCAAAwD6WL1+ugIAAffzxx2rZsqVq166t559/XitWrJC/v7/mz59v74gAAAvQcALsaMuWLapXr57q1q2rEiVKaP369Wb7LF26VE2bNlXlypX16quv6ttvv1W5cuW0b98+0z5nzpxRr169VL16dVWvXl39+vXTpUuXMvOpAAAAZEhERISMRqMSEhKSjPv4+Ojdd99Vy5YtTWNff/212rZtq0qVKql+/foaP368oqKiJEkPHjzQU089pRYtWigmJkaSZDQa1alTJ9WvX1+3b9/OvCcFANkYDSfATs6ePatjx46pTZs2kqQ2bdrom2++UUREhGmfefPmafr06WrZsqXmz5+vKlWqaODAgUmOExYWpldffVW3bt3SlClTNGHCBF26dEmvvfaabt26lYnPCAAAIP2aNGmiK1eu6NVXX9XatWt1/vx5Ja7+0aJFC4WGhkqSdu7cqX79+qlUqVL66KOP1L9/f+3YsUN9+/aV0WiUr6+vJkyYoPDwcC1cuFCStGrVKu3bt08TJ05U7ty57fYcASA7cbN3ACC72rJli/z9/RUSEiJJCg0N1dy5c7V582b17t1bUVFR+vjjj9WhQwcNHTpUktSgQQNFR0drw4YNpuPMmzdP3t7eWrFihXx9fSVJ9erVU7NmzbRkyRK98847mf/kAAAALNS+fXvdvHlTS5cu1bhx4yRJAQEBatCggTp16qTKlSvLaDRq+vTpatiwoaZPn256bGBgoDp37qwffvhBTZo00ZNPPqlXXnlFixcvVpUqVTRz5kx16NBBjRs3ttfTA4BshxlOgB3ExsZqx44datasmR4+fKjIyEjlyJFDNWrU0MaNG5WQkKAjR47o4cOHatGiRZLHtm7dOsnnv/76q2rXri0vLy/FxcUpLi5Ovr6+qlmzpn755ZfMfFoAAAAZ8tZbb+mnn37SjBkz9OKLL8rX11c7d+7Uyy+/rFWrVunChQu6du2aQkJCTHVPXFycatWqJV9fX/3888+mYw0bNkwFChRQ7969VaRIEQ0bNsyOzwwAsh9mOAF28P333+vWrVvavHmzNm/ebLb9p59+0v379yXJbNp3njx5knx+9+5dffHFF/riiy/MjsOUcQAA4Gxy5cql1q1bm95kO3HihN5++21NmzZNTzzxhCRp7NixGjt2rNljb9y4Yfp7jhw51Lx5cy1btkz16tWTl5dX5jwBAIAkGk6AXWzZskXFihXThAkTkowbjUb1799f69evV7du3SRJt27dUqlSpUz7/HehSz8/Pz355JPq0qWL2Xnc3PgRBwAAju/69etq166d3nrrLb300ktJtgUHB2vQoEHq16+f4uPjJf0ze6l27dpmx8mVK5fp72fOnNHq1atVoUIFffLJJ3r++edVpUoV2z4RAIAJl9QBmezmzZv66aef9Oyzz6pOnTpJPurWrasWLVrohx9+UKFCheTn56fdu3cnefxXX32V5PPatWvr3LlzqlChgipVqqRKlSqpYsWKWrFihdljAQAAHFHevHnl5uamdevW6dGjR2bbL1y4IE9PT5UpU0Z58uTRX3/9Zap7KlWqpAIFCmjGjBk6ceKEJCkuLk7Dhw9X8eLFtX79epUvX17vvPNOsscGANgG0x+ATLZt2zbFxcXp2WefTXZ7mzZttGnTJm3dulXdu3fXnDlz5O3trdq1a2v//v365JNPJEkuLv/0i/v27atXX31VvXr10muvvSZPT09t2LBBX3/9tebMmZNpzwsAACC9XF1dNWbMGPXr10/t2rVThw4dFBQUpOjoaP38889au3at3nrrLQUEBGjQoEF6//335erqqqeeekqRkZGaP3++rl+/brrkbuHChTpx4oTWrVsnLy8vffDBB3rppZf04Ycfavjw4XZ+tgCQPRiMifcaBZApWrZsKVdXV3322WfJbjcajWrWrJliY2P13XffafHixdqwYYMiIiJUpUoVPf3005o0aZK2bt1qKqr++OMPffjhhzp06JCMRqPKli2rnj17qmnTppn51AAAADLkjz/+0NKlS3Xw4EHdvn1bHh4eCg4OVseOHdW8eXPTfl988YWWLFmis2fPysfHR9WrV9fAgQNVrlw5nTp1Si+++KJeeukljR492vSYyZMna+XKlVqzZo1q1Khhj6cHANkKDSfAQcXFxemzzz5TnTp1VKhQIdP42rVrNX78eO3bt085c+a0Y0IAAAAAAJJHwwlwYM8++6w8PDzUp08fBQQE6MyZM5o1a5aaNWumSZMm2TseAAAAAADJouEEOLBLly5p5syZ2rdvnyIjI1W4cGE9//zz6tWrl9zd3e0dDwAAAACAZNFwAgAAAAAAgFW52DsAAAAAAAAAshYaTgAAAAAAALAqGk4AAAAAAACwKjd7B8hshw8fltFoZMFlAMgEsbGxMhgMqlatmr2jAEgBtREAZB5qI2Qn2W6Gk9FolC3WSTcajYqJibHJsa3NmbJK5LUlZ8oqOVdeZ8oq2S6vrf7NBWA9tvg55d9A23KmvM6UVSKvLTlTVonaCLCGbDfDKfHdu0qVKln1uFFRUTp58qRKly4tHx8fqx7b2pwpq0ReW3KmrJJz5XWmrJLt8h47dsxqxwJgG7aojfg30LacKa8zZZXIa0vOlFWiNgKsIdvNcAIAAAAAAIBt0XACAAAAAACAVdFwAgAAAAAAgFXRcAIAAAAAAIBV0XACAAAAAACAVWW7u9QBALKu+Ph4xcbG2jsGrMzd3V2urq72jgEAgNOhNnJM2aW2oeEEAHB6RqNR165d0927d+0dBTbi7++vggULymAw2DsKAAAOj9rI8WWH2sbuDafr16+rUaNGZuOTJk1S27ZtdfLkSU2YMEHHjx9X7ty51blzZ3Xq1MkOSQEAjiqxoMqfP798fHyy9H/c2Y3RaFRUVJRu3LghSSpUqJCdEwEA4PiojRxXdqpt7N5wOnXqlDw9PfX1118n+SHw8/PTnTt31KVLF4WEhGjs2LE6cuSIxo4dqxw5cqhdu3Z2TA0AcBTx8fGmgipPnjz2jgMb8Pb2liTduHFD+fPnzxZT0AEASC9qI8eXXWobuzeczpw5o8DAQOXPn99s28qVK+Xu7q5x48bJzc1NQUFBunjxohYvXkzDCQAgSaZ1CXx8fOycBLaU+PWNjY3NskUZAADWQG3kHLJDbWP3u9SdPn1aQUFByW47cOCAateuLTe3/+uL1a1bV+Hh4YqIiMisiAAAJ8BU8ayNry8AAJbh/07Hlh2+PnZvOJ05c0a3b99Whw4d9OSTT+q1117Tjz/+KOmf604LFiyYZP/EmVBXr17N9KwAANiS0Wi0dwQAAIBMQ+2Ttdn1krq4uDhduHBBpUuX1vDhw+Xr66vPP/9cPXv21PLly/Xw4UN5eHgkeYynp6ck6dGjR+k+b+IiXdYUHR2d5E9H5shZk+vyxsTEyN3d3SHzJseRX9//cqasknPldaasku3yGo3GLP/uzbFjx7Rq1Sr99ttvun37tvLnz6969eqpZ8+eKlasWJqPc/bsWY0aNUrr16+3YVoAwOO4u7tn+f+7gIyg9kFa2bXh5Obmpn379snV1VVeXl6SpIoVK+rs2bNaunSpvLy8FBMTk+QxiY2mjFyPGhsbq5MnT6Y/eCrCw8NtclxbcLSs7u7uCg5+Qm5uSa9f9fb2lq+vn06c+MN0PbIzcLTXNzXOlFVyrrzOlFWyTd7/vnGQlaxdu1YTJ05UnTp1NGTIEOXPn18XL17U0qVL9dVXX2nlypUqX758mo71v//9T4cPH7ZxYgBAagwGQ7L1aEriE4xydaE5heyD2geWsLjhdO7cOe3cuVO//vqr/vrrL92/f18BAQEqXLiwGjVqpObNm6e4JlNycuTIYTZWpkwZ7dmzRwULFjTdKjBR4ucFChSwNLqJu7u7Spcune7HJyc6Olrh4eEKDAw0rTjvqBw1q8FgkJubq0ZvPqrwiAem8cC8vhr7YmUVL17cKX5xddTXNznOlFVyrrzOlFWyXd5z585Z7ViO5uDBg5owYYI6dOigkSNHmsbr1KmjZs2aqU2bNnr33Xe1detWO6ZEZrB2bQTAvpKrR5OTWKMC2QW1DyyV5obThQsXNG3aNH3//fcqUKCAKlasqKpVq8rb21uRkZG6du2aVq1apTlz5qhp06YaOHDgY5s6Z8+e1SuvvKIFCxaoTp06pvHjx4+rdOnSqlChgtavX6/4+HjTqu2//vqrSpYsmaHbOxoMBput2O/t7e00dwNw1KzhEQ90+up9s3FPT0+n+MU9kaO+vslxpqySc+V1pqyS9fNm5UsSli5dKj8/Pw0ePNhsW+7cuTV8+HCFhYUpKipKLi4u+uijj7Rr1y5duXJFHh4eqlKlioYNG6YKFSpo7ty5mjdvniSpXLly6t+/vwYMGKCEhAQtWbJEmzZt0tWrV1WkSBG9/vrr6tixo1mWdevW6ebNmwoODlbPnj3Vp08frVq1yvT/67FjxzRr1iwdP35csbGxql27toYMGaIyZcpIkvbt26dOnTpp7NixWrRoke7du6epU6eqX79+Wrp0qRo0aGA634EDB9ShQwetW7dONWrUsNVL7PBsURsBcAwp1aNAdkbtQ+1jqTQ1nJYsWaKPP/5YrVu31ieffKKqVaumuO/Ro0e1fv16tW/fXj169FCPHj1S3DcoKEilSpXSuHHjNHbsWAUEBGjjxo06cuSItmzZojx58mjJkiUaOXKkunfvrqNHj2rFihUaO3asxU8UAABrMRqN2rNnj0JCQlJshLdq1cr09zfffFMHDhzQ4MGDVbx4cV28eFGzZ8/WkCFD9Pnnn+ull17StWvXtHnzZm3YsMF0w4wxY8Zo69at6tWrl6pVq6bffvtNEydOVGRkpPr16ydJmjdvnj766CN169ZNdevW1U8//aSBAwcmyfLrr7+qe/fuqlOnjiZOnKhHjx5p0aJFevXVV7Vx48Yks2/mzZun9957Tw8fPtSTTz6p/Pnza/v27UmKrm3btikwMDBbF1y2qo0AAHBE1D7UPumRpobT2bNntXPnTtMd4lJTuXJlVa5cWQMGDNDMmTNT3dfFxUULFy7UjBkzNHDgQEVGRio4OFjLly9X2bJlJf1T0E2YMEGhoaHKly+fhg0bptDQ0LTEBgDAJu7cuaNHjx6paNGij903JiZGf//9t9577z1TIVa7dm09ePBAkydPVkREhAoWLGgqtBIbF2FhYdq4caMGDx6snj17SpIaNGggg8GgRYsWqX379vL09NTHH3+sDh06aOjQoaZ9oqOjtWHDBlOGGTNmqESJElq8eLFpxnCDBg309NNPa86cOZo9e7Zp3/bt26tFixamz0NDQ7V69Wr9/fffypEjhx4+fKgvv/zSlCm7slVtBACAI6L2ofZJjzQ1nKZMmWLxgQsVKqRp06Y9dr+8efNq0qRJKW6vXLlykm8cAADsLbFwiY+Pf+y+Hh4eWrp0qSTp+vXrCgsLU3h4uL777jtJMrs5RqJff/1VRqNRISEhiouLM42HhIRowYIFOnjwoHx8fPTw4cMkRZIktW7d2vR/Z1RUlI4dO6b+/fubcktSzpw59dRTT+mHH35I8tgKFSok+bxdu3ZatGiRdu/erTZt2mj37t2KiopSmzZtHvvcszJb1kYAADgaah9qn/TI8F3qjEaj7ty5I39/f7m4uFgjEwAADi1XrlzKkSOHrly5kuI+UVFRio2NVa5cufTTTz9p4sSJunDhgnLkyKHy5cub1soyGo3JPv7u3buSpGeffTbZ7devX1euXLkk/bNuwr/9e53D+/fvy2g0Km/evGbHyJs3r+7fT7pGyX/X8CpRooRq166tbdu2qU2bNtq2bZuefPLJDN28I6ujNgIAZDXUPtQ+6ZHuhtPdu3f1/vvv69tvvzUt6t24cWONHj06TdPLAQBwZg0aNNC+ffv06NEjeXp6mm3fuHGjpkyZok2bNqlfv35q1qyZFi1apGLFislgMGjt2rX66aefUjx+zpw5JUkrV65M9o6uhQsXVlhYmCTp1q1bKlWqlGnb7du3TX/38/OTwWBQRESE2TFu3rwpf3//xz7Xdu3a6d1339X58+e1d+9eTZ8+/bGPyY6ojQAAWRm1DyyV7rfdPvjgA3l4eGjdunX63//+p+XLl+vhw4d65513rJkPAACH1LVrV929e1ezZs0y23bz5k0tW7ZMpUuX1p9//qlHjx6pZ8+eKl68uOnOfYkFV+K7fP+dCVOzZk1J/6yZUKlSJdPH7du3NXv2bN29e1fly5eXn5+fdu/eneSxX331lenvPj4+qlixor788ssk0+Dv37+v77//Pk2LXz7zzDPy9vbWmDFjlCNHDjVr1iwNr1D2Q20EAMjKqH1gqTTNcPrss8/UunXrJGMnTpzQvHnzTKu7lyhRQlFRUaaFuwAAyMqqVq2qt956S7NmzdL58+fVpk0bBQQE6OzZs1q6dKkePXqkWbNmyc3NTW5ubpo2bZq6du2qmJgYbd26Vd9//72kf6afS//3rt5nn32mKlWqqFy5cnr++ec1atQoXb58WRUrVlRYWJg+/PBDFS1aVIGBgXJ1dVX37t01Z84ceXt7q3bt2tq/f78++eQTSf9XyA0ZMkTdunVTz5491b59e8XGxmrx4sWKiYkx3fElNd7e3nr22We1YcMGvfbaa/Lw8LDBK+pcqI0AANkNtQ8slaYZTsuXL1fbtm31888/m8YaNWqkQYMGaenSpdq8ebMWLlyosWPHqmnTpjYLCwCAI+nTp48WL14sSZo4caJ69uypNWvWqEmTJtq2bZuCgoJUokQJzZgxQ9evX1efPn30/vvvS5JWr14tg8GgAwcOSJKaN2+uSpUqafjw4aaFNidNmqQuXbpo/fr16t69uxYuXKhWrVpp2bJlpkUwe/XqpQEDBmj79u3q1auXDhw4YGpwJK5JUK9ePdNsm8GDB2vUqFEqUKCANm7caLor7OM0adJEktS2bVvrvHhOjtoIAJAdUfvAEmma4bRlyxZ99tlnGj16tIoUKaK3335b77zzjtasWaOvvvpKt27dUp48edS+fXt16tTJ1pkBAHAYjRs3VuPGjVPdp0WLFmZ3U5GkU6dOmf5eoEABbd68Ocl2Nzc39evXL8V34uLi4vTZZ5+pTZs26tOnj2l87dq1cnFxUbFixUxjderU0dq1a1PMWKdOHZ0+fTrF7T/88IMqVKigypUrp7hPdkJtBADIrqh9kFZpXjS8devWeuaZZ7R27Vr16NFDtWvX1qBBgyiiAACwEzc3N3388cdauXKl+vTpo4CAAJ05c0azZs1SmzZtTFPVM2LVqlW6cOGCNm7cqGnTplkhddZBbQQAQOai9nEuFi0a7u7urs6dO2v37t0qUaKE2rVrpzFjxiS7+jsAALC9hQsXKjAwUGPGjFGXLl20cuVKvfHGGxo3bpxVjn/gwAFt375db7zxhtmaRaA2AgAgs1H7OI80z3A6f/689u3bp4SEBNWoUUODBw9Whw4dNHv2bD3zzDN6/fXX1aNHD/n6+toyLwAA+JdixYrpww8/tNnx58yZY7NjOztb1UZ3797VzJkz9f333+vBgwcqV66chgwZYrp7T5cuXfTLL78keUzt2rW1evVqqz03AAAcFbWP80jTDKdPP/1UL7zwgjZs2KBt27bp1Vdf1fLly1WgQAFNnDhR69ev1+nTp9WsWTOtWLHCxpEBAADsy5a10eDBg3X48GHNnDlTW7ZsUYUKFdStWzdduHBBknT69GmNGTNGe/bsMX3MnTvXBs8SAAAg/dLUcJo7d65GjBih7du3a/PmzVq1apXmzJkjo9EoSSpTpowWLlyoOXPm6IsvvrBpYAAAAHuzVW108eJF/fzzzxozZoxq1qypkiVLatSoUcqfP7927typW7du6datW6pSpYry5ctn+vD397fRMwUAAEifNDWcHj58mKSQyZUrl2JjYxUXF5dkv9q1a2vjxo1WDQgAAOBobFUbBQQEaPHixapUqZJpzGAwyGAwKDIyUqdPn5bBYFDJkiUz/BwAAABsKU1rOL388st699139dlnn8nLy0v79+/Xiy++KHd3d1vnAwAAcDi2qo1y5sxpdqvpXbt26eLFi3r33Xd15swZ+fn5ady4cfr555/l4+OjFi1aqG/fvvLw8MjQuQEAAKwpTQ2ngQMHqmrVqtq7d68MBoNGjRqlFi1a2DobAACAQ8qs2ujQoUMaMWKEmjdvriZNmujdd9/Vo0ePVLlyZXXp0kUnT57U1KlTdeXKFU2dOjXd5zEajYqKirJa7ujo6CR/Ojry2o4zZZWkmJgYeXt7W/SY6Oho0+W0mc2ZXl9nyirZLq/RaJTBYLDqMQFHlea71DVp0kRNmjSxYRQAAADnYeva6Ouvv9bQoUNVvXp1TZ8+XZI0btw4vfPOO8qVK5ckqWzZsnJ3d9egQYM0bNgw5c2bN13nio2N1cmTJ62WPVF4eLjVj2lL5LUdZ8nq7e1t8ZpoYWFhdm+iOMvrKzlXVsk2eZmRiuwiTQ2n2bNnq0+fPhb9YERFRWnRokUaNGhQusMBAAA4IlvXRmvWrNGECRPUokULTZkyxXQeNzc3U7MpUZkyZSRJ165dS3fDyd3dXaVLl07XY5MTHR2t8PBwBQYGWjxbxB7IazvOlFX6Z4aTpUqWLGnXGU7O8vo6U1bJdnnPnTtntWMBji5NDafIyEg1b95cnTt3VqtWrZQ/f/4U971586Y2bdqkTz75RM2aNbNaUAAALGWMj5fB1dUpz/3gwQPVr19fOXLk0A8//PDYtYFCQkIUGhqqAQMGpOn4/93/u+++U7FixazadMjKbFkbrVu3Th988IE6duyokSNHJrn0omPHjipatKgmTZpkGjt27Jjc3d0VGBiY7udjMBjk4+OT7senxNvb2ybHtRXy2o6zZE3PpU6O0DxxltdXcq6skvXz2vtyOnvVRuk9b8eOHbV///5kt3Xt2lXvvPNORqNZ7K+//lLTpk21atUq1alTJ9PP70zS1HAaNWqUmjVrpsmTJ2vq1KmqUqWKKleurKJFi8rb21v379/X1atXdfDgQZ0+fVpBQUGaOHGiGjZsaOv8AACkyODqqtv9ByjubOa+m+hWprRyz5uboWN8/vnnypMnj27evKndu3erVatWVkpn7vLly+rdu7dWrVpFwymNbFUbhYWFaeLEiXr66afVq1cvRUREmLZ5eXnpmWee0cSJE1W5cmU1aNBAx44d09SpU9WtWzf5+vra+mkDAJycPWqjjNZFLVu21MiRI83GHaHZi9SleQ2nevXqafv27fr++++1c+dOffbZZ7p165Zpe968edWgQQP1799fTz31lE3CAgBgqbiz5xR7/Li9Y1hsy5Ytatiwoa5cuaL169fbtOFkr0tBnJ0taqNdu3YpNjZWu3fv1u7du5NsCw0N1eTJk2UwGLR69WpNnDhR+fLlU+fOndWzZ0+rPjcAQNblbLWRl5eX8uXLZ+8YSIc0N5wS/XuBzOjoaN2/f1/+/v4sfAYAgJWcP39ev//+u7p376579+7pvffeU1hYmEqWLClJun//vsaPH69vvvlGbm5u6tWrV5LHb926VSNGjNDp06dTHZP+b1q4JHXq1En9+/dP82V5+Ic1a6PevXurd+/eqe7ToUMHdejQIT1RAQDIUoxGo5YsWaL169crIiJCgYGB6tatm55//nlJ0r59+9SlSxfNnj1b06dP19WrV1W1alVNmTJFS5cu1bZt2+Tu7q5OnTqpT58+kv5Zy+3DDz/Url27dOPGDfn4+KhevXoaPXq0cufOnWyOLVu2aMmSJbp8+bKKFCmiV199VR07dpSLi0umvRaOKEPP3tvbW/nz56fZBACAFW3evFk+Pj5q1KiRnn76abm7u2v9+vWm7QMHDtTRo0e1cOFCLV++XN9//70uX76crnMVKlRImzZtkiTNnTtXXbt2tcpzyK6ojQAAyDwffvihPvnkE40aNUo7d+5Up06dNGbMGK1du9a0T3x8vBYsWKDp06dr5cqVOnXqlF544QW5u7tr06ZNevXVVzVr1izTm3JTp07VV199pcmTJ2vXrl2aPHmyfv31Vy1YsCDZDBs2bNDUqVPVv39/ff755xo4cKA+/vhj0x1mszOLZzgBAADbiYuL044dOxQSEiIvLy95eXmpQYMG2rZtmwYPHqzLly9rz549WrFihWrWrClJmjFjRrovZ3d1dTW9W5crVy7lyJHDas8FAAAgo3bu3Kldu3YlGatRo4bmzJmjFStWaObMmaaZxsWLF9fly5e1dOnSJLOB33rrLVWqVEmSVLduXf3+++8aNmyYDAaDevXqpfnz5+vs2bMqV66cKlWqpBYtWpjqrCJFiujJJ5/UmTNnks03f/589enTR88++6wkqVixYnrw4IHGjh2rt956S56entZ+SZwGDScAABzIDz/8oIiICFPRIknPPvusvvvuO3355Zfy8vKSJFPRJP2zVlCxYsUyPSsAAICthYSEaOjQoUnGvLy8dO7cOT169EhDhgxJculaXFycYmJi9PDhQ9NYiRIlTH/38fFR0aJFTXcMTKytYmJiJEkvvPCCfvnlF02fPl3h4eG6cOGCwsLCTA2of7t9+7auXbummTNnavbs2abxhIQEPXr0SH/99ZeCgoKs8Co4JxpOAAA4kK1bt0qS+vfvb7Zt/fr16tKli6R/Cpl/c3NL/b/0+Ph4KyUEAADIPDly5EjSMEp07do1SdKsWbNUqlQps+3/vrz9v3VSamsrvf/++9q1a5fatGmjkJAQ9evXT0uXLtX169fN9k2sx0aMGKEnn3zSbHuhQoVSPE92YPEaTnv37k3SKQQAANZx69Yt/fDDD2rbtq22bduW5KNdu3Y6fPiwqeA6dOiQ6XGRkZH6888/TZ+7u7tLkh48eGAaCw8PT/G8ie/wIX2ojQAAyHylSpWSm5ubrly5ohIlSpg+fvjhBy1dujRdC3bfuXNHGzZs0OjRozVixAi1bdtWFSpU0IULF5K9q2+ePHmUO3duXbp0KUmGP/74Q7NmzbLCs3RuFn8FBgwYoK+++soWWQAAyNZ27NihuLg49ejRQ2XLlk3y0bt3b7m4uGjjxo1q0aKFxo0bp19++UVnzpzRsGHDTNPAJalq1aoyGAyaO3eu/vrrL3355Zf69NNPUzyvj4+PJOnMmTO6f/++zZ9nVkNtBABA5vPz89Orr76q2bNna/v27bp06ZI2b96sadOmKX/+/Ok6pq+vr/z8/PTNN9/o4sWLOn36tEaNGqU//vgjSa2VyGAwqEePHlq9erXWrFmjP//8U7t379aYMWPk5eWV7W8iYvEldTlz5jRd4wgAgKNzK1Paac65detWPfnkk8lOCy9evLiaNWumHTt26Mcff9S0adM0aNAgJSQk6JVXXtHt27dN+xYrVkxjx47VokWLtG7dOtWoUUPDhg3TO++8k+x5AwIC1K5dO02dOlUXL17Ue++9l6782RW1EQDAmWR2bWTL840YMUIBAQGaPXu2bty4oUKFCunNN99U9+7d03U8d3d3zZ49W5MnT9Zzzz2nXLlyqU6dOho8eLAWLVqk6Ohos8d07dpVnp6eWr16tSZPnqy8efPq5Zdf1ptvvpnRp+f0LG449erVS+PHj1dYWJjKly9velf032rVqmWVcAAAZIQxPl65582127kNrq4WPWbnzp2pbp879/+ey+jRozV69OgU933llVf0yiuvJBlr06aN6e/ffvttkm0TJ07UxIkTLUiLRNRGAABnYa/aKD11kSStXr061e1ubm7q379/smtfSlKdOnV0+vTpJGOTJ0822+/f+9SvXz/ZmqxXr16SpKJFi5ods0OHDknuiod/WNxwSixuP/zwQ0lJ130wGo0yGAw6efKkleIBAJB+6SlsssK5kbmojQAAzsJe9Ql1UfZkccNp1apVtsgBAADglKiNAAAAzFnccKpdu7YtcgAAADglaiMAAABzFjecJOn27dtaunSpfvnlF928eVNLlizR119/rfLly6tZs2bWzggAAODQqI0AAACScrH0AZcuXdLzzz+vjRs3qkCBArp165bi4+MVFhamN998U99//70NYgIAADgmaiMAAABzFs9wmjJlivLkyaPVq1fLx8dHFStWlCTNmDFDjx490sKFC9WkSRNr5wQAAHBI1EYAAADmLJ7htHfvXvXt21c5c+ZMchcW6Z9bMJ89e9Zq4QAAABwdtREAAIA5ixtOkuTmlvzEqJiYGLNCCwAAIKujNgIAAEjK4oZTzZo1tWjRIkVFRZnGDAaDEhIS9Mknn6h69epWDQgAAODIqI0AAADMWdxwGjJkiM6fP6/mzZtr2LBhMhgMWrp0qdq2bauDBw9q0KBBtsgJAECWN3z4cHXs2DHN+y9YsEC1a9dWtWrVdOzYMatkOHv2LItcW4jaCAAA2wgJCVG5cuW0fPnyZLe///77KleunObOnZvm46V1X2ScxQ2nsmXLavPmzapTp4727dsnV1dX/fLLLypevLjWr1+vChUq2CInAAAWi08wZtlz379/X7Nnz1b79u312WefqXz58lY5bq9evazWvMouqI0AAM7CXrVRRs7r7u6uXbt2mY3HxcXpq6++4tJ1B2bxXeokqWTJkpoxY0ay265du6aCBQtmKBQAANbg6mLQ6M1HFR7xIFPPG5jXV2NfrGzTc0RGRspoNKpu3boqUqSITc+Fx6M2AgA4A3vURhmti+rVq6effvrJ7P/TX3/9VT4+PvL29rZGTNiAxQ2nChUqaMOGDapc2fwb5sCBA+rRo4cOHz5slXAAAGRUeMQDnb56394xLBYSEqIOHTroyJEj2rNnjzw8PPTcc89p+PDhOnjwoDp16iRJeuONN1S7dm2tXr1aBw4c0Jw5c3T8+HHFxMSoWLFi6t27t1544QVJ/1yyJ0kBAQHatm2boqKiVLduXY0bN04FChRQSEiILl++rHnz5mn//v1avXq13Z6/M6E2AgA4E2erjSpXrqzz58/rf//7nzp37mwa/+KLL9SyZUt9+eWXprFNmzZp1apVunjxolxcXBQcHKwRI0aoUqVKyR770KFDmjFjho4dO6bcuXPrqaee0pAhQ+Tr62vrp5UtpKnhtGzZMtNCmEajUZs2bdKPP/5ott/hw4fl4eFh3YQAAGRTs2fP1tChQzVs2DDt379fI0eOVMWKFdWqVStt2rRJL730kubOnavatWvr+vXr6tatm15//XV98MEHio2N1ccff6yRI0eqfv36yps3ryTps88+03PPPac1a9bo1q1bGjx4sGbNmqVJkyZp8+bNCg0NVatWrdSrVy87P3vHRm0EAEDmadmyZZKGU0xMjL7++mutWLHC1HDavXu3xo0bp/Hjx6tmzZq6efOmPvjgA7333nvavn272TFPnTqlLl26qE+fPpowYYIiIiI0depUde3aVRs2bOBSPStIU8Pp0aNHmjdvnqR/7rqyadMms31cXFzk5+enPn36WDchAADZVIMGDUwzmYoVK6bVq1fr0KFDatOmjXLnzi1JypUrl/z9/fXnn39qwIAB6tatm6lA6tmzp7Zt26bw8HBTw8nPz0/jxo2Tu7u7goKC1KpVK/3www+SpNy5c8vV1VU+Pj7y9/fP/CfsRKiNAADIPC1bttTSpUt1/fp1FShQQD///LNy586t4OBg0z7+/v6aMGGCnn/+eUlSkSJF9OKLL2rcuHHJHnPp0qWqX7++evfuLUkKDAzUjBkz1KxZM+3fv1916tSx/RPL4tLUcOrTp4+pWCpfvrw2bNigKlWq2DQYAADZXVBQUJLP/fz8FBsbm+y+xYsXV9u2bbVq1SqdOXNGf/75p06dOiVJio+PT7Kfu7t7mo6JlFEbAQCQeSpWrKhixYpp165d6tSpk7744gs9++yzSfapVauWzp8/r48++kgXLlzQxYsXdfr0aSUkJCR7zBMnTujixYuqVq2a2bbz58/TcLICi9dwSixe/y0uLk4PHjzg3VAAAKwouUuxjMbk7/Jy7tw5tW/fXk888YSefPJJNW/eXAEBAXrppZcee0xkDLURAAC2l3hZ3SuvvKJvvvnGbHbxzp07NXz4cD333HOqXr26Xn31VZ05cybFGU4JCQl67rnnTDOc/i1xJjkyxsXSB8THx2vevHnauXOnJGnfvn2qX7++6tWrpzfeeEP37t2zekgAAJC69evXK0+ePFq+fLl69Oihxo0bKyIiQlLKTSpYB7URAAC217JlSx06dEhbtmxRsWLFzGaCL168WC+++KImT56sDh06qFatWrp06ZKk5GuhMmXK6Ny5cypRooTpIy4uTpMmTdLVq1cz5TlldRY3nGbPnq0FCxYoMjJSkjR+/Hj5+/trxIgR+vPPP1O8JXBahIWFqVq1atq6datp7OTJk3r99ddVtWpVhYSEaNWqVek+PgAAWVXBggV17do1/fDDD7p8+bK++uorjRkzRtI/C2umVY4cORQeHm5qVuHxbFkbAQCAf1SoUEElSpTQjBkzzC6nk6RChQrp0KFD+uOPP/Tnn39qxYoVWrNmjaTka6GuXbvqxIkTGjt2rM6fP6/Dhw9ryJAhCg8PV2BgoK2fTrZg8SV1n3/+uQYPHqwOHTro/PnzOnv2rCZPnqw2bdrI399fU6dOTXHKWmpiY2M1dOhQ0x1fJOnOnTvq0qWLQkJCNHbsWB05ckRjx45Vjhw51K5dO4vPAQDIfgLzZv5tbe1xzk6dOunChQsaNmyYYmJiFBgYqMGDB2vOnDk6duyYGjVqlKbjdOzYUVOmTNHZs2e1Y8cOG6fOGmxVGwEAYAuZXadY83wtW7bUggUL1KpVK7Nto0aN0vvvv6/XX39dHh4eKl++vKZOnapBgwbp2LFjqlmzZpL9q1atqiVLlmj27NkKDQ2Vj4+P6tWrp3feeYclCKzE4obTjRs3TItifv/993JxcTEVsQULFtT9+/fTFWTu3Lny9U36jbhx40a5u7tr3LhxcnNzU1BQkC5evKjFixfTcAIAPFZ8glFjX6xst3O7ulh2O93Jkyeb/v7tt9+abV+9erXp70WLFtXp06dNn3t4eGjixIlmj2nTpk2yx080YMAADRgwwPT5yy+/rJdfftmi3NmdrWojAACszV61UXrqIsm8Hho4cKAGDhyY4j7Lly83O8a/m1P/PV69evVUr149i3MhbSy+pC5//vz666+/JP3zxapQoYJpQa3Dhw+rYMGCFof47bfftGHDBrNC+MCBA6pdu7bc3P6vL1a3bl2m+gMA0iQ9hU1WODcyly1qIwAAbMFe9Ql1UfZkccOpdevWmjRpkrp166aDBw+aZhpNmDBBc+fO1XPPPWfR8SIjIzVs2DC99957KlSoUJJt165dMyvS8ufPL0ks4gUAAByCtWsjAACArMDiS+oGDhwoHx8f/fbbbxoyZIjat28vSTp27Ji6du2qPn36WHS8MWPGqFq1askWYw8fPjS7dtLT01OS9OjRI0ujmxiNxiRrRVlDdHR0kj8dmaNmNRgM8vb2TnH7o0ePnOJOS476+ibHmbJKzpXXmbJKtstrNBplMPCOFrI2a9dGAAAAWYHFDSeDwaBevXqpV69eScbXr19v8cm3bdumAwcOmG4j/F9eXl5mq8knNpp8fHwsPl+i2NhYnTx5Mt2PT014eLhNjmsLjpbV29tbwcHBKW6/cuWK0/zyLjne65saZ8oqOVdeZ8oq2SYviy4iq7NmbQQAAJBVWNxw+u233x67T61atdJ0rC1btujWrVtq0qRJkvHRo0friy++UMGCBXXjxo0k2xI/L1CgQNoCJ8Pd3V2lS5dO9+OTEx0dbbp9YmqzdByBo2Z93CyIwoULO8Uvro76+ibHmbJKzpXXmbJKtst77tw5qx0LcFTWrI0AAACyCosbTh07dpTBYDBd2pRckyCts4emT5+uhw8fJhlr3ry53nzzTT3//PPavn271q9fr/j4eLm6ukqSfv31V5UsWVJ58uSxNLqJwWDI0Ayp1Hh7e9vs2NbmTFmlfy6ndIZf3BM50+vrTFkl58rrTFkl6+flcjpkB9asjQDAEgaDQe7u7vaOAQflDMuRZGfZ4etjccNp1apVZmNRUVE6cOCAtm/frrlz56b5WCnNUsqTJ48KFCigdu3aacmSJRo5cqS6d++uo0ePasWKFRo7dqylsQEAAGzCmrURAEiSMT5ehv//hntqvL29VbFCBT2Ki8uEVHAWiXd5j+P7wqElfn0Sv15ZkcXPrHbt2smON2nSRD4+PlqwYIEWLVqU4WDSP42nJUuWaMKECQoNDVW+fPk0bNgwhYaGWuX4AAAAGZWZtRGA7MHg6qrb/Qco7mzql6a7lSmt3PPmSjQW8C+urq5ydXVVZGSk/Pz87B0HKYiMjDR9rbIqq7bSatasqY8//jhDxzh9+nSSzytXrqwNGzZk6JgAAAD2YI3aCED2FHf2nGKPH7d3DDghg8Gg/Pnz6+rVq/L09FSOHDlY5sCBGI1G/f3334qMjFShQoWy9NfGqg2nb7/9Vjly5LDmIQEAAJwWtREAwB5y5cql6OhoRURE6ObNm/aOg/8wGAzy9/dXrly57B3FpixuOHXq1MlsLCEhQdeuXdPly5fVo0cPqwQDACCjEowJcjG4OO25y5Urp0mTJqlt27ZWSpV+W7du1YgRI0wzkR0pm71RGwEAHI3BYFChQoWUP39+xcbG2jsO/sPd3T1LX0qXyOKGU3Irqbu4uKhs2bLq1auX2rVrZ5VgAABklIvBRTMOTNOl+5cy9bzF/IppSM23M/WcttaqVSs1bNjQ3jEcErURAMBRZfU1guDYLG44rV692hY5AACwiUv3L+nCvfP2juH0vLy85OXlZe8YDonaCAAAwFya5vpfuXLFog8AAGA9c+fOVUhISIpjw4cP10svvZRk++XLl1W+fHn98ssvkqRDhw6pQ4cOqly5spo0aaKxY8fqwYMHpv1DQkK0dOlSDRgwQNWqVVOdOnU0fvx40y17t27dqnLlyqWY8bvvvlPbtm1VuXJlPf3005o1a5ZiYmKs8vwdkS1ro7t37+r9999Xo0aNVL16db322ms6cOCAafvevXvVtm1bValSRS1atNDnn39u7acHAACQYWma4RQSEmLRyuknT55MdyAAAGCZtm3bqmPHjvrzzz9VvHhxSdLOnTtVsGBB1a1bV6dOnVKXLl3Up08fTZgwQREREZo6daq6du2qDRs2mP6Pnz17toYOHaphw4Zp//79GjlypCpWrKg2bdqkev4ff/xRAwcO1IgRI/Tkk0/qzz//1AcffKCwsDDNnj3b1k/fLmxZGw0ePFg3b97UzJkzlSdPHq1evVrdunXTp59+KqPRqF69eqlLly6aNm2avv/+ew0bNky5c+dWvXr10vNUAAAAbCJNDaeJEydm6Vv1AQDgzGrVqqVixYppx44d6t+/v6R/Gk4vvPCCXFxctHTpUtWvX1+9e/eWJAUGBmrGjBlq1qyZ9u/frzp16kiSGjRoYFoAu1ixYlq9erUOHTr02IbTwoUL9fLLL+vVV1+VJBUvXlxjx47VG2+8ob/++ktFixa10TO3H1vVRhcvXtTPP/+sdevWqUaNGpKkUaNG6aefftLOnTt169YtlStXToMGDZIkBQUF6cSJE1qyZAkNJwAA4FDS1HD67x1oEhISdObMGZUvX16SdPPmTZ04cUL169eXm5vFy0IBAIAMMBgMatOmjXbu3Kn+/fvrxIkTOnfunObPny9JOnHihC5evKhq1aqZPfb8+fOmhlNQUFCSbX5+fmm6s82JEyd09OhRbd682TSWuJD2+fPns2TDyVa1UUBAgBYvXqxKlSqZxgwGgwwGgyIjI3XgwAE1a9YsyWPq1q2rCRMmyGg08gYhAABwGBZ3h65fv67u3bsrOjpaX3/9taR/Cs1evXqpatWqWrhwofz9/a2dEwCALC0iIsI0e0X6v4ZNSneWSVxbKVFoaKjmzZunY8eO6YsvvlD16tVVokQJSf80Q5577jnTDKd/y507t+nvHh4eZtuTuwPbfyUkJKh79+4KDQ0125YvX77HPt7ZWbM2ypkzpxo3bpxkbNeuXbp48aLeffddffrppypYsGCS7fnz51d0dLTu3LmT5OsJAABgTxY3nKZOnaqYmBhNnz7dNNa4cWNt3bpVgwcP1owZM/TBBx9YNSQAAFnd0qVL9eOPP5oWgL53756kfxpCV69e1d9//51k/4sXLyb5vEiRIqpTp4527dqlL7/8Uv369TNtK1OmjM6dO2dqQEn/zDyaNm2aBg8eLD8/vwxlL1OmjMLCwpIcf9++fVq1apXGjBkjHx+fDB3f0dmyNjp06JBGjBih5s2bq0mTJnr48KFZYzDx84ws0m40GhUVFZXux/9XdHR0kj8dHXltx5mySv/8HHl7e1v0mOjo6DQ159PKYDBYnOHRo0dWzWALzva9YKu8zEZFdmJxw+mXX37RuHHjVLVq1STjwcHBeuuttzRhwgRrZQMAINt48skntWzZMn366aeqWbOmli9frpw5c6patWpyd3fX3bt3tXTpUj3zzDPas2ePfvzxR7NZM6GhoRo3bpwSEhLUsmVL03jXrl3VoUMHjR07Vq+//roiIyM1duxYPXz4UIGBgRnO3qNHDw0cOFDz5s3Ts88+q2vXrmnkyJEqWrRotpjhZKva6Ouvv9bQoUNVvXp1UzPL09PTrLGU+Lmlv6D+W2xsrE1u+hIeHm71Y9oSeW3HWbJ6e3tbfLVGWFiYVZsS3t7eCg4OtugxV65ccZpGjrN8LySyRd7kZhQDWZHFDaeYmJgUp/d7e3ubvQMLAIA9FfMr5hTnbNiwoYYPH665c+cqIiJCZcuW1YIFC+Tr66u6detqwIABWrZsmebMmaNGjRrpzTff1KpVq5Ic45lnntG4cePUrFkz+fr6msarVq2qJUuWaPbs2QoNDZWPj4/q1aund955xypFb4sWLfThhx9q0aJFpsvHQkJCNHTo0Awf2xnYojZas2aNJkyYoBYtWmjKlCmmr1OhQoV048aNJPveuHFDPj4+GZqp5u7urtKlS6f78f8VHR2t8PBwBQYGZqgRllnIazvOlFVK30zBkiVLWn2Gk6UKFy7s8E0MZ/tesFXec+fOWe1YgKOzuOFUpUoVLV++XA0bNpS7u7tpPC4uTqtWrVLlypWtGhAAgPRKMCZoSM237XZuF4OLRY/p0qWLunTpkuy2/v37m+5Al6hz585JPvf29tahQ4eSfXy9evVSvYvZt99+aza2evVq09/btm2bZKHs06dPJ9m3ZcuWSWZVZSfWro3WrVunDz74QB07dtTIkSOT/PJZs2ZN7d+/P8n+v/76q6pXry4XF8u+3/7NYDDY5NJHb29vp7qkkry24yxZ09PscYTmiaenp0PkSAtn+V5IZO28XE6H7MTihtObb76pjh07qmnTpmrUqJHy5Mmj27dv6+eff9atW7eSFKcAANiTpQ2frHJuZC5r1kZhYWGaOHGinn76afXq1UsRERGmbV5eXurYsaNCQ0M1ffp0hYaG6ocfftD//vc/LVmyxBZPDQAAIN0sbjhVrVpVGzZs0MKFC/X999/r7t278vPzU82aNdW3b19VqFDBFjkBAAAckjVro127dik2Nla7d+/W7t27k2wLDQ3V5MmTNX/+fE2bNk0rV65U0aJFNW3atFRnrwEAANiDxQ0n6Z9FMOfMmWPtLAAAAE7JWrVR79691bt371T3adSokRo1apThcwEAANgS8/0BAAAAAABgVTScAAAAAAAAYFU0nAAAAAAAAGBVNJwAAAAAAABgVRY3nLp166a9e/faIgsAAIDToTYCAAAwZ3HD6dChQzIYDLbIAgAA4HSojQAAAMxZ3HBq2LChduzYodjYWFvkAQAAcCrURgAAAObcLH2Ap6enduzYoS+//FJBQUHy8fFJst1gMGjlypVWCwgg6zPGx8vg6vrYMcBSCQkJcnGxz3KFGTn3jh07tGbNGp05c0YGg0GlSpXSSy+9pFdffTVNj9+6datGjBih06dPp+v8kvTqq6/q8OHD2r59u8qXL//Y/YcPH67Lly9r9erVaTr+f/c/e/asLl++rCZNmqQ7s71QGwEAAJizuOF07do1VatWzfS50WhMsv2/nwPA4xhcXXW7/wDFnT0nSXIrU1q55821cypkBS4uLvpm5h7dvRSZqef1L5ZTTQc3SNdjN2/erAkTJmjkyJGqUaOGjEajfv75Z40fP14RERHq37+/ldOaCwsL0+HDhxUYGKhPPvlEY8eOtfk5e/XqpdDQUKdsOFEbAQAAmLO44ZTWdy4BwBJxZ88p9vhxe8dAFnT3UqQiLty2d4w0W7dundq1a6cXX3zRNFaqVCldv35dq1atypSG05YtW1SqVCm1bdtWCxYs0LBhw5QjRw6bn9dZURsBAACYS/d1Bvfu3dM333yjTz75RLdv39aFCxd4Bw8AgAxycXHR4cOHde/evSTjPXv21IYNGyRJISEhmjs36SzA5MY2btyohg0bqkqVKurdu7cuX7782PPHx8dr+/btql+/vpo3b66///5bn332WZJ9jEaj5s+fr0aNGqlq1aoaMWKEHj16ZNr+119/qVy5ctq3b1+qY//OfvnyZc2bN08dO3Z8bEZHRW0EAADwf9LVcFqwYIEaN26sfv36ady4cbp69aomTZqkl156SZGRmXvZAgAAWUn37t114sQJNWrUSD179tTixYt19OhR+fn5qWTJkhYda/Xq1Zo9e7bWrl2rO3fuqF+/fo9tgPz000+6ceOGWrRooRIlSuiJJ54wNboSLV68WEuWLNGwYcO0detW5cyZU1988YXFzzXR5s2bVbBgQXXt2tWsaeYsqI0AAACSsrjhtGbNGs2dO1ddunTRxo0bTYXr66+/rkuXLmn27NlWDwkAQHbRokULffLJJ2ratKl+//13zZgxQy+99JJatGihgwcPWnSsadOmqXr16qpYsaKmTJmikydPau/evak+ZuvWrSpYsKBq1KghSWrdurX++OMPHT16VNI/s5tWr16tTp06qXXr1ipVqpRGjBihChUqpO8JS8qdO7dcXV3l4+Mjf3//dB/HXqiNAAAAzFnccFq9erV69uypt956S0888YRpvHHjxho4cKC+/fZbqwYEACC7qVq1qmbOnKm9e/fq008/1cCBA/XgwQP16NFDt27dStMxcuTIkeTucoGBgcqVK5fOnDmjHTt2qFq1aqaP7t27S5Ju376tb7/9Vi1btpTBYJAktWrVSgaDQevXr5ck3blzRzdv3lSlSpXMMmdX1EYAAADmLF40/MqVK6pdu3ay20qVKqWIiIgMhwIAIDu6du2aFi1apF69eqlgwYJycXFRcHCwgoOD1axZM7Vu3Vq//fZbso+Ni4tL8rmrq6vZPgkJCfLw8FBISIiqVKliGvfy8pIk7dy5U7GxsVq5cqVWrVpl2m40GvXFF19oxIgRpkbUfy/Nc3NLvaSIj49PdbszozYCAAAwZ/EMp0KFCunw4cPJbjt+/LgKFSqU4VAAAGRHHh4e2rRpk3bs2GG2LWfOnJKkvHnzyt3dXQ8ePDBte/DggdnMp8jISP3555+mz0+fPq379++rbNmy8vX1VYkSJUwfBQoUkPTP5XRly5bV9u3btW3bNtPHmDFjFB0dre3btysgIECFChUyu7zv+L/uMunu7m7KlSg8PDydr4rjozYCAAAwZ/EMpxdffFFz586Vl5eXmjRpIkmKiorSrl27tGjRInXp0sXaGQEAyBZy586t7t27a/bs2fr777/VokUL+fr66ty5c5o/f77q1KmjmjVrqmrVqvriiy/0zDPPKGfOnJozZ47ZjCYXFxcNHDhQ77//viRp9OjRql27tmrWrJnsuf/44w+dOnVK48aNU9myZZNsK126tJYuXaoNGzbo9ddfV48ePTRlyhSVKlVKNWvW1Pbt23X06FHTuk/58+dXkSJFtHLlSgUGBuru3buaPXu2aXZUcnLkyKHw8HBFREQob968GXkZMx21EQAAgDmLG049evTQX3/9penTp2v69OmSpE6dOkmSnnvuOfXq1cu6CQEAyAD/Yjmd6pwDBw5UYGCgNm7cqLVr1+rhw4cqXLiwWrZsafo/dvDgwbp79666dOkiPz8/de3a1exOaLlz59YLL7ygvn37Kjo6Wk899ZTee++9FM+beLe5559/3mybi4uL3njjDY0fP14HDhxQhw4dlJCQoAULFigiIkINGzbUiy++qLCwMEmSwWDQ1KlTNXHiRL3wwgsqUaKERowYoZ49e6Z4/o4dO2rKlCk6e/ZssjO8HBm1EQAAgDmLG04Gg0Hjxo1Tly5dtG/fPt29e1d+fn6qVauW2TuiAADYU0JCgpoObmC3c7u4WHzluiSpTZs2atOmTYrbCxQooEWLFiUZ69q1q+nvbdu2Vdu2bSVJb7zxRprOOWrUKI0aNSrF7R07dlTHjh1T/Py/atasqa1btyYZO3HihOnvkydPTrLt5Zdf1ssvv5ymrI6G2ggAAMCcxQ2nRCVLlpTBYNC9e/eUO3duFStWzJq5AADIsPQ2fJz93LAPaiMAAID/k66G0/r16zV//nzdvHnTNFa4cGENGTJErVq1slo4AAAAZ0BtBAAAkJTFDadPPvlEY8eOVdOmTdW8eXPlyZNHERER+t///qchQ4bIw8NDzZo1s0VWAAAAh0NtBAAAYM7ihtOKFSvUvn17011vErVp00bvv/++PvroI4oqAACQbVAbAQAAmLN4gYlr166padOmyW575plndOHChQyHAgAAcBbURgAAAOYsbjhVqlRJP/30U7LbDh8+rHLlymU4FAAAljIajfaOABty5K8vtREAAIA5iy+p69OnjwYPHqy///5bL7zwggoUKKA7d+7om2++0fLly/Xuu+/qt99+M+1fq1YtqwYGAODf3N3dJUlRUVHy9va2cxrYSlRUlKT/+3o7EmojAAAAcxY3nLp16yZJ2rRpkzZv3mwaT3zncezYsabPDQaDTp48aY2cAAAky9XVVf7+/rpx44YkycfHRwaDwc6pYC1Go1FRUVG6ceOG/P395erqau9IZqiNAAAAzFnccFq1apUtcgAAkG4FCxaUJFPTCVmPv7+/6evsaKiNAAAAzFnccKpdu7YtcgAAkG4Gg0GFChVS/vz5FRsba+84sDJ3d3eHnNmUiNoIAADAnMUNJwAAHJWrq6tDNyYAAACA7MLiu9QBAAAAAAAAqbF7w+nWrVt6++23VbduXVWrVk09e/bU+fPnTdtPnjyp119/XVWrVlVISAjrJAAAAAAAADg4uzec+vXrp4sXL2rx4sXavHmzvLy81LlzZ0VHR+vOnTvq0qWLihcvri1btqhfv36aPn26tmzZYu/YAAAASIHBYJC7u7u9YwAAADuyuOH0/vvv6/fff7fKye/du6ciRYpo/Pjxqly5soKCgtS3b1/duHFDZ8+e1caNG+Xu7q5x48YpKChI7dq1U+fOnbV48WKrnB8AACCjrFkbOTpjfHya9vP29lbFChVkMBhsnAgAADgqixcN//XXX7Vp0yYFBgaqbdu2euGFF5Q/f/50nTxXrlyaMWOG6fPbt29rxYoVKliwoEqXLq25c+eqdu3acnP7v5h169bVokWLFBERobx586brvAAAANZizdrI0RlcXXW7/wDFnT2X6n5uZUor97y5UlxcJiUDAACOxuKG01dffaWDBw9q27ZtWrx4sWbNmqX69eurbdu2atq0abqnT48aNUobN26Uh4eHFixYIB8fH127dk1ly5ZNsl9iAXf16tV0N5yMRqOioqLS9diUREdHJ/nTkTlqVoPBIG9v7xS3P3r0SEajMRMTpY+jvr7JiY6Olru7u2JiYpJ9FzozXu/Uvu7R0dFJMjjba/vvPx2drfIajUZmOCDLs1Vt5Kjizp5T7PHj9o4BAAAcnMUNJ0mqUaOGatSooffee0+7d+/W//73P40YMUIeHh567rnn9PLLL5s1ih7njTfe0CuvvKK1a9eqX79+WrdunR4+fCgPD48k+3l6ekr6p/mQXrGxsTp58mS6H5+a8PBwmxzXFhwtq7e3t4KDg1PcfuXKFaf55V1yvNc3Oe7u7gp+Ilhurub/FMTFx+nEHycUGxtr0wypfd3DwsKS/Zo7w2ubyJmySrbJ+99/x4GsyBa1EQAAgDNLV8Mpkaenp2rVqqWIiAhdu3ZNx48f186dO7V27Vo1aNBA48ePV4ECBdJ0rNKlS0uSJkyYoN9//11r1qyRl5eXYmJikuyX2Gjy8fFJd253d3fT+awlOjpa4eHhCgwMTHWWjiNw1KyPmwVRuHBhp/jF1VFf3+TExMTIzdVNMw5M06X7l0zjxfyKaUjNt1WmTBmbz3JK7etesmRJsxlOzvLaOlNWyXZ5z51L/bIbIKuxZm0EAADgzNLVcIqKitKuXbu0fft27d+/Xz4+PmrZsqVGjRqlKlWq6Pfff9egQYP01ltvaf369Ske5/bt29q7d6+eeeYZ0zpNLi4uKl26tG7cuKGCBQvqxo0bSR6T+HlGijWDwZChhlVqvL29bXZsa3OmrNI/Rbwz/OKeyBle38Rmz6X7l3Th3nmz7fZ+vVM6vzO8tomcKatk/bxcTofswlq1EZDVcMdCAMi+LG44DR06VN98842io6NVo0YNTZgwQS1atEjyi2GVKlXUpk0brVixItVjRUREaPDgwVqyZIkaNmwo6Z/L3U6cOKGQkBDlzZtX69evV3x8vFxdXSX9szBnyZIllSdPHkujAwAAWJ01ayMgrdzd3e3a1DfGx8vw/+vz1CTesfARC8gDQLaTrrvUvf7662rXrp0CAwNT3K9u3boqV65cqscqW7asGjVqpPHjx2v8+PHKlSuXFi1apMjISHXu3Fmenp5asmSJRo4cqe7du+vo0aNasWKFxo4da2lsAAAAm7BmbQSkhcFgUHDwE3Jze3zDR5LiE4xydbFuc4o7FgIAHsfihlOTJk3UrFmzVAsqSapdu3aajjdz5kzNmDFDgwYN0v3791WzZk2tXbtWhQsXliQtWbJEEyZMUGhoqPLly6dhw4YpNDTU0tgAAAA2Ye3aCEgLNzdXjd58VOERD1LdLzCvr8a+WNkmGbhjIQAgNRY3nHbu3KlWrVpZLYCfn5/GjBmjMWPGJLu9cuXK2rBhg9XOBwAAYE3Wro2AtAqPeKDTV+/bOwYAAMlysfQB1apV06+//mqLLAAAAE7HlrXRokWL1LFjxyRj7733nsqVK5fkIyQkxCbnzy5Y2BoAAOuzeIZTuXLltGzZMu3atUvly5c3u5uRwWDQxIkTrRYQAADAkdmqNlq7dq1mzZqlmjVrJhk/ffq0evfurddff9005pqGxZuzIxa2BgDAfixuOO3evVv58+dXbGysjh07ZradW2ADAIDsxNq10fXr1zV69Gjt27fPbF0oo9Goc+fOqWfPnsqXL19GYmcLLGwNAID9WNxw+vbbb22RAwAAwClZuzb6448/5O7urh07duijjz7S5cuXTdv+/PNPRUVFqVSpUlY9Z1bGwtYAANiHxQ2nRJGRkTpy5Iju37+v3Llzq1KlSvL19bVmNgAAAKdhrdooJCQkxTWZzpw5I0lavXq1fvzxR7m4uKhRo0YaNGiQ/Pz80p3daDQqKioq1X0MBoO8vb0tOu6jR49kNBrTnSsjnC2vJWJiYix+btHR0VZ7bry2SVnztZWy7usbHR2d5E9HZ6u8RqORq4KQbaSr4bR48WLNnz9fDx8+NI15eHioV69e6tevn9XCAVlJQkKCXFxcHjsGAHA+mVUbnTlzRi4uLsqfP78WLlyoP//8U1OnTtXZs2e1cuXKdP+fEhsbq5MnT6a6j7e3t4KDgy067pUrV+z2y6Wz5bWEt7e3/P39LXpMWFiY1Z4br21S1nxtEzNk1ddXksLDw+0dwSK2yOvh4WH1YwKOyOKG05YtWzRz5ky9+OKLev7555U3b17dvHlT27dv17x581S4cGGFhobaIivg1FxcXPTNzD26eylSkuRfLKeaDm5g51QAgIzKzNqoT58+at++vQICAiRJZcuWVb58+fTyyy/r2LFjqlKlSrqO6+7urtKlS6e6T3rekS9cuLDdfrFytryWiImJsfgxJUuWtOoMJ0sVK1ZMbm5p+9XDnjN17P3aSln3ezc6Olrh4eEKDAy0eAaXPdgq77lzqa8pB2QlFjecVqxYoddee02jR482jZUqVUp16tSRl5eXVq1aRcMJSMHdS5GKuHDb3jEAAFaUmbWRi4uLqdmUqEyZMpKka9eupbvhZDAYzO6uZw2enp5O8YtlImfJm56GhL2el0u+fIpPMKb5ks/4BKNcXex3uZEzvbb/5izfu9I/r5ct/r2xFWvn5XI6ZCcWN5wuXryo4cOHJ7utadOm2rJlS4ZDAQAAOIvMrI2GDRumGzduaMWKFaaxxDvjPW6GEmAPLrlyytXFoNGbjyo84kGq+wbm9dXYFytnUjIAgK1Z3HAqUKCArly5kuy2v/76i4XDAQBAtpKZtdEzzzyjvn37at68eXr++ecVFhamcePGqXXr1goKCrLaeQBrC494oNNX79s7BgAgE1m8smRISIhmz56to0ePJhn//fffNXfu3BTvqgIAAJAVZWZt1LRpU82aNUvffPONnnvuOY0cOVLNmzfXxIkTrXYOAAAAa7B4htOAAQP0yy+/6JVXXlGRIkWUN29eRURE6PLlywoKCtKQIUNskRMAAMAh2bI2mjx5stlYy5Yt1bJly4xEBgAAsDmLG06+vr7avHmztmzZot9++0337t1TpUqV1LVrV7Vt21ZeXl62yAkAAOCQqI0AAADMWdxwkv65C0L79u3Vvn17a+cBAABwOtRGAAAASVnccBoxYkSK21xcXOTj46PAwEC1atXK7La9AAAAWQ21EQAAgDmLG07Xrl3ToUOH9OjRI9M6Bbdu3dJff/0lV1dX0+cLFizQJ598omLFitkiNwAAgEOgNgIAADBn8V3qnnrqKfn5+Wn9+vX65ptvtGHDBn399dfaunWrChQooL59+2rv3r0qXry4Zs6caYvMAAAADoPaCACyJnd3dxkMBnvHAJyWxQ2nFStWaMiQIapatWqS8eDgYL311ltatGiR/Pz81KVLF+3bt89aOQEAABwStREAZD0Gg0HBwU/I29s7TfvHJxhtnAhwPhZfUnfnzh3lzp072W25cuXSrVu3JEkBAQGKiorKWDoAAAAHR20EAFmTm5urRm8+qvCIB6nuF5jXV2NfrJxJqQDnYXHDKTg4WEuWLFG9evXk4eFhGo+JidGyZctUoUIFSdIff/yhQoUKWS8pAACAA6I2AoCsKzzigU5fvW/vGIBTsrjhNHToUHXp0kVNmzZV48aNlSdPHkVEROjHH3/UgwcPtGTJEh04cEAzZ85Unz59bJEZAADAYVAbAQAAmLO44VStWjVt3bpVCxcu1E8//aTbt2+rYMGCatiwoXr37q3ixYtr7969evPNN9WtWzdbZAYAAHAY1EYAAADmLG44SVKpUqU0derUFLfXq1dP9erVS3coAAAAZ0JtBAAAkFS6Gk4xMTHavHmzfvnlF928eVMTJ07U/v379cQTT6hyZRZLAwAA2Qu1EQAAQFIulj7g9u3bateunSZMmKCLFy/q6NGjevjwob777jt17NhRhw8ftkVOuzHGx6dpDAAAZE/ZrTYCAABIC4sbTlOnTtXff/+tL774Qp9++qmMRqMkae7cuapUqZLmzJlj9ZD2ZHB11e3+A3TjmZa68UxL3e4/QAZXV3vHAgAADiK71UYAAABpYXHD6bvvvtNbb72lEiVKyGAwmMY9PT3VtWtX/fHHH1YN6Ajizp5T7PHjij1+XHFnz9k7DgAAcCDZsTYCAAB4HIsbTo8ePZK/v3+y21xdXRUbG5vRTAAAAE6D2ggAAMCcxQ2nSpUqad26dclu27lzpypWrJjhUAAAAM6C2ggAAMCcxXepe+utt9S5c2e98MILaty4sQwGgz777DPNnTtXe/bs0ZIlS2yREwAAwCFRGwEAAJizeIZTzZo1tXz5cnl7e2vJkiUyGo1asWKFbt68qUWLFqlu3bq2yAkAAOCQqI0AAADMWTzDSZJq1aql9evX6+HDh7p37558fX2VI0cOSVJ8fLxcuYsbAADIRqiNAAAAkrJ4hlPTpk116tQpSZKXl5cKFChgKqiOHj2qJ5980roJAQAAHBi1EQAAgLk0zXD67LPPFBcXJ0m6fPmydu/ebSqs/m3v3r3ciQUAAGR51EYAAACpS1PD6dixY1q5cqUkyWAw6KOPPkpx3y5dulgnGQAAgIOiNgIAAEhdmhpOQ4YMUadOnWQ0GtWsWTPNmzdPFSpUSLKPq6urfH195evra5OgAAAAjoLaCAAAIHVpajh5eHioSJEikqRvvvlG+fPnl7u7u02DAQAAOCpqIwAAgNRZfJe6IkWK6OjRo9q3b59iYmJkNBolSUajUVFRUTp48KA2btxo9aAAAACOiNoIAADAnMUNp7Vr12r8+PGmYurfXFxc1KBBA6sEAwAAcAbURgAAAOZcLH3AmjVr1KhRI+3bt09du3bVyy+/rCNHjmj27Nny9PTU888/b4ucAAAADonaCAAAwJzFDae//vpL7du3V65cuVSxYkUdPHhQXl5eeuaZZ9SzZ0+tWrXKFjkBAAAcErURAACAOYsbTu7u7vLy8pIklShRQhcvXlRsbKwkqUaNGgoPD7dqQAAAAEdGbQQAAGDO4oZThQoV9N1330mSSpYsqYSEBP3++++SpGvXrlk3HQAAgIOjNgIAADBn8aLhXbp0Uf/+/RUZGamJEyeqadOmGjZsmJo3b66dO3eqRo0atsgJAADgkKiNAAAAzFk8w6lZs2ZauHChgoKCJEnjxo1TYGCg1q9fr1KlSun999+3ekgAAABHRW0EAABgzuIZTpLUpEkTNWzYUJIUEBCgjz76SHFxcfLz87NqOAAAAGdAbQQAAJCUxTOcYmNjNXr0aL388sumscOHD6tevXqaMmWKEhISLDre3bt39f7776tRo0aqXr26XnvtNR04cMC0fe/evWrbtq2qVKmiFi1a6PPPP7c0MgAAgM1YuzYCAADICixuOM2dO1c7duxQ69atTWPBwcEaOnSoNm7cqCVLllh0vMGDB+vw4cOaOXOmtmzZogoVKqhbt266cOGCzp8/r169eqlhw4baunWrXnrpJQ0bNkx79+61NDYAAIBNWLs2AgAAyAosvqRu586deuedd/Tqq6+axvz9/dW5c2e5ublp1apV6tmzZ5qOdfHiRf38889at26daUHNUaNG6aefftLOnTt169YtlStXToMGDZIkBQUF6cSJE1qyZInq1atnaXQAAACrs2ZtBAAAkFVYPMPpzp07KlasWLLbSpUqZdHtfwMCArR48WJVqlTJNGYwGGQwGBQZGakDBw6YNZbq1q2rgwcPymg0WhodAADA6qxZGwEAAGQVFs9wKlWqlHbt2qX69eubbfv2229VokSJNB8rZ86caty4cZKxXbt26eLFi3r33Xf16aefqmDBgkm258+fX9HR0bpz545y585taXxJktFoVFRU1GP3MxgM8vb2TnZbdHR0kqZXdHR0kj8dmaNmTe31lqS4uDizzI7YeEzu9bXkeykzxcTEpPqaZ0Y2fs4cg63yGo1GGQwGqx4TcDTWrI0AAACyCosbTp06ddLw4cN19+5dNWvWTHny5NHt27f13Xff6csvv9SkSZPSHebQoUMaMWKEmjdvriZNmujhw4fy8PBIsk/i5zExMek+T2xsrE6ePPnY/by9vRUcHJzstrCwsGR/MQsPD093rszmaFlTer1z+3oowZhgdqefuPg4nfjjhGJjYzMrokX+/fqm53spM3h7e8vf3z/F7ZmRjZ8zx2KLvP/9dxzIamxZGwEAADgrixtObdq00d9//6358+frq6++Mo0HBARo1KhRatOmTbqCfP311xo6dKiqV6+u6dOnS5I8PT3NGkuJn6c2K+Nx3N3dVbp06cful9q78iVLljSbeREeHq7AwMAMZcsMjpo1pdfbz8tdLgYXzTgwTZfuX5IkFfMrpiE131aZMmUcbpZTcq+vJd9LmelxjdvMyMbPmWOwVd5z585Z7ViAo7JVbQQAAODMLG44SVKHDh3Uvn17hYWF6e7du8qZM6dKlSolFxeLl4SSJK1Zs0YTJkxQixYtNGXKFNO74YUKFdKNGzeS7Hvjxg35+PiYzXaxhMFgkI+PT7ofL6Xc8PL29s7wsTOLM2WVpEv3L+nCvfNJxhz5F/m0vr72fA6Pu9TJ3q8vP2eZz9p5uZwO2YW1ayMAAABnl66Gk/TPLxGlSpXKcIB169bpgw8+UMeOHTVy5Mgkv5zUrFlT+/fvT7L/r7/+qurVq1PAAQAAh2Kt2ujfFi1apD179mj16tWmsZMnT2rChAk6fvy4cufOrc6dO6tTp05WPS8AAEBG2bVrExYWpokTJ+rpp59Wr169FBERoZs3b+rmzZu6f/++OnbsqKNHj2r69Ok6f/68li1bpv/973/q3r27PWMDAADY3Nq1azVr1qwkY3fu3FGXLl1UvHhxbdmyRf369dP06dO1ZcsW+4QEAABIQbpnOFnDrl27FBsbq927d2v37t1JtoWGhmry5MmaP3++pk2bppUrV6po0aKaNm2a6tWrZ6fEAAAAtnX9+nWNHj1a+/btU2BgYJJtGzdulLu7u8aNGyc3NzcFBQXp4sWLWrx4sdq1a2efwFkIlwEDAGA9dm049e7dW7179051n0aNGqlRo0aZlAgAAMC+/vjjD7m7u2vHjh366KOPdPnyZdO2AwcOqHbt2nJz+78Srm7dulq0aJEiIiKUN29ee0R2ei758ik+wSgvL6807R+fYJSrC80pAABSk6aGU+vWrTVlyhQ98cQT2rZtmxo3bqyAgABbZwMAAHBItqyNQkJCFBISkuy2a9euqWzZsknG8ufPL0m6evVquhtORqNRUVFRqe5jMBgsvpnEo0eP7HYnVkvyuuTKKVcXg0ZvPqrwiAep7huY11djX6ys6Ohou95l1tKvhTXzpud7wRLZ+bWVnO9nLa2io6OT/OnobPW9YDQamU2JbCNNDaeLFy/q9u3bkqQRI0Zow4YNNJwAAEC2Za/a6OHDh6a7+Sby9PSU9M8vnOkVGxurkydPprqPt7e3goODLTrulStX7PbLZXryhkc80Omr99O0b1hYmF2fm7+/v0WPsWbe9Ly2lsjOr21iBmf6WbNUeHi4vSOkiS2/F/777ziQVaWp4RQUFKShQ4eqbNmyMhqNGjNmjHx9fZPd12AwaOXKlVYNCQAA4EjsVRt5eXkpJiYmyVhio8nHxyfdx3V3d1fp0qVT3Sc978gXLlzYbr9Y2XoGQcmSJe06C8dS1szLa5uUtfM6289aWkVHRys8PFyBgYE2nSFnLbb6Xjh37lx6IwFOJ00Np6lTp2revHm6e/eu6R/AlH6QHH0qJwAAQEbZqzYqWLCgbty4kWQs8fMCBQqk+7gGgyFDDauUeHp6OsUvlulhz+eVnoaEM30deG0t50w/a97e3jb598babPW9wOV0yE7S1HAqW7as5syZI0kqX768xowZo8qVK9s0GAAAgKOyV21Uq1YtrV+/XvHx8XJ1dZUk/frrrypZsqTy5Mlj8/MDWR3NAACwHhdLH3Dq1ClTQXX+/HkdOXJEFy9etHowAAAAZ5CZtVG7du304MEDjRw5UufOndPWrVu1YsUK9erVyybnA7IKY3x8mvZL650KAQCPl6YZTv/12WefacqUKYqIiDCN5c2bV0OGDFGbNm2slQ0AAMApZFZtlCdPHi1ZskQTJkxQaGio8uXLp2HDhik0NNRq5wCyIoOrq273H6C4s6mvn+P5VBPlGv5O5oQCgCzO4obTt99+q7ffflt169bV4MGDlTdvXt24cUM7duzQiBEj5O/vryZNmtggKgAAgOOxZW00efJks7HKlStrw4YNGUwNZD9xZ88p9vjxVPdxKx2USWkAIOuzuOG0YMECtWjRQh9++GGS8Xbt2mnQoEFatGgRDScAAJBtUBsBAACYs3gNpzNnzqQ4bTs0NFSnTp3KcCgAAABnQW0EAABgzuKGU0BAgO7du5fstrt378rDwyPDoQAAAJwFtREAAIA5ixtO9erV07x583Tt2rUk41evXtVHH32k+vXrWy0cAACAo6M2AgAAMGfxGk6DBw9Wu3bt1Lx5c1WrVk158+ZVRESEDh8+rFy5cmnIkCG2yAkAAOCQqI0AAADMWTzDKV++fPr000/VsWNHRUdH6/jx44qOjlbHjh316aefqkiRIrbICQAA4JCojQAAAMxZPMNJkvLkyaO3337b2lkAAACcErURAABAUhbPcELWlpCQkKYxAAAA2IfBYLB3BAAAHitdM5yQdbm4uOibmXt091KkJMm/WE41HdzAzqkAAACyPmN8vAyuro/dz8vLKxPSAACQMTScYObupUhFXLht7xgAAADZisHVVbf7D1Dc2XOp7uf5VBPlGv5O5oQCACCdaDgBNuDu7s50dwBAtsf/hZaLO3tOscePp7qPW+mgTEoDAED60XACrMxgMCj4iWC5ufLjBQDInlzy5VN8gjHNl37FJxjl6kJzCgCArMTi34gfPnyoBQsW6LvvvlN0dLTZgtIGg0Fff/211QICzsjN1U0zDkzTpfuXJEk1CtRQx+A37JwKAGAL1EbmXHLllKuLQaM3H1V4xINU9w3M66uxL1bOpGQA7MVgMMjd3d3eMQBkIosbThMmTNDmzZtVu3ZtVahQQS4u3OgOSM6l+5d04d55SVJR36J2TgMAsBVqo5SFRzzQ6av37R0DgA2ldbF7b29vVaxQQY/i4jIhFQBHYHHD6auvvtKgQYPUs2dPW+QBAABwKtRGALKztC5271amtHLPmyvRcAKyDYsbTrGxsapcmWnPAAAAErURAKRlsXsA2Y/Fc74bNGigH3/80RZZAAB2xN0VgfShNgIAADBn8QynVq1aafTo0bp9+7aqVKkib29vs33atGljjWywUErXT6f1umpYhzP9wu5MWZ2Nsy2MmdLdFROMCXIxsB4NkBpqIwAAAHMWN5wGDhwoSdq2bZu2bdtmtt1gMFBU2Uly10+brpWGTSR3G+e03gLaHv6b15GzOpv/vrbe3t4KDn5CsbExdkxlmf/eXbGYXzENqfm2nVMBjo/aCAAAwJzFDadvvvnGFjlgJVw/nbmSu+VzvdL51LtZGTumStl/8zpyVmfz39c28TbfsbF2Dmahf99dEUDaUBsBAACYs7jhVKRIEdPfo6Oj9eDBA/n7+zvVpSOANf33ls8l8uawY5rH+3deR8/qbLj9N5A9URsBAACYs7jhJEkHDhzQ1KlTdfz4cRmNRklS5cqVNWjQINWtW9eqAQEAABwdtREAAEBSFjecDh06pM6dO6tYsWLq27ev8ubNqxs3bujzzz9X9+7dtXr1alWrVs0WWQEAABwOtREAAIA5ixtOs2bNUs2aNbV06VK5/uvOZ/3791e3bt00d+5cLVu2zKohAQAAHBW1EQAAgDmL73V97NgxderUKUlBJUkuLi56/fXXdfToUauFAwAAcHTURgAAAOYsbjjlyJFDcXFxyW6Li4szrVsAwHn5ewYoISHBbDy5MQDI7qiNAAAAzFl8SV316tW1ePFiNWzYUN7e3qbxqKgoLV68WDVr1rRqQACZz9c9h1xcXPTNzD26eylSkuRfLKeaDm5g52QA4HiojQAAAMxZ3HAaMmSI2rZtq6ZNm6pJkybKly+fbt68qe+//14PHz7UhAkTbJETgB3cvRSpiAu37R0DABwatREAAIA5ixtOJUqU0IYNGzRv3jz98MMPunfvnnLlyqXatWurf//+Kl26tC1yAgAAOCRqIwAAAHMWN5wkqXTp0po1a5aVowAAADgnaiMAAICk0tRw2rZtmxo3bqyAgABt27btsfu3adMmg7EAAAAcF7URAKSPwWCwdwQAmSRNDafhw4dr48aNCggI0PDhw1Pd12AwUFQBAIAsjdoIACzjki+f4hOM8vLyStP+8QlGubrQnAKcWZoaTt98843y5ctn+jsAAEB2Rm0EAJZxyZVTri4Gjd58VOERD1LdNzCvr8a+WDmTkgGwlTQ1nIoUKWL6+2+//WaaQv5fN2/e1LZt29SjRw/rJQQAAHAw1EYAkD7hEQ90+up9e8cAkAlcLH3AiBEjdOnSpWS3nTx5UnPmzMlwKAAAAGdBbQQAAGAuTTOcevbsqfPnz0uSjEaj+vXrJw8PD7P9bt26peLFi1s3IQAAgIOhNgIAAEhdmhpOvXv31qZNmyRJn376qYKDg5U7d+4k+7i4uChnzpxq27at9VMCAAA4EGojAACA1KWp4VS9enVVr15dkhQdHa0hQ4aoWLFiNg0GAADgqKiNAAAAUmfxGk579+7VwYMHbZFFixYtUseOHZOMnTx5Uq+//rqqVq2qkJAQrVq1yibntpWEhASLxgEAgHOxZW0EAADgrNI0w+nf3N3dk70LS0atXbtWs2bNUs2aNU1jd+7cUZcuXRQSEqKxY8fqyJEjGjt2rHLkyKF27dpZPYMtuLi46JuZe3T3UqRpzL9YTjUd3MCOqQAAgLXYqjYCAABwZhY3nN566y1NnTpV9+/fV/ny5eXj42O2T+HChdN8vOvXr2v06NHat2+fAgMDk2zbuHGj3N3dNW7cOLm5uSkoKEgXL17U4sWLnabhJEl3L0Uq4sJte8cAAAA2YO3aCAAAICuwuOE0ZswYxcfH6+23305xn5MnT6b5eH/88Yfc3d21Y8cOffTRR7p8+bJp24EDB1S7dm25uf1fzLp162rRokWKiIhQ3rx5LY0PABlmjI+XwdXV3jEAOAhr10YAAABZgcUNp/Hjx1s1QEhIiEJCQpLddu3aNZUtWzbJWP78+SVJV69eTXfDyWg0Kioq6rH7GQwGeXt7J7stOjpaRqMxyef//vNxj0/uGBlhybmSy/q4Y1gza0oe9xySkxm5UpKevMnJrOfgqHkz+nOW2RLz3u4/QHFnz0mSPJ9qolzD30l2/0ePHtnte9QSMTExKX4dHj58mOxzSMvzMhqNMhgMGc4HODJr10Zpcf36dTVq1MhsfNKkSdwVDwAAOASLG06hoaG2yJGshw8fysPDI8mYp6enpH9+iUuv2NjYNL3T6O3treDg4GS3hYWFJftLb3h4eJoen9ox0iM95/p31scdw5pZU/K455CczMiVkvTkTU5mPQdHzZvRn7PMlpg37uw5xR4/LklyKx2U4v5Xrlyxa4Msrby9veXv759kzN8zQAkJCfLy8jLbPz4uXn+c+EOxsbGPPfZ//x0HsprMrI0SnTp1Sp6envr666+TNHX9/PwyPQsAAEByLG44SdLt27e1bNky7d+/X5GRkQoICFDNmjXVuXNn5cmTx2rhvLy8FBMTk2QssdGU3PoIaeXu7q7SpUs/dr/U3pUvWbKk2cyL8PBwBQYGmmYJPO5d/f8eIyMsOVdyWR93DGtmTUl6ZkFkRq6UWGvWRmY9B0fNm9Gfs8xm6etYuHBhp2i4/PffWknydc+R6o0PypQp89jvhXPnzlk9K+CIMqs2SnTmzBkFBgaaZn4DAAA4GosbTteuXdMrr7yi27dvq2rVqgoODtbNmze1fPlybdu2TZs3b1aBAgWsEq5gwYK6ceNGkrHEzzNyDoPBkKGGlaQUf9n19vZO87Ez8xfm5M7lqFkt4ai5LOFsz8He37eJ4xn9Gc4snp6eTvE1Tq2RltKND9LyvLicDtlBZtZGiU6fPq2goJRnVwIAANibxQ2nadOmyc3NTV988YWKFStmGr906ZK6du2qDz/8UJMnT7ZKuFq1amn9+vWKj4+X6/9foPfXX39VyZIlbfJuIQAAgKUyszZKdObMGQUEBKhDhw4KCwtTiRIl1KdPn2TXdUqLtKxvaa21AFOSmWsEWkN2zutMWaXskddea0Y622tridTWt0xJWvKyviWyE4sbTnv27NG7776bpKCSpGLFiqlfv36aOnWq1cK1a9dOS5Ys0ciRI9W9e3cdPXpUK1as0NixY612DgAAgIzIzNpIkuLi4nThwgWVLl1aw4cPl6+vrz7//HP17NlTy5cvV7169Sw+ZlrWt7TWWoApycw1Aq0hO+d1pqxS9shrrzUjne21tURy61s+TlrzOsNyC4A1WNxwio+PV0BAQLLbcufOrQcPHmQ4VKI8efJoyZIlmjBhgkJDQ5UvXz4NGzbMLotzAgAAJCczayNJcnNz0759++Tq6mpa1L9ixYo6e/asli5dmq6GU1rWt7T1O/KZuUagNWTnvM6UVcoeee21ZqSzvbZS2jPHxcVZfOy05GV9S2QnFjecypUrp507dyY7ZXv79u0qW7ZsusMkN928cuXK2rBhQ7qPCQAAYEu2rI1SkiNHDrOxMmXKaM+ePek6njXWt8woZ1jv7t/IazvOlFVyjLzOsmakpWzxnIzx8TL8/+VarI31LYGkLG449e3bV926ddO9e/fUqlUr5cuXTzdv3tTnn3+uPXv2aM6cObbICQAA4JAyuzY6e/asXnnlFS1YsEB16tQxjR8/fjxNd+EFgOzM4Oqq2/0HKO5s6jONPJ9qolzD38mcUEAWZXHDqX79+po8ebKmT5+uH3/80TSeN29eTZw4UU8//bRVAwIAADiyzK6NgoKCVKpUKY0bN05jx45VQECANm7cqCNHjmjLli1WPRcAZEVxZ88p9vjxVPdxK82dQIGMsrjhJElt2rTRCy+8oAsXLujevXvKlSuXSpUqxfRAZBp/zwAlJCTIxcXFbFtK48i6+F4AYG+ZWRu5uLho4cKFmjFjhgYOHKjIyEgFBwdr+fLlNrl8DwAAID3S1XCSpJ9//lm//fab7t69qzx58ujJJ59UzZo1rZkNSJGvew65uLjom5l7dPdSpGncv1hONR3cwI7JYA98LwBwBJlZG+XNm1eTJk2yybEBAACsweKG071799SrVy8dOXJEbm5u8vf31927d7VgwQI1atRIc+fO5TaPyDR3L0Uq4sJte8eAA+B7AYC9UBsBAACYs/hak4kTJyosLEzz5s3TsWPHtGfPHh09elSzZ8/WkSNH9OGHH9oiJwA4JS41BrI+aiMAAABzFjecvv/+ew0dOlTNmjUz/SLl4uKi5s2ba9CgQdq5c6fVQwKAs8nt66EEY4K8vLzMtiUYE+yQCICtUBsBAACYs/iSOqPRqLx58ya7rVChQoqKispwKABwdn5e7nIxuGjGgWm6dP+SabyYXzENqfm2HZMBsDZqIwAAAHMWz3AKDQ3VggUL9PfffycZj4uL05o1axQaGmq1cADg7C7dv6QL986bPv7dfAKQNVAbAQAAmLN4hpO3t7fCw8PVtGlTNW3aVAUKFNCdO3f0ww8/6Nq1a8qVK5dGjBgh6Z+1SyZOnGj10MgYg8Egd3d3e8cAACBLoDYCAAAwZ3HDaceOHfL19ZUk7d27N8m2ggUL6tChQ6bPWSzX/lzy5VN8glGuLv/3tfD29lZw8BOKjY2xYzIAALIGaiMA9sa/LQAckcUNp2+//dYWOWAjLrlyytXFoNGbjyo84oEkKTCvr8a+WFmxsXYOBwBAFkBtBMBeEt9cTu4mJf/13zehAcDWLG44JYqMjNSRI0d0//595c6dW5UqVTK9uwfHEx7xQKev3k8yxjshAABYD7URgMyW3JvLyUl8wxkAMlO6Gk6LFy/W/Pnz9fDhQ9OYh4eHevXqpX79+lktHGwjtdu1AwAAy1EbAbCn5N5cBgB7s7jhtGXLFs2cOVMvvviinn/+eeXNm1c3b97U9u3bNW/ePBUuXJi7sTi4lG7XXqNADXUMfsOOyQAAcD7URgAAAOYsbjitWLFCr732mkaPHm0aK1WqlOrUqSMvLy+tWrWKospJJN6uPVFR36KZen5jfLwMrq6Zek4AAKyN2ggAAMCci6UPuHjxopo1a5bstqZNm+rChQsZDoXsweDqqtv9B+jGMy1145mWujd5ir0jAQBgMWojAAAAcxbPcCpQoICuXLmS7La//vqLxTFhkbiz5xR7/Lgkya10kJ3TAABgOWojAAAAcxbPcAoJCdHs2bN19OjRJOO///675s6dq5CQEKuFAwAAcHTURgAAAOYsnuE0YMAA/fLLL3rllVdUpEgR5c2bVxEREbp8+bKCgoI0ZMgQW+QEAABwSNRGAAAA5ixuOPn6+mrz5s3asmWLfvvtN927d0+VKlVS165d1bZtW3l5edkiJwAAgEOiNgIAADBnccOpW7du6t69u9q3b6/27dvbIhOADOIOgACQeaiNAAAAzFnccDp06JAMBoMtsgCwksQ7AMadPWca83yqiXINf8d+oQAgi6I2AgAAMGfxouENGzbUjh07FBsba4s8AKwk8Q6AiR/xly7ZOxIAZEnURgAAAOYsnuHk6empHTt26Msvv1RQUJB8fHySbDcYDFq5cqXVAjoTd3d33uEEkCp/zwAlJCTIxSVpvz+5MQDOgdoIAADAnMUNp2vXrqlatWqmz41GY5Lt//08uzAYDAp+Ilhurha/pACyEV/3HHJxcdE3M/fo7qVISZJ/sZxqOriBnZMBSC9qIwAAAHMWd0dWr15tixxZgpurm2YcmKZL9/+5dKlGgRrqGPyGnVMBcER3L0Uq4sJte8cAYAXURgAAAOYsajgdPXpUly9fVokSJRQcHGyrTE7t0v1LunDvvCSpqG9RO6cBAAC2RG0EAACQvDQ1nCIjI9WrVy8dOXJERqNRBoNB1apV04wZM1SoUCFbZwQAAHAo1EYAAACpS9MKtbNmzdKJEyc0YMAALV68WO+8844uXLig999/39b5AAAAHA61EQAAQOrSNMPpu+++0+DBg/XGG/+sR9SoUSMVKFBAQ4cOVVRUlNndWAAAALIyaiMAAIDUpWmG082bN/XEE08kGatTp47i4+N19epVmwQDAABwVNRGAAAAqUtTwykuLk4eHh5JxnLlyiVJevTokfVTAQAAODBqIwAAgNSlqeGUGqPRaI0cgMMzxsfbOwIAwAlQGwEAAKRxDafUGAwGa+QAHJ7B1VW3+w9Q3NlzkiTPp5oo1/B37BsKAOBwqI0AAAAsaDiNGTNGvr6+ps8T370bNWqUcuTIYRo3GAxauXKlFSMCjiPu7DnFHj8uSXIrHWTnNAAAe6I2AgAASFmaGk61atWSZD5FPLlxppEDAICsjtoIAAAgdWlqOK1evdrWOQAAAJwGtREAAEDqMrxoOAAAAAAAAPBvNJwAAAAAAABgVTScAAAAAAAAYFU0nAAAAAAAAGBVNJwAAAAAAABgVTScAAAAAAAAYFU0nAA4FXd3dxkMBnvHAAAAAACkws3eAQAgrQwGg4KfCJabK/90pZUxPl4GV9c0jwMAAACANTjFb20JCQmaN2+eNm3apPv376tWrVp6//33VaxYMXtHA5DJ3FzdNOPANF26f0mSVKNADXUMfsPOqRyXwdVVt/sPUNzZc6YxtzKllXveXDumApBR1EYAAMDROcUldfPnz9e6dev0wQcfaP369UpISFD37t0VExNj72gA7ODS/Uu6cO+8Ltw7r+t/X7d3HIcXd/acYo8fN338u/kEwDlRGwEAAEfn8A2nmJgYLVu2TG+++aaaNGmi8uXL68MPP9S1a9f01Vdf2TseAABApqI2AgAAzsDhG06nTp3S33//rXr16pnGcubMqeDgYP322292TAYAAJD5qI0AAIAzMBiNRqO9Q6Tmq6++0oABA/T777/Ly8vLNP7WW2/p4cOHWrRokUXHO3TokIxGo9zd3dO0v8FgUMKtWzLGxv7zubu7XPLkUXIvm8Fg0L1H9xRnjJMkebp6ytfdV9H3HiohLsG0n4ubi7xzeSV7jIz4b1ZJMnh5ycXfX3f+jlFc/D/n83R3UU5v9yRZU8prq6zJ5U0ua0p5M/u1tUXezMyaUl6H/V5w4J+zZPM60Wsrpfz62uK1jY2NlcFgUPXq1a30jADYszZK7t+UZPdL4f/I5Li5GhSQwyPT/g1Mdj/yWsyZskrkdaasUtbOS22E7MThFw2Pjo6WJHl4eCQZ9/T01L179yw+XuLt1C25rbpLnjwpHue/cnnmMhvzzuWVzJ6WZUir5LJKUkAOD7Ox5LJKyee11W3ok8ubXFbJ/q+tZJu8mZlVcq7vBUf9OZPS/r3gqK9tSuez9mtrMBhs9ryA7MretVFK/6YkJ6X/I1PLYW3k/Udm1p3JsXdWibySc2WVsmZeaiNkJw7fcEp85y4mJibJu3iPHj2St7e3xcerVq2a1bIBAABkNmojAADgDBx+DadChQpJkm7cuJFk/MaNGypQoIA9IgEAANgNtREAAHAGDt9wKl++vHx9fbVv3z7TWGRkpE6cOKFatWrZMRkAAEDmozYCAADOwOEvqfPw8NDrr7+u6dOnK3fu3CpSpIimTZumggULqnnz5vaOBwAAkKmojQAAgDNw+IaTJL355puKi4vTe++9p4cPH6pWrVpaunRpmu80BwAAkJVQGwEAAEdnMNriXpMAAAAAAADIthx+DScAAAAAAAA4FxpOAAAAAAAAsCoaTgAAAAAAALAqGk4AAAAAAACwKhpOAAAAAAAAsCoaTgAAAAAAALAqGk4ZlJCQoDlz5qhhw4aqWrWqevTooUuXLtk7VposWrRIHTt2tHeMFN29e1fvv/++GjVqpOrVq+u1117TgQMH7B0rRbdu3dLbb7+tunXrqlq1aurZs6fOnz9v71iPFRYWpmrVqmnr1q32jpKi69evq1y5cmYfjpx527ZtatWqlSpVqqRnn31WX375pb0jmdm3b1+yr2u5cuXUtGlTe8cD4KSctTZy9LpIojbKLNRG1ucMdZFEbQRYm5u9Azi7+fPna926dZo8ebIKFiyoadOmqXv37tq5c6c8PDzsHS9Fa9eu1axZs1SzZk17R0nR4MGDdfPmTc2cOVN58uTR6tWr1a1bN3366acqVaqUveOZ6devnxISErR48WLlyJFDs2fPVufOnfXVV1/J29vb3vGSFRsbq6FDhyoqKsreUVJ16tQpeXp66uuvv5bBYDCN+/n52TFVyrZv366RI0fq3XffVcOGDfX5559r8ODBKliwoKpVq2bveCbVqlXTnj17kowdOXJEAwYMUN++fe2UCoCzc8bayBnqIonaKDNQG1mfs9RFErURYG3McMqAmJgYLVu2TG+++aaaNGmi8uXL68MPP9S1a9f01Vdf2Ttesq5fv67evXtr+vTpCgwMtHecFF28eFE///yzxowZo5o1a6pkyZIaNWqU8ufPr507d9o7npl79+6pSJEiGj9+vCpXrqygoCD17dtXN27c0NmzZ+0dL0Vz586Vr6+vvWM81pkzZxQYGKj8+fMrX758pg8vLy97RzNjNBo1e/ZsderUSR06dFDx4sXVp08fPfnkk9q/f7+94yXh4eGR5PXMkSOHJk2apNDQULVr187e8QA4IWerjZylLpKojTILtZF1OVNdJFEbAdZGwykDTp06pb///lv16tUzjeXMmVPBwcH67bff7JgsZX/88Yfc3d21Y8cOValSxd5xUhQQEKDFixerUqVKpjGDwSCDwaDIyEg7Jkterly5NGPGDJUtW1aSdPv2ba1YsUIFCxZU6dKl7Zwueb/99ps2bNigyZMn2zvKY50+fVpBQUH2jpEmYWFhunz5sp577rkk40uXLlWvXr3slCptFi5cqOjoaL3zzjv2jgLASTlbbeQsdZFEbZQZqI2sz5nrIonaCMgoLqnLgGvXrkmSChUqlGQ8f/78pm2OJiQkRCEhIfaO8Vg5c+ZU48aNk4zt2rVLFy9e1LvvvmunVGkzatQobdy4UR4eHlqwYIF8fHzsHclMZGSkhg0bpvfee8/s+9cRnTlzRgEBAerQoYPCwsJUokQJ9enTR40aNbJ3NDNhYWGSpKioKHXr1k0nTpxQ0aJF1adPH4f+2Uv8RWDIkCHy9/e3dxwATsrZaiNnqYskaiNbozayDWetiyRqI8AamOGUAdHR0ZJkth6Bp6enHj16ZI9IWdahQ4c0YsQINW/eXE2aNLF3nFS98cYb2rJli1q3bq1+/frpjz/+sHckM2PGjFG1atXM3m1yRHFxcbpw4YLu3bunAQMGaPHixapatap69uypvXv32juemQcPHkiS3nnnHbVu3VrLli1T/fr11bdvX4fMm2jdunXy8/PTK6+8Yu8oAJwYtVHmoTayLmoj23DWukiiNgKsgRlOGZB4jXRMTEyS66UfPXrksAshOqOvv/5aQ4cOVfXq1TV9+nR7x3msxGniEyZM0O+//641a9Zo0qRJdk71f7Zt26YDBw445HoPyXFzc9O+ffvk6upq+jmrWLGizp49q6VLlya5bMMRuLu7S5K6deum0NBQSVKFChV04sQJLV++3OHyJtq2bZvatGnjcGs/AHAu1EaZg9rIuqiNbMdZ6yKJ2giwBmY4ZUDidNsbN24kGb9x44YKFChgj0hZzpo1azRgwAA99dRTWrhwoTw9Pe0dKVm3b9/W559/rri4ONOYi4uLSpcubfb9YW9btmzRrVu31KRJE1WrVs10d5DRo0ere/fudk6XvBw5cpj9Z1+mTBldv37dTolSlvizn7hmRaLSpUvrr7/+skekxzp16pQuXbrkFO/qAnBs1Ea2R21kfdRGtuOMdZFEbQRYCw2nDChfvrx8fX21b98+01hkZKROnDihWrVq2TFZ1rBu3Tp98MEH6tChg2bOnOmwt1KWpIiICA0ePDjJ1ODY2FidOHHC4RZ0nD59ur744gtt27bN9CFJb775piZMmGDfcMk4e/asqlevnuTnTJKOHz/ukIuOPvHEE8qRI4d+//33JONnzpxR8eLF7ZQqdQcOHFCePHlUvnx5e0cB4OSojWyL2sg2qI1sxxnrIonaCLAWLqnLAA8PD73++uuaPn26cufOrSJFimjatGkqWLCgmjdvbu94Ti0sLEwTJ07U008/rV69eikiIsK0zcvLS35+fnZMZ65s2bJq1KiRxo8fr/HjxytXrlxatGiRIiMj1blzZ3vHSyKld5jz5MnjkO8+BwUFqVSpUho3bpzGjh2rgIAAbdy4UUeOHNGWLVvsHc+Ml5eXunfvro8++kgFChRQ5cqV9fnnn+vnn3/WihUr7B0vWSdOnFC5cuXsHQNAFkBtZDvURrZDbWQ7zlgXSdRGgLXQcMqgN998U3FxcXrvvff08OFD1apVS0uXLjVdr4z02bVrl2JjY7V7927t3r07ybbQ0FCHvF3tzJkzNWPGDA0aNEj3799XzZo1tXbtWhUuXNje0Zyai4uLFi5cqBkzZmjgwIGKjIxUcHCwli9fbjY921H07dtX3t7e+vDDD3X9+nUFBQVp7ty5qlOnjr2jJevmzZvcfQWA1VAb2Qa1ERI5W23kbHWRRG0EWIvBaDQa7R0CAAAAAAAAWQdrOAEAAAAAAMCqaDgBAAAAAADAqmg4AQAAAAAAwKpoOAEAAAAAAMCqaDgBAAAAAADAqmg4AQAAAAAAwKpoOAEAAAAAAMCqaDgBWYjRaLR3BAAAAIdBbQQA9kPDCbCyY8eO6e2331aTJk1UuXJlNWvWTKNGjdKlS5dM+5QrV05z58616nkPHjyonj17WvWYAAAAGUVtBADZEw0nwIrWrl2rV199Vbdu3dKQIUP08ccfq2fPntq/f79efPFFnTp1ymbn3rRpk86fP2+z4wMAAFiK2ggAsi83ewcAsoqDBw9qwoQJ6tChg0aOHGkar1Onjpo1a6Y2bdro3Xff1datW+2YEgAAIHNQGwFA9sYMJ8BKli5dKj8/Pw0ePNhsW+7cuTV8+HA1bdpUUVFRSbZt3bpV5cqV019//ZVkPCQkRMOHDzd9/vPPP+vll19WtWrVVKtWLfXp08f0rt3w4cP16aef6vLlyypXrpypcHv06JGmTp2qxo0bq2LFinruuef0xRdfmJ1n4sSJeuONN1S5cuUkBSEAAEB6URsBQPbGDCfACoxGo/bs2aOQkBB5e3snu0+rVq3SffxLly6pb9++ateunQYPHqzIyEjNnDlTPXv21O7du9W3b1/dvn1bJ06c0Lx581S8eHEZjUb169dPhw4d0ptvvqmgoCDt3r1bgwYNUkxMjNq0aWM6/tq1a9WlSxf16NFDOXLkSHdOAAAAidoIAEDDCbCKO3fu6NGjRypatKhNjn/06FE9fPhQvXr1UoECBSRJBQsW1DfffKOoqCgVL15cuXPnloeHh6pWrSrpn3f9fvrpJ3344Yemgq5hw4aKjo7W9OnT1bp1a7m5/fNPQOHChTV06FCbZAcAANkPtREAgIYTYAWurq6SpPj4eJscv0qVKvL09NSLL76oFi1aqFGjRqpTp44qV66c4mP27t0rg8Ggxo0bKy4uzjQeEhKiHTt26OzZs6pQoYIkmf4EAACwBmojAAANJ8AKcuXKpRw5cujKlSsp7hMVFaXY2FjlypXL4uMXLVpUa9as0eLFi7V582atWrVKOXPmVPv27TVw4EAZDAazx9y9e1dGo1HVq1dP9pg3btwwFVM+Pj4WZwIAAEgJtREAgIYTYCUNGjTQvn379OjRI3l6eppt37hxo6ZMmaLNmzcnGU8siBISEpKM//3330k+r1y5subNm6eYmBgdPHhQGzZs0MKFC1W+fHm1bNnS7Hx+fn7y8fHRqlWrks1bokQJi54fAACAJaiNACB74y51gJV07dpVd+/e1axZs8y23bx5U8uWLVPp0qX1xBNPJNnm6+srSbp27Zpp7Pz587p79/+xd99xVdb//8cfBwQBxQVuMRQnufceaGqa5mo4c+UeqWmYExP3HuXAPXLvyp2WfRxh5sqJM9MUt4GKcH5/+ON8PTLkwDkyfN5vN25wrvk8p0O+eJ339b7umx4vWrSImjVr8uzZMxwdHalYsSLffPMNgOmTQzs781/ncuXKERISgtFopGjRoqavc+fOMWvWLLOh5CIiIiLWptpIROTtphFOIlZSokQJ+vTpw9SpUwkKCqJx48ZkzJiR8+fPM3/+fJ4+fRptwVW+fHmcnJwYO3Ysffr04b///mP69OlkyJDBtE2FChWYOHEiPXr0oHXr1tjb27Ny5UocHR2pWbMmAOnSpSM4OJh9+/ZRuHBhqlevTtmyZenevTvdu3fHy8uL48ePM336dKpWrUqmTJne0CsjIiIibyPVRiIibzeNcBKxom7dujF37lwARo8eTefOnVm2bBk1atRg48aNeHl5RdknXbp0zJgxg/DwcHr06MG0adPo0aMHRYoUMW1TqFAhZs+ezePHj+nXrx89e/bk/v37LFiwgLx58wLQtGlTcubMSY8ePdi4cSN2dnbMnTuXBg0aMGfOHDp27MjKlStp3749U6ZMeTMviIiIiLzVVBuJiLy9DEaj0ZjYIUREREREREREJOXQCCcREREREREREbEqNZxERERERERERMSq1HASERERERERERGrUsNJRERERERERESsSg0nERERERERERGxKjWcRERERERERETEqtRwEhERERERERERq1LDSURERERERERErEoNJxERERERERERsSo1nERERERERERExKrUcBIREREREREREatSw0lERERERERERKxKDScREREREREREbEqNZxERERERERERMSq1HASERERERERERGrUsNJRERERERERESsSg0nERERERERERGxKjWcRERERERERETEqtRwEhERERERERERq1LDSUREREREkjSj0ZjYEURExEJqOInEoE2bNhQsWNDsq0yZMrRt25bDhw9bdKwZM2ZQsGBBGyU15+vri4+Pzxs5l4iIiAhErZsKFSpEyZIladq0KUuWLOH58+embX18fPD19Y3zsXfv3s1XX31li9giImJDqRI7gEhS5u3tzfDhwwEIDw/n3r17fP/993Ts2JH169eTP3/+RE4YVffu3Wnbtm1ixxAREZG3zKt104MHD/jll18YM2YMgYGBTJ06FTs7O2bOnEnatGnjfNxFixbZKLGIiNiSGk4isUibNi0lSpQwW1apUiUqVqzI+vXrk+Snbblz507sCCIiIvIWiq5u8vHxIW/evPj7+7N161YaNWqEt7d34gQUEZE3SpfUiVjI2dmZ1KlTYzAYTMt27dpF06ZNKVq0KJUrV2bUqFGEhIRE2Xfv3r00atSIokWLUrduXTZu3Gi2/syZM/Ts2ZMKFSrw7rvvUrVqVUaNGsWTJ08A6NChA02bNo1y3O7du9OoUSMg6iV14eHhLF++nIYNG1KsWDFq1KjBxIkTefr0qWmbNm3a0KZNG7NjHjp0iIIFC3Lo0CEA1q9fj7e3N2vWrKFy5cqUK1eOCxcuWPjqiYiIyNumdevWZM2alZUrVwJRL6mLbEQVK1aMChUq8OWXX/Lvv/8CL2qUw4cPc/jwYVNdElmjrFy5kpo1a1KqVCl+++03AAIDA2ndujXFixenXLlyfPXVV9y9e9d0roiICKZMmYKPjw9FihTBx8eHSZMmERYWFqc8IiISd2o4icTCaDTy/Plznj9/TlhYGLdv32bSpEk8e/aMZs2aAbBlyxZ69OhB3rx5mTVrFj179mTz5s107949ygSXw4YNo127dnz33Xdky5YNX19fzpw5A8CtW7do1aoVoaGhjB07lnnz5tGgQQOWLl3KkiVLAGjUqBGnTp3iypUrpmM+fPiQX375hQ8//DDa5zBs2DDGjBlD7dq1+e6772jVqhXLli2LNt/rhIeHs2DBAvz9/Rk0aBBeXl4W7S8iIiJvHzs7OypWrMjx48fN5nICOHLkCAMHDqROnTrMmzePQYMGcfDgQfr37w/A8OHD8fb2xtvbm1WrVvHuu++a9p05cyZfffUVw4YNo2TJkvz++++0a9cOJycnpk6dytdff83hw4dp27at6cO7efPm8f3339OjRw8WLFhAixYtmD9/Pt99912c8oiISNzpkjqRWPz+++9mhU2kfv364eXlhdFoZOLEiVStWpWJEyea1nt6etKuXTv27dtHjRo1TMtHjRpFtWrVgBeXvr333nscPnyYQoUKce7cOQoXLsy0adNM8xpUqlSJ3377jUOHDtG5c2fq1KmDn58fW7dupUePHgDs2LGD8PBwPvjggyg5L1y4wNq1a+nfvz+dO3cGoHLlymTJkoWBAwfyyy+/UL16dYtek65du5o9JxEREZHXcXd3JywsjPv375stP3LkCE5OTnTu3BlHR0cAMmTIwIkTJzAajeTLl89UF716uV7Lli2pV6+e6fGkSZPIkycPc+bMwd7eHoDixYvToEED1q1bR6tWrTh8+DBFihQxfXBYrlw5nJ2dcXV1jVOel0e4i4hI7DTCSSQW7777LmvXrmXt2rWsWbOG+fPn89lnnzFlyhSmTJnCxYsXuXnzJj4+PqaRUM+fP6ds2bKkTZvWNLw7UpkyZUw/58qVC3gxQgmgSpUqLFu2jNSpU3PhwgV2797Nd999x927d3n27BkALi4u1K5dmx9//NF0nB9++IGKFSuSNWvWKPkj76bXoEEDs+UNGjTA3t7edLmcJQoXLmzxPiIiIvJ2ixxV/WrDpmzZsoSGhvLBBx8wadIkAgMDqVKlCj179nxtc+flmiQ0NJRjx45RvXp1sxHqHh4eeHl5mWqy8uXL89tvv9GyZUsCAgK4cOECrVu3No0UT0geERExpxFOIrFIkyYNRYsWNVtWpUoVQkJCCAgIoHLlygD4+fnh5+cXZf9bt26ZPXZxcTH9bGf3ot8bWYBFREQwefJkli9fTkhICNmzZ6dYsWKkTp3a7Bgffvghmzdv5syZM7i7u3Po0CFGjx4dbf4HDx4AkDlzZrPlqVKlImPGjDx69Oi1r8GrXn4OIiIiInHx77//4uTkRIYMGcyWlyxZkrlz57Jo0SIWLlzI3LlzcXd3p2vXrlHml3zVyzXJw4cPiYiIYN68ecybNy/KtpH1VKdOnUiTJg3r1q1j4sSJTJgwgfz58zNkyBAqVKiQoDwiImJODSeReChSpAhr1qwxPR44cCDlypWLsl369OnjfMzI4sbPz486deqYhnY3b97cbLuKFSuSOXNmfvrpJzJnzkzq1KmpU6dOtMeMPP/t27fJmTOnaXlYWBj37t0jY8aMpmXh4eFm+0Y36bmIiIiIpZ4/f86hQ4coVaqU6VK3l1WtWpWqVasSGhrKwYMHWbJkCaNGjaJ48eIUK1YsTudIkyYNBoOBdu3aRRnZDS9u+gIvPvBr1aoVrVq14s6dO+zbt4/Zs2fTq1cvfvvtNxwdHa2SR0REdEmdSLwcP34ce3t78ubNi5ubG3///TdFixY1fWXNmpVJkybx119/xfmYR44cIV++fDRr1szUbPr33385d+4cERERpu3s7e1p2LAhP//8M9u2baN27doxjjqKbIL98MMPZst/+OEHwsPDKV26NPDiNsY3b96MkkdEREQkoVatWsXt27dp0aJFlHXjxo2jWbNmGI1GnJ2dqVmzJl999RUA//zzD/B/o8JjkzZtWry9vbl48aJZTZY/f35mzJhhmkbg008/ZdSoUQC4ubnRtGlTWrVqxcOHD3n8+HGc8oiISNxohJNILB4/fsyff/5pevzs2TP27NnDunXr+OSTT3B3d6dv374MGzYMe3t7atasycOHD/n222/5999/o51wPCbFihXj22+/Ze7cuZQoUYIrV64wZ84cnj17RmhoqNm2H374IQsWLMDOzi7aYeOR8uXLR5MmTZg+fTqhoaGULVuW06dPM3PmTMqXL0/VqlUBqFmzJnv27GHMmDH4+PgQGBjIxo0bLXqtRERE5O32ct0UERHBvXv32L9/P6tWraJRo0bRjsiuUKECCxcuxNfXl0aNGhEWFkZAQAAZMmSgQoUKAKRLl46jR49y4MABvL29Yzx/v3796Ny5M/3796dRo0amu+seO3aM7t27Ay/maFqwYAHu7u6ULFmSf//9l4ULF1KuXDkyZcoUpzwiIhI3ajiJxOKvv/7ik08+MT1OnTo1uXPnpm/fvnTs2BGAjz76iDRp0hAQEMCqVatwcXGhVKlSTJw4EQ8Pjzifq0uXLty7d48lS5Ywa9YssmfPzocffojBYGDOnDk8fPiQdOnSAVCoUCEKFCjAvXv3qFixYqzH9ff355133mHdunXMmzePLFmy0LZtW7p37276xLBZs2ZcvXqVDRs2sHLlSsqWLcv06dOj/SRSREREJDov100Gg4E0adJQoEABRowYwUcffRTtPtWrV2fixIksWLDANDF36dKlWbJkiWm+p1atWnHy5Ek+//xzxowZQ5YsWaI9VpUqVZg/fz4zZ86kd+/eODg48O6777Jw4ULTHe769OmDo6Mj69atY9asWbi6uuLj40P//v3jnEdEROLGYIycsVhERERERERERMQKNIeTiIiIiIiIiIhYlRpOIiIiIiIiIiJiVWo4iYiIiIiIiIiIVanhJCIiIiIiIiIiVqWGk4iIiIiIiIiIWJUaTiIiIiIiIiIiYlWpEjvAm3b06FGMRiMODg6JHUVEJMULCwvDYDBQsmTJxI4iIjFQbSQi8uaoNpK3yVs3wsloNGI0Gm1y3GfPntnk2NaWnLKC8tpScsoKyStvcsoKtstrq//nioj12OL3VP8PtK3klDc5ZQXltaXklBVUG4lYw1s3winy07uiRYta9bghISGcPn2afPny4eLiYtVjW1tyygrKa0vJKSskr7zJKSvYLu+JEyesdiwRsQ1b1Eb6f6BtJae8ySkrKK8tJaesoNpIxBreuhFOIiIiIiIiIiJiW2o4iYiIiIiIiIiIVanhJCIiIiIiIiIiVqWGk4iIiIiIiIiIWJUaTiIiIiIiIiIiYlVv3V3qREQk5QoPDycsLCyxY0gS4eDggL29fWLHEBERSTSqjcSaLK2t1HASEZFkz2g0cvPmTe7fv5/YUSSJyZAhA9myZcNgMCR2FBERkTdGtZHYiiW1lRpOIiKS7EUWVFmyZMHFxUXNBcFoNBISEsKtW7cAyJ49eyInEhEReXNUG4m1xae2UsNJRESStfDwcFNB5ebmlthxJAlxdnYG4NatW2TJkkWX14mIyFtBtZHYiqW1lSYNFxGRZC1yXgIXF5dETiJJUeT7QvNXiIjI20K1kdiSJbWVGk4iIpIiaKi4REfvCxEReVvp30CxBUveV2o4iYiIJANGozGxI4iIiIgkG6qdEp8aTiIiIjZy4sQJBgwYQI0aNShWrBi1a9dm6NChXLt2zaLjnD9/nhYtWtgopYiIiEjSoNopZVHDSUQwGAxEREREWR7dMhGJm+XLl/Ppp59y584d+vfvz7x58+jcuTOHDx+mefPmnDlzJs7H2rZtG0ePHrVhWhFJCozh4fFaJyKSEqh2SnksvkvdhQsX2LJlCwcPHuTvv//m0aNHZMyYkRw5clCtWjXq1KmDl5eXLbKKiA3Z2dmxe/J+7l97CEAGj3TU6lclkVOJJE9HjhzB39+fVq1aMXjwYNPy8uXLU7t2bRo3bszXX3/N+vXrEzGlWItqI7EWg709d3v24vn5C2bLU+XPR6aZMxIplYiI7al2SpniPMLp4sWLdOvWjYYNG7Jp0yYyZ87MBx98QIcOHahVqxYZM2ZkyZIlfPDBB/Ts2ZMLFy68/qAikqTcv/aQ4It3Cb5419R4EhHLzZ8/H1dXV/r16xdlXaZMmfD19aVWrVqEhITw5MkTJk2aRJ06dShSpAilSpWiffv2nD59GoAZM2Ywc+ZMAAoWLMiMGS/+6IyIiGDu3Lm89957FClShLp167J06dJos9SqVYtixYrx6aefsmfPHgoWLMihQ4dM25w4cYKOHTtSvnx5SpUqRdeuXTl//rxp/aFDhyhYsCArV66kZs2alCpVil27dlGwYEH2799vdr7AwEAKFizIkSNHEv5CJnGqjcQWnp+/QNjJk2ZfrzagRERSGtVOKbN2itMIp4CAAObNm8cHH3zA999/T4kSJWLc9vjx46xcuZKWLVvy+eef8/nnn1srq4iISJJnNBrZv38/Pj4+ODs7R7tN/fr1TT/37t2bwMBA+vXrR+7cubly5QrTpk2jf//+/PDDD3z00UfcvHmTtWvXsmrVKrJlywbAiBEjWL9+PV26dKFkyZL8/vvvjB49mocPH9KjRw8AZs6cyaxZs+jYsSMVKlTg119/5YsvvjDLcvDgQTp16kT58uUZPXo0T58+Zc6cOXz66aesXr3abGTOzJkzGTJkCE+ePKFSpUpkyZKFTZs2UaXK/42G3LhxI56enpQuXdpaL2mSpNpIRETEOlQ7pdzaKU4Np/Pnz7NlyxayZMny2m2LFStGsWLF6NWrF5MnT05wQBERkeTk3r17PH36lFy5cr1222fPnvHff/8xZMgQUyFVrlw5Hj9+zNixYwkODiZbtmymQimyqXHp0iVWr15Nv3796Ny5MwBVqlTBYDAwZ84cWrZsSerUqZk3bx6tWrXiyy+/NG0TGhrKqlWrTBkmTZrEO++8w9y5c7G3tzdt99577zF9+nSmTZtm2rZly5bUq1fP9LhJkyYsXbqU//77jzRp0vDkyRN++uknU6aUTLWRiIiIdah2Srm1U5wuqRs3blycCqqXZc+enQkTJsQrlIiISHIVWXiEx2GCX0dHR+bPn0/9+vX5999/OXjwICtXruTnn38GXhRV0Tl48CBGoxEfHx+eP39u+vLx8eHp06ccOXKEP//8kydPnpgVOQAffPCB6eeQkBBOnDjB+++/b8oNkC5dOmrWrMnhw4fN9i1cuLDZ42bNmhESEsLOnTsB2LlzJyEhITRu3Pi1zz25U20kIiJiHaqdUm7tZPGk4a8yGo3cu3ePDBkyYGdn+U3v/v33X6pVqxZl+ZgxY2jatCmnT5/G39+fkydPkilTJtq1a0fbtm0TGltERMQm0qdPT5o0afjnn39i3CYkJISwsDDSp0/Pr7/+yujRo7l48SJp0qShUKFCuLi4AC/+jY3O/fv3AWjQoEG06//991/Sp08PvJj34GVubm6mnx89eoTRaMTd3T3KMdzd3Xn06JHZsshckd555x3KlSvHxo0bady4MRs3bqRSpUpkzZo1xuf+NkhobSQiIvI2Ue2UcmuneDec7t+/z7Bhw9izZw/h4eHY29tTvXp1hg8fbtEnfmfOnCF16tTs2rULg8FgWu7q6sq9e/do3749Pj4++Pn58eeff+Ln50eaNGlo1qxZfKOLiIjYVJUqVTh06BBPnz4lderUUdavXr2acePGsWbNGnr06EHt2rWZM2cOHh4eGAwGli9fzq+//hrj8dOlSwfA4sWLSZMmTZT1OXLk4NKlSwDcuXOHvHnzmtbdvXvX9LOrqysGg4Hg4OAox7h9+zYZMmR47XNt1qwZX3/9NUFBQRw4cICJEye+dp+Uylq1kYiIyNtGtVPKFO+P3b755hscHR1ZsWIF27ZtY+HChTx58oSvvvrKouOcO3cOT09PsmTJQubMmU1fTk5OrF69GgcHB0aOHImXlxfNmjWjXbt2zJ07N76xRUREbK5Dhw7cv3+fqVOnRll3+/ZtFixYQL58+bh69SpPnz6lc+fO5M6d2/TBS2TBFPkp3aujZMqUKQO8mPOgaNGipq+7d+8ybdo07t+/T6FChXB1dTUN2Y60Y8cO088uLi4UKVKEn376yWwY+6NHj9i7d2+cJq+sW7cuzs7OjBgxgjRp0lC7du04vEIpk7VqIxERkbeNaqeUKU4jnLZu3Wp23SLAX3/9xcyZM00zsL/zzjuEhISYJteKq7Nnz5rN4v6ywMBAypUrR6pU/xezQoUKzJkzh+Dg4GiHsYmIiCS2EiVK0KdPH6ZOnUpQUBCNGzcmY8aMnD9/nvnz5/P06VOmTp1KqlSpSJUqFRMmTKBDhw48e/aM9evXs3fvXuDF8HH4v0/ltm7dSvHixSlYsCCNGjVi6NChXL9+nSJFinDp0iWmTJlCrly58PT0xN7enk6dOjF9+nScnZ0pV64chw8f5vvvvwf+rxDr378/HTt2pHPnzrRs2ZKwsDDmzp3Ls2fPTHdsiY2zszMNGjRg1apVtGjRAkdHRxu8okmPLWsjERGRt41qp5QpTg2nhQsXsmDBAvr370/lypUBqFatGn379uXDDz8kffr0BAcHs3r1amrVqmVRgHPnzpExY0ZatWrFpUuXeOedd+jWrRvVqlXj5s2bFChQwGz7yCHpN27ciHfDyWg0mt6I1hIaGmr2PSlLTllBeW0pMuPTp09jvAVpaGhojNdCv2nJ8bVNDlnBdnmNRqPZ5dJvk27duuHt7c3y5csZPXo0Dx48IHv27NSoUYOuXbuSPXt24MWdTmbOnEm3bt1Inz49JUqUYOnSpbRp04bAwEAKFixInTp12LRpE76+vjRv3pwRI0YwZswY5syZw8qVK7l58yZubm7Ur1+fL774wjSJZZcuXTAajaxatYr58+dTvHhxvvzyS8aMGWOaU6BixYosXLiQ6dOn069fPxwdHSlTpgzjxo0jf/78cXquNWrUYNWqVTRt2tQ2L2YSZMvaSERE5G2k2inliVPDad26dWzdupXhw4eTM2dOBgwYwFdffcWyZcvYsWMHd+7cwc3NjZYtW1o0offz58+5ePEi+fLlw9fXl7Rp0/LDDz/QuXNn0zD0V7t9kddzPn361IKnaS4sLIzTp0/He//YXL582SbHtYXklBWU15b++eefGK83vnTpUpJrmiSn1zY5ZQXb5E3Jn9q8TvXq1alevXqs29SrVy/K3VDgxRyHkbJmzcratWvN1qdKlYoePXrE+Ena8+fP2bp1K40bN6Zbt26m5cuXL8fOzg4PDw/TsvLly7N8+fIYM5YvX56zZ8/GuH7fvn0ULlyYYsWKxbhNSmOr2khERORtptopZYnzpOEffPABdevWZfny5Xz++eeUK1eOvn37JqiISpUqFYcOHcLe3h4nJycAihQpYho25+TkFOW2hpGNpldne7eEg4MD+fLli/f+0QkNDeXy5ct4enrGOFIkqUhOWUF5bSkya44cOWLcJk+ePElqhFNye22TQ1awXd4LFy5Y7VhimVSpUjFv3jwWL15Mt27dyJgxI+fOnWPq1Kk0btzYNNQ8IZYsWcLFixdZvXo1EyZMsELq5MUWtZGIiIgkDtVO1mfRXeocHBxo164dzZs3Z+7cuTRr1oyGDRvSs2fPeF/eFt0M8fnz52f//v1ky5aNW7duma2LfJyQ2wYaDIYENaxi4+zsbLNjW1tyygrKa0vR3QkiUlJsliSn1zY5ZQXr531bL6dLKmbPns3kyZMZMWIEDx8+JEeOHHz22Wd06dLFKscPDAzk119/5bPPPosyn9Hbwha10f3795k8eTJ79+7l8ePHFCxYkP79+5smPG3fvj3/+9//zPYpV64cS5cuTfDzEREReZupdrKuODecgoKCOHToEBEREZQuXZp+/frRqlUrpk2bRt26dWndujWff/45adOmjfPJz58/zyeffMJ3331H+fLlTctPnjxJvnz5KFy4MCtXrjTdWhjg4MGD5MmTBzc3NwuepoiIyNvHw8ODKVOm2Oz406dPt9mxkwNb1EYA/fr14/bt20yePBk3NzeWLl1Kx44d2bBhA3nz5uXs2bOMGDHC7K42Dg4O1n56IiIibx3VTtZl9/pNYMOGDXz44YesWrWKjRs38umnn7Jw4UKyZs3K6NGjWblyJWfPnqV27dosWrQozif38vIib968jBw5ksDAQIKCghgzZgx//vkn3bp1o1mzZjx+/JjBgwdz4cIF1q9fz6JFi6zWXRQRERGJD1vVRleuXOG3335jxIgRlClThjx58jB06FCyZMnCli1buHPnDnfu3KF48eJkzpzZ9BXTPHwiIiIiiSVODacZM2YwaNAgNm3axNq1a1myZAnTp083zeuSP39+Zs+ezfTp0/nxxx/jfnI7O2bPnk2xYsX44osvaNKkCceOHWPhwoUUKFAANzc3AgICuHTpEk2aNGHmzJkMHDiQJk2axO/ZioiIiFiBrWqjjBkzMnfuXIoWLWpaZjAYMBgMPHz4kLNnz2IwGMiTJ4/Vn5OIiIiINcXpkronT56YfXKWPn16wsLCeP78udkQ7nLlyrF69WqLAri7uzNmzJgY1xcrVoxVq1ZZdEwRERERW7JVbZQuXbood+fZvn07V65c4euvv+bcuXO4uroycuRIfvvtN1xcXKhXrx7du3dP0B0hjUYjISEh8d7/VZF3N01qdzmNSVLIazAYXjtvYmhoKEajMUnkjavklBWU15aSU1awXV6j0ag5LuWtEaeG08cff8zXX3/N1q1bcXJy4vDhwzRv3lzzBYiIiMhb6U3VRn/88QeDBg2iTp061KhRg6+//pqnT59SrFgx2rdvz+nTpxk/fjz//PMP48ePj/d5wsLCOH36tBWTv3D58mWrH9OWEjOvs7Mz3t7esW5z6dIlsz9+k9Prm5yygvLaUnLKCrbJm5APCESSkzg1nL744gtKlCjBgQMHMBgMDB06lHr16tk6m4iIiEiS9CZqo127dvHll19SqlQpJk6cCMDIkSP56quvSJ8+PQAFChTAwcGBvn37MnDgwHjfGc/BwYF8+fJZLXtoaCiXL1/G09MzSd7t9FVJIW9cRjzkyZPHNMIpsfPGVXLKCsprS8kpK9gu74ULF6x2LJGkLs53qatRowY1atSwYRQRERGR5MOWtdGyZcvw9/enXr16jBs3zvRpeKpUqUzNpkj58+cH4ObNm/FuOBkMBlxcXBIWOhrOzs42Oa6tJPW8r/7Rm9Tzviw5ZQXltaXklBWsn1eX08nbJE6Thk+bNo1nz55ZdOCQkBCb3k5QREREJLHYsjZasWIF33zzDa1atWLy5Mlml160adOGQYMGmW1/4sQJHBwc8PT0tCiPiIiIiC3FqeH08OFD6tSpw6JFi7h161as296+fZtvv/2WunXr8vDhQ6uEFBERiQ9jeHiyPffjx48pXrw4lSpVIiwszGzdtWvXaNasGUWKFKFPnz5RHvv6+tKmTRsADh06RMGCBfn777+BFw0LX1/fOOe4d+8ea9asMT22dP+Uyla10aVLlxg9ejTvvfceXbp0ITg4mNu3b3P79m0ePXpE3bp12bRpE99//z3Xrl3jxx9/ZPz48XTs2JG0adNa8ymKiEgKlFi1UXzP6+PjQ8GCBU1fRYoUoUaNGgwfPpy7d++atitYsCDr16+P0zFfrW3EduJ0Sd3QoUOpXbs2Y8eOZfz48RQvXpxixYqRK1cunJ2defToETdu3ODIkSOcPXsWLy8vRo8eTdWqVW2dX0REJEYGe3vu9uzF8/Nvdr6EVPnzkWnmjAQd44cffsDNzY3bt2+zc+dO6tevb1q3bNkybt68yaZNm8iQIQNz5841e+zo6Eh4DIXdjBkzsLe3j3OO8ePH8/fff/PRRx/Fa/+Uyla10fbt2wkLC2Pnzp3s3LnTbF2TJk0YO3YsBoOBpUuXMnr0aDJnzky7du3o3LmzLZ+uiIikEIlRGyW0LurQoQMdOnQAXtwl9ty5c0yYMIHWrVuzatUqXF1d2b9/P66urnE63qu1jdhOnOdwqlixIps2bWLv3r1s2bKFrVu3cufOHdN6d3d3qlSpQs+ePalZs6ZNwoqIiFjq+fkLhJ08mdgxLLZu3TqqVq3KP//8w8qVK80aTg8fPiRPnjx4eXlF+zg2GTJksCiH0WhM0P4pmS1qo65du9K1a9dYt2nVqhWtWrVKUHaRhDKGh2OIpfn8uvUikniSW23k4uJC5syZTY89PDwoXLgwDRo0ICAggL59+5qtf51XaxuxnTg3nCK9PEFmaGgojx49Mn2aKiIiIgkXFBTEsWPH6NSpEw8ePGDIkCFcunSJPHny0KZNGw4fPgy8GD5erlw5s8dLlixhw4YNXL9+naVLl0Y5dps2bciZMydjx44FYM2aNSxZsoQrV65gZ2eHt7c3gwYNomjRovj6+rJhwwbTsc+ePRtl/6NHjzJlyhROnTpFqlSp8PHxYeDAgWTMmBF4MRS+VatW/Pnnn+zfvx9HR0caNmyIr68vqVJZXIYkSaqN5G0U2ygJa4zyFBGJTY4cOXjvvff44Ycf6Nu3LwULFmTMmDE0bdqUO3fu4Ofnx6FDhwgNDcXb25t+/fpRrly5GGsbT09Pzpw5w6VLlxg2bBiNGjVi3bp1BAQEcP36dXLmzMmnn35KmzZtsLN7MTPRxo0bmTdvHlevXiVDhgzUq1ePAQMGmEaaT5482fRhVK5cufjss89o0aJFYr5sb1yc5nCKibOzM1myZFFBJSIiYkVr167FxcWFatWq8d577+Hg4MDKlSuBF5e0vf/++5QsWZL9+/czdepUs8clS5aM83l27tzJyJEj6dSpEz/99BOLFi3i6dOnDBkyBIDBgwebHftVx48fp02bNuTPn5/Vq1czbdo0jh07RseOHc0u6Zs2bRply5Zl8+bNDBw4kGXLlrF169YEvkpJk2ojeZtEjpJ49etNX8YsIm+nAgUKcO3aNf777z+z5SNGjODp06csW7aMLVu2kCdPHrp3705ISEiMtc2aNWto27YtK1asoGrVqqxatYrx48fTs2dPfvjhB7744gvmzZvHxIkTAThz5gxDhgyhV69ebN++ndGjR7Np0yYCAgKAFzcA2bZtG1OmTGH79u20bt2aESNGEBgY+OZeoCQgZXy0KCIikkI8f/6czZs34+Pjg5OTE05OTlSpUoWNGzfSr18/MmTIgJOTEw4ODqbh468+jqsMGTLg7+9Po0aNAMiZMyfNmzdn5MiRALi6usZ67AULFlCwYEGGDh0KgJeXF5MnT+bDDz9k//79VK9eHYAqVarQtm1b4MUw+KVLl/LHH3/QuHHjeL1GIiIiIunSpQNe3GjlZVevXqVAgQJ4eHjg5OTE4MGDadiwIfb29ri4uERb2xQuXJiGDRuaHn/77bd069aNBg0aAC/ql8ePH+Pn50efPn34+++/MRgM5MyZkxw5cpAjRw7mz59vuoHH1atXcXFxIVeuXGTJkoXWrVuTN29e8uTJY+uXJUlRw0lERCQJ2bdvH8HBwaYCB6BBgwb8/PPP/PTTT1Zt0pQtW5agoCBmzZrFxYsXuXLlCmfPniUiIiJO+587d47KlSubLStUqBCurq6cPXvW1HB6dW4pV1fXKHfeExEREbHEo0ePAKLcpbVnz54MGDCA7du3U7p0aapUqcIHH3xA6tSpYzzWO++8Y/r57t273Lx5k8mTJzNt2jTT8oiICJ4+fcrff/9N1apVKVmyJM2bNydXrlxUrlyZWrVqUaRIEeDFfIu7du2ievXqFC5cmMqVK9OgQQPc3Nys+RIkeWo4iYiIJCGRt/Tt2bNnlHUrV660asNpy5Yt+Pr60rBhQ0qVKsWnn37KuXPnTCOcXiemSTeNRiMODg6mx9FdXqYJO0VERCQhTp06haenJ2nSpDFb/t577/Hrr7/y66+/8r///Y+FCxcyc+ZMVq9eTf78+aM9lpOTk+nnyA/eBg0aRKVKlaJsmz17dhwdHVmyZAl//fUX+/fvZ//+/XTt2pXGjRszZswYPD092bFjB4cPH+a3335j7969zJs3jzFjxtCkSRMrvgpJm8VzOB04cIAnT57YIouIiMhb7c6dO+zbt4+mTZuyceNGs69mzZpx9OhRzp07Z7XzzZ07l+bNmzN27FhatWpF2bJluXbtGvB/DSGDwRDj/gULFuTIkSNmy86cOcPjx4/jdMe8lEK1kYiIyJt18+ZNdu/ebXYZHMCzZ88YM2YM165do379+owaNYpdu3ZhZ2fH3r17gdhrGwA3NzcyZcrEtWvXeOedd0xfp06dYurUqcCLEekzZ87E29ubzp07s2TJEnr37s2PP/4IwJIlS9ixYweVK1dm4MCBbNmyhYoVK5rWvy0sbjj16tWLHTt22CKLiIjIW23z5s08f/6czz//nAIFCph9de3aFTs7O9Pk4daQPXt2/vjjD06dOsXVq1dZtGgRy5YtA14UbPDiVsS3bt0yNaJe1r59e86ePcs333xDUFAQhw4d4ssvv8Tb25uKFStaLWdSp9pIRETEdkJCQrh9+za3b9/m2rVr7Nq1i06dOpErVy7at29vtq2joyMnTpxg6NCh/Pnnn/z999+sX7+ekJAQ041VYqtt4EVD6vPPP2fp0qUsW7aMq1evsnPnTkaMGIGTkxOOjo44ODgwa9YsFi1axLVr1zh58iR79+41nePu3buMHDmS3bt3c/36dX799VdOnz5t0c1dUgKLL6lLly6d2XAzERGRpCxV/nzJ5pzr16+nUqVK5M2bN8q63LlzU7t2bTZv3hzt8O74GDp0KMOGDaN169Y4OjpSqFAhxo8fT9++fTlx4gRlypShcePG7Ny5kw8++CBKU6V48eIEBAQwdepUGjduTNq0aalduzb9+/c3u6QupVNtJCIiycmbro0Ser4FCxawYMECABwcHMiePTv169enQ4cOUS6nA5gyZQpjxoyhW7duPHr0iLx58zJx4kTKlCkDEGttE6lDhw6kTp2apUuXMnbsWNzd3fn444/p3bs3AJUqVcLf358FCxYwZcoUnJycqF69Or6+vsCLqRHCwsIYNWoUt2/fJnPmzLRo0YIuXbok6LVIbixuOHXp0oVRo0Zx6dIlChUqhIuLS5RtypYta5VwIiIiCWEMDyfTzBmJdm6Dvb1F+2zZsiXW9TNmRP9cxo4dG+Pj8uXLc/bsWdPjpUuXmn728PBg4cKFUY5Xv359089Fixbll19+iXZ/gIoVK8Y6mmnPnj1Rlr16jOROtZGIiCQXiVUbxacugujriOi8XOtkzZrVdOlbdF5X20Rq1aoVrVq1ivE4TZs2pWnTptGuS5UqFQMGDGDAgAGvSZ6yWdxwGj58OPCiawjm1z8ajUYMBgOnT5+2UjwREZH4i09hkxLOLW+WaiMREUkuEqs+UV30drK44bRkyRJb5BARERFJllQbiYiIiERlccOpXLlytsghIiIikiypNhIRERGJyuKGE7yYcX3+/Pn873//4/bt2wQEBLBr1y4KFSpE7dq1rZ1RREREJElTbSQiIiJizs7SHa5du0ajRo1YvXo1WbNm5c6dO4SHh3Pp0iV69+7N3r17bRBTREREJGlSbSQiIiISlcUjnMaNG4ebmxtLly7FxcWFIkWKADBp0iSePn3K7NmzqVGjhrVzioiIiCRJqo1EREREorJ4hNOBAwfo3r076dKlM7sLC8Ann3zC+fPnrRZOREREJKlTbSQiIiISlcUNJ4BUqaIfGPXs2bMohZaIiIhISqfaSERERMScxQ2nMmXKMGfOHEJCQkzLDAYDERERfP/995QqVcqqAUVERESSMtVGIiIiIlFZ3HDq378/QUFB1KlTh4EDB2IwGJg/fz5NmzblyJEj9O3b1xY5RURE3hrPnz9n8eLFNG3alJIlS1KhQgU6dOjAwYMHEzvaa927d481a9aYHrdp0wZfX1/T459//pkLFy4AcOjQIQoWLMjff//9xnNak2ojERER2/Dx8cHHx4fHjx9HWefr60ubNm2ser5Xa5NX65jYhIWFsWjRIqvmSe4sbjgVKFCAtWvXUr58eQ4dOoS9vT3/+9//yJ07NytXrqRw4cK2yCkiImKx8Ahjsjv306dPadu2LYsWLaJNmzZs2LCBRYsW4eXlRfv27dmyZYuVk1rX+PHj2bx5s+nxjBkzGDx4MADXr1+na9eu3LlzB4CSJUuyf/9+smfPnihZrUW1kYiIJBeJVRsl5LzXr19n/PjxVkwTdy/XMa+zdetWxowZY+NEyYvFd6kDyJMnD5MmTYp23c2bN8mWLVuCQomIiFiDvZ2B4WuPczk46qdituTpnha/5sXite+0adM4e/YsW7duNWvEDB48mMePHzNq1Ch8fHxIkyaNteJaldFoXlBmyJAhxnWOjo5kzpz5TcSyOdVGIiKSHCRGbZSQugjAw8ODVatWUa9ePSpVqmTFZK/3ch3zOq/WORKPhlPhwoVZtWoVxYpFfcMEBgby+eefc/ToUauEExERSajLwY85e+NRYseIk7CwMNatW0fTpk2jHfXzxRdf0KJFC5ycnLh//z7Tpk1jz5493Lt3D29vb/r27Uv58uWBF5/IHTlyhDJlyrBixQpCQ0Np2LAh3bp1Y8SIERw8eJAsWbIwePBgatSoAbwYtt68eXOOHDnC77//TtasWencuTMfffSRKcMff/zBpEmTOHHiBJkyZaJmzZr079+ftGnT4uvry4YNGwAoWLAgZ8+epU2bNuTMmZOePXtSq1YtANq2bUvPnj0pV64cbdu2Zffu3eTKlYsnT54we/ZstmzZwq1bt8ibNy/du3enbt26AKxfv57vvvuObt268d1333Hjxg0KFCjA4MGDKV26tC3/08RKtZGIiCQnyak2AmjUqBF//PEHgwcPZsuWLaRNmzba7V5XG0UnMDCQcePGcfbsWTw9PWnWrJnZ+sg6ZuzYsYSHhzN58mS2bt3KnTt3yJUrF5999hktWrRg/fr1DBo0CHhRAy1ZsoTDhw9z6NAhMmfOzL59+2jSpAlDhw6NtZYCOH78OGPHjuX06dOkSpWKChUqMGjQIHLkyAHAxo0bmTdvHlevXiVDhgzUq1ePAQMG4OjoaI2X26ri1HBasGCBaSJMo9HImjVr+OWXX6Jsd/To0ST5JEVERJKDa9eucf/+/Rgnmc6aNStZs2YlPDycDh06EBYWxoQJE8iUKRNLliyhY8eOrFixwtT4CAwMxM3NjeXLl/PHH3/w9ddfs3v3bgYMGMDAgQOZMGECvr6+HDhwwHQntW+//ZauXbsyePBgfvnlF4YNG0aaNGmoX78+Z86coX379nTr1g1/f3+Cg4MZP348HTp0YNWqVQwePJgnT55w8+ZNZsyYYZY9e/bsrFmzho8++ogZM2ZQuXJlTp48abZNv379+OuvvxgxYgTvvPMOW7dupU+fPsycOZPatWsDcOPGDVauXMmECRNIkyYNI0aMwNfXlx07drzRu8GpNhIREXkzDAYD/v7+NGzYkHHjxvHNN99E2SautdHLrl27RocOHWjcuDFjx47lwoULDBs2LMYcK1asYNu2bUyZMoWsWbPy888/M2LECPLnz0/9+vV59OgRo0ePZv/+/aRPn57Dhw/z+++/07ZtWzZt2kR4ePhra6mIiAi6dOnCxx9/zLhx43j48CHDhg3j66+/ZtGiRZw5c4YhQ4YwceJEihUrRlBQEP379ydjxox0797dqq+7NcSp4fT06VNmzpwJvPiP/fJkoJHs7OxwdXWlW7du1k0oIiLylnjw4AEA6dOnj3W7/fv3c+rUKbZs2UKBAgUA8PPz48SJE8yfP59p06YBEBERgZ+fH2nTpiVPnjxMmDCBChUq0LhxYwBatGjBzz//zO3bt8mSJQsAVapUoWfPngDkzZuXY8eOsXjxYurXr8/8+fOpXLkyXbt2BcDT05NJkyZRu3ZtDh8+TPny5XFycsLBwSHKpXL29vZkypTJ9PxevSQwKCiI3bt3M3v2bNOIq169enHmzBlmz55tajiFhYXh5+dnmhepffv29OjRw+w5vAmqjURERN6cnDlz8tVXXzFs2DDq1q1LlSpVzNbHtTZ62erVq3F3d2f48OHY29vj5eXFjRs3YpyH6erVq7i4uJArVy6yZMlC69atyZs3L3ny5MHJyQlXV1eAKDVQ7969TesGDBgQay1VqFAh7t27R5YsWciZMyceHh5MnTrVNP/l33//jcFgIGfOnOTIkYMcOXIwf/78GEd9JbY4NZy6detmKpYKFSrEqlWrKF68uE2DiYiIvG0iGzL379+Pdbtz587h6upqKqjgRdOjTJky7N+/37TMzc3NrABxcXEhd+7cpsdOTk4APHv2zLTs1WHnJUuWZO/evQD89ddfXLlyhZIlS0bJFBQUFOuQ9dc5e/YsQJRL48qWLcvkyZPNlnl5eZl+jizgwsLC4n3u+FBtJCIi8mZ98sknbN++nSFDhrB161azdXGtjV7dx9vbG3t7e9OymEaZA7Rq1Ypdu3ZRvXp1ChcuTOXKlWnQoAFubm4x7uPm5maqVSButVSnTp345ptvmD59OhUqVKB69eq8//77AFStWpWSJUvSvHlzcuXKReXKlalVqxZFihSJMUNisngOpzNnzkRZ9vz5cx4/fmzRhFoiIiJizsPDA3d3d/744w/q168fZX1QUBD+/v4xFkNGo5FUqf7vn3YHB4co29jZxX6D2pf3hxejpCL3iYiIoGHDhqZP5V4W2SyztlefExDtJWqJOVGnaiMREZE3Y9SoUTRs2DDKKKSY6oDo6ohIBoOBiIgIs2UxbQsvRiPt2LGDw4cP89tvv7F3717mzZvHmDFjaNKkSbT7RH64FykutdSXX35Jy5Yt2bdvHwcOHOCbb74hICCAjRs3kjp1apYsWcJff/3F/v372b9/P127dqVx48ZJ8g55sVed0QgPD2fmzJmm2zIfOnSIypUrU7FiRT777DPT5QAiIiJiGTs7O5o3b8769eu5ceNGlPUBAQGcOHGCIkWK8OjRI86dO2daZzQaOXLkCPny5UtQhhMnTpg9/uOPP/D29gYgf/78XLhwgXfeecf09fz5c8aMGWPKG9s8SrGtK1iwIABHjhwxWx4YGJjg52Rrqo1ERETejBw5cuDr68vatWsJDAw0LS9YsKDFtVGhQoU4efKk2UjvV+eXfNmSJUvYsWMHlStXZuDAgWzZsoWKFSvy448/ArHXOZFeV0tdvHiR4cOH4+bmRosWLZg+fToBAQEEBQVx5swZ9u3bx8yZM/H29qZz584sWbKE3r17mzIkNRY3nKZNm8Z3333Hw4cPgRcdxgwZMjBo0CCuXr0a4y2BRURE5PW6du2Kp6cnLVu2ZOPGjVy9epXjx48zaNAgNm7cyDfffEPVqlUpXLgw/fv35/DhwwQFBTFy5EjOnTvHZ599lqDz//DDDyxfvpzLly8TEBDAzp076dSpEwAdOnTgr7/+ws/Pj6CgII4ePUr//v25fPkynp6ewIvL9m7dusW1a9eiHNvFxQV4MYT90SPzu+N4eXlRs2ZN/Pz82Lt3L5cuXWLmzJns3r2bDh06JOg52ZpqIxERkTfno48+okqVKma1RpUqVSyujVq0aEFoaChff/01QUFB/Pzzz1FuevKyu3fvMnLkSHbv3s3169f59ddfOX36tOnyuMg65+TJkzx58iTaY7yulsqYMSM//PADw4YNIygoiEuXLrFhwwbSp09P3rx5cXBwYNasWSxatIhr165x8uRJ9u7dG+0lekmBxZfU/fDDD/Tr149WrVoRFBTE+fPnGTt2LI0bNyZDhgyMHz+ekSNH2iKriIiIxTzd3/wkigk5p7OzM8uWLWPBggXMmzePf/75BycnJ7y9vVm6dCllypQBXtwlbdy4cfTs2ZNnz55RpEgRFi1aRIkSJRKUvUmTJuzcuZOxY8fi6enJ1KlTqV69OgAlSpQgICCAadOm0aRJE1xcXKhYsSJfffWV6TK3xo0bs3PnTj744AN27NhhduyMGTPSrFkzxo8fz5UrV3jvvffM1k+ePJnJkyczePBgHj58SIECBZgxY0aU7ZIa1UYiIpKcvOnayBbni7y0LpK9vb3FtVHWrFlZvHgxo0ePpkmTJmTPnp1u3brh5+cX7fY9e/YkLCyMUaNGcfv2bTJnzkyLFi3o0qULABUqVKB48eJ8+umnTJgwIdpjvK6WcnR0ZN68eUyaNImPP/6Y8PBwSpQowcKFC0mbNi2VKlXC39+fBQsWMGXKFJycnKhevTq+vr4Je0FtxOKG061bt0yTYu7duxc7OzuqVasGQLZs2aJ8YikiIpJYwiOM+DWPehvcN3Vue7vXD62OjouLCz179jTdLS46mTJlYty4cTGu79WrF7169TJbtmfPHrPH5cuXN03WHSlr1qyMGjUqxuNWrFiRihUrxri+aNGi/PLLL6bHS5cuNVs/evRoRo8ebXr88vldXFwYMmQIQ4YMifbYTZs2pWnTpq99Dm+aaiMREUkuEqs2im9d9GrtEil79uxml9TB62uj6Hh7e7Ns2TKzZS1btjT9/HIdkypVKgYMGMCAAQOiPVb69OlZvXq12bJXazF4fS1VsmTJKJleFl09lFRZfEldlixZ+Pvvv4EX//ELFy5smtzq6NGjZMuWzboJRURE4im+DZ/kfm55s1QbiYhIcpFY9YnqoreTxQ2nDz74gDFjxtCxY0eOHDlCs2bNAPD392fGjBlmw9pEREREUjrVRiIiIiJRWXxJ3RdffIGLiwu///47/fv3Nw03O3HiBB06dKBbt25WDykiIiK2F9OwdYmdaiMRERGRqCxuOBkMBrp06WKaGCvSypUrrRZKREREJLlQbSQiIiISlcUNp99///2125QtWzZeYURERESSG9VGIiIiIlFZ3HBq06YNBoMBo9EIvPhU71WnT59OeDIRERGRZEC1kYiIiEhUFjeclixZEmVZSEgIgYGBbNq0iRkzZsQ7zKVLl2jatClDhw413ebv9OnT+Pv7c/LkSTJlykS7du1o27ZtvM8hIiIiYk22rI1EREREkiuLG07lypWLdnmNGjVwcXHhu+++Y86cORYHCQsL48svvyQkJMS07N69e7Rv3x4fHx/8/Pz4888/8fPzI02aNKY7wIiIiIgkJlvVRiIiIiLJmZ01D1amTBkOHz4cr31nzJhB2rRpzZatXr0aBwcHRo4ciZeXF82aNaNdu3bMnTvXGnFFREREbCohtZGIiIhIcmbVhtOePXtIkyaNxfv9/vvvrFq1irFjx5otDwwMpFy5cqRK9X8DsSpUqMDly5cJDg5OcF4RERERW4pvbSQiIiKS3Fl8SV108ydFRERw8+ZNrl+/zueff27R8R4+fMjAgQMZMmQI2bNnN1t38+ZNChQoYLYsS5YsANy4cQN3d3cL04uIyNskwhiBncGqn63Y9Nw+Pj40adKEXr162SSTr68v169fZ+nSpa/d1mg0snHjRqpVq4abmxvr169n0KBBnD171ibZkjNr10YiIiK2kli1UXzP6+PjA8DmzZujXBFlSV0DUWubmDx+/JjKlSuTJk0a9u3bh4ODg8W5X+flms/aNdbr6sk3WdNZ3HCKvAPLy+zs7ChQoABdunSxeG6lESNGULJkSRo2bBhl3ZMnT3B0dDRbljp1agCePn1q0XleZjQazeaKsobQ0FCz70lZcsoKymtLkRmfPn2Ks7NzjNtE93ufGJLja5scsoLt8hqNxmjv2PWm2BnsmBQ4gWuPrr3R83q4etC/zIA3es64GDx4MOHh4XHa9vfff8fX15fdu3cDUL9+fapWrWrLeMmWtWsjERERW0mM2iihddH169cZP348I0eOTFCOV2ubmPzwww+4ublx+/Ztdu7cSf369RN03tdJyTWWxQ2nuHYP42Ljxo0EBgayZcuWaNc7OTnx7Nkzs2WRjSYXF5d4nzcsLMxmtye+fPmyTY5rC8kpKyivLf3zzz9kyJAh2nWXLl1Kck2T5PTaJqesYJu8r35w8KZde3SNiw+CEjVDUuHq6hrnbV9tojg5OeHk5GTtSCmCNWsjgPv37zN58mT27t3L48ePKViwIP3796dMmTIAHDhwgAkTJhAUFET27Nnp1asXDRo0sGoGERFJuZJbbeTh4cGqVauoV68elSpVivdx4voh+rp166hatSr//PMPK1eutHnDKSXXWHFqOP3zzz8WHTRHjhxx2m7dunXcuXOHGjVqmC0fPnw4P/74I9myZePWrVtm6yIfZ82a1aJML3NwcCBfvnzx3j86oaGhXL58GU9PzxhHiiQVySkrKK8tRWaN7Xc2T548SWqEU3J7bZNDVrBd3gsXLljtWPLig5oFCxZw+fJl3N3dad68OV26dMHe3h6Aq1ev8s033xAYGEjatGnp0KEDK1asoFu3bjRt2jTK0PP58+fz/fffc/PmTbJkyUKzZs3o3r07hw8fNl0mVqtWLcaMGQNgNvz6v//+Y/LkyWzfvp3//vuPd999F19fX4oUKUJ4eDiTJ09m69at3Llzh1y5cvHZZ5/RokWLRHjVbMNWtRFAv379uH37NpMnT8bNzY2lS5fSsWNHNmzYgNFopEuXLrRv354JEyawd+9eBg4cSKZMmahYsaKlT0NERCTJa9SoEX/88QeDBw9my5YtUS6ti3T//n2mTZvGnj17uHfvHt7e3vTt25fy5ctz6NChKLVN06ZNoxwjKCiIY8eO0alTJx48eMCQIUO4dOkSefLkMW3j4+ND8+bNOXLkCL///jtZs2alc+fOfPTRR8CLS9amTp1Kz549mT59Ov/99x+VKlVi2LBh0fYxXr3ELbYaC2DNmjUsWbKEK1euYGdnh7e3N4MGDaJo0aKmY96+fZtOnTpx6NAhMmfOTMeOHWnVqlW0r9uzZ8+YNm0amzdv5vHjx+TPn5/evXtTpUqVuPzniVWcGk4+Pj4WXRIR19FDEydO5MmTJ2bL6tSpQ+/evWnUqBGbNm1i5cqVhIeHm4rpgwcPkidPnlivuXwdg8GQoBFSsXF2drbZsa0tOWUF5bWlyEtVo5MUmyXJ6bVNTlnB+nkT83K6lGbRokVMmjQJX19fKleuzLFjxxg5ciT37t1j8ODBhIaG0q5dO/LkycP333/P48eP8fPz49q16IfM79mzhzlz5jBlyhTy5MnDn3/+ycCBA8mVKxfvv/8+M2bMoFevXqxZs4YCBQrw448/mu3/xRdfcPnyZcaMGUPu3LmZPXs2HTp0YOfOnWzevJlt27YxZcoUsmbNys8//8yIESPInz+/aZROcmer2ujKlSv89ttvrFixgtKlSwMwdOhQfv31V7Zs2cKdO3coWLAgffv2BcDLy4u//vqLgIAANZxERCRFMhgM+Pv707BhQ8aNG8c333wTZZvw8HA6dOhAWFgYEyZMIFOmTCxZsoSOHTuyYsUKSpYsGaW2ic7atWtxcXGhWrVqPHnyBD8/P1auXMmgQYPMtvv222/p2rUrgwcP5pdffmHYsGGkSZPGNBrq7t27LF68mKlTp+Lo6MiIESPo1KkTGzZsMLspWnRiq7EOHz7MyJEjGTVqFGXKlOH27dt88803DBkyhE2bNpmOsXr1ar744gsGDx7M/v378ff3J0uWLLz33ntRzjdo0CCCgoKYOHGiqW7r2rUrM2fOjDI4yFJxajiNHj3aJn80xDRKyc3NjaxZs9KsWTMCAgIYPHgwnTp14vjx4yxatAg/Pz+rZxEREUmqjEYj8+bNo3Xr1qZPpzw9Pbl//z4TJkygd+/e7Nixg7t377J+/XrTJbITJkzgww8/jPaYV69exdHRkZw5c5IjRw5y5MhBlixZyJEjB46OjqRPnx6ATJkyRRnmffHiRX755Rfmz59v+vRrxIgRpEuXjnv37nH16lVcXFzIlSsXWbJkoXXr1uTNm9fs08Hkzla1UcaMGZk7d67Zp5QGgwGDwcDDhw8JDAykdu3aZvtUqFABf3//RJ8zTURExFZy5szJV199xbBhw6hbt26U0Tf79+/n1KlTbNmyxdRM8vPz48SJE8yfP59p06bFWtsAPH/+nM2bN+Pj42O6zK1KlSps3LiRfv36mX1IX6VKFXr27AlA3rx5OXbsGIsXLzY1nMLCwhg3bpxpVNKECROoX78+Bw4ciHW+ptfVWBkyZMDf359GjRqZXpfmzZtHmd+qdu3adO3aFcD0weKCBQuiNJyuXLnC1q1b2bhxI4ULFwagffv2nDlzhvnz57+ZhtOrQ80iIiI4d+4chQoVAl4M1/rrr7+oXLnya7t1lnBzcyMgIAB/f3+aNGlC5syZGThwIE2aNLHaOUTeFuERRuztDGaP9YeJSPJw9+5dgoODTSNeIpUrV46wsDAuXrzIX3/9RZ48eczmYytUqFCM8zY1atSIdevWUbduXfLly0elSpWoW7dunC79OnfuHAAlSpQwLUudOrXp079WrVqxa9cuqlevTuHChalcuTINGjRI0OjkpMZWtVG6dOmoXr262bLt27dz5coVvv76azZs2EC2bNnM1mfJkoXQ0FDu3btHpkyZ4vV8rH1DFd04wXIGg+G1o4ojb+SR2HnjkhVe5EvsrJZSXttJTlkh5d5QJTn75JNP2L59O0OGDGHr1q1m686dO4erq6vZyCWDwUCZMmXYv39/nI6/b98+goODzeZFbNCgAT///DM//fQTjRs3Ni0vX7682b4lS5Zk7969psdp0qQxNZvgxYjk9OnTc+7cuVgbTq+rsTw9PQkKCmLWrFlcvHiRK1eucPbsWSIiIsyO82rNWLx4cfbt2xflfH/99RcALVu2NFseFhZGunTpYswZVxZ3h/799186depEaGgou3btMoXs0qULJUqUYPbs2TFOPhwXr96ar1ixYqxatSrexxORF+ztDAxfe5zLwY/xdE+LX/NiiR1JROIopnnUIouLVKlSYW9vH6XYiE2mTJnYtGkTR48e5bfffmP//v0sWbKEXr16mT6xi8nrGiienp7s2LGDw4cP89tvv7F3717mzZvHmDFjUuSHRrasjf744w8GDRpEnTp1qFGjRrR38I18/OqNVixhqxuq6MYJcefs7Iy3t3es27x6I4/EyhuXrGCeV+8F20pOeZNTVkiZN1RJzkaNGkXDhg1N80tGiqlWMhqNcf7gZ/369QDR1kErV640azi9esyIiAjs7OxMjx0cHKIc4+WpgmLyuqxbtmzB19eXhg0bUqpUKT799FPOnTsXZYTTy1ki80X3vot83ZYvX06aNGliPUZ8WNxwGj9+PM+ePWPixImmZdWrV2f9+vX069ePSZMmRXtNpYgkvsvBjzl741FixxARC7m7u+Pu7s6RI0fMLqcKDAzEwcGB3LlzU6hQIVavXs39+/dNzY2goCAePYr+d37z5s08evSIVq1aUbp0aXr37s2QIUP48ccf6dmzZ6yfvnp5eQFw4sQJ07xBz58/p06dOgwcOJBbt27h5uZGgwYNqFy5MgMHDqR9+/b8+OOPKbLhZKvaaNeuXXz55ZeUKlXKdOzUqVNHaSxFPk7InHvWvqGKbpxgubiMeIi8kUdi543r6Iw8efIQEhKS6K+tJRL7tbVUcsqbnLKCbqiSVOXIkQNfX1+GDBmCh4cH2bNnB6BgwYI8evSIc+fOmUY5GY1Gjhw5Yvr3Lbb/d925c4d9+/bRtGlT2rdvb7Zu0aJFrFu3zuzYJ06cMNvmjz/+MGvE379/n2vXruHh4QHA+fPnefz48Wub9a+rsebOnUvz5s3NphnavXu36flGPsdTp06ZHffIkSPkz58/yvkil92+fdss25QpU7Czs6NPnz6x5n0dixtO//vf/xg5cqTZEC8Ab29v+vTpg7+/f4ICiYiIvM2uXLnCL7/8YrbMycmJjh07MmXKFDw8PKhcuTLHjx9n5syZfPLJJ7i6uvLBBx8wY8YMvvzyS7788kuePHli+rQrugLr6dOnjBs3jjRp0lCmTBlu3rzJ77//bprUO3Ly+DNnzpAxY0azffPkyUOdOnXw8/NjxIgRZM2alblz5/L06VPKlSvHkiVLmDVrFk5OThQqVIiLFy9y+vRp091hUhpb1EbLli3D39+fevXqMW7cONOnktmzZ4/2Dr4uLi4xXj4ZF7a6ocrbfuMEa3v1j97kkjcyZ3JoMkReLphc8kZK6u+FlyWnrKAbqiRFH330Edu2bWP//v2mhlOVKlUoXLgw/fv3Z+jQobi5ubFs2TLOnTvH8OHDgai1zcsjejZv3szz58/5/PPPyZs3r9n5unbtyoYNG1i5ciXDhg0D4IcffqB48eJUrlyZXbt2sXPnTmbPnm2234ABAxgyZAjPnz/Hz8+PkiVLUrZs2Vif2+tqrPXr1/PHH39w6tQpXF1d2bNnD8uWLQNefAAVOc/UDz/8QKFChahRo4Yp3+LFi6OcL3/+/NSsWZPhw4czbNgw8ufPz7Zt25gzZ06UUWTxYXHD6dmzZzEOA3N2dua///5LcCgRERFr8XD1SFbn3LJlC1u2bDFbljNnTvbs2YOjoyOLFy9m9OjRZMuWjc8//5yOHTsCL4bnBwQEMHLkSD7++GPSp09P165dOXXqVLTDuj/66CPu37/Pt99+y40bN0ifPj1169blyy+/BKBAgQJUr16dL774gn79+kW5JGz06NGMHz+ePn368OzZM4oXL878+fPJlCkTPXv2JCwsjFGjRnH79m0yZ85MixYt6NKlS7xfl6TM2rXRihUr+Oabb2jTpg2DBw82++OkTJkyHD582Gz7gwcPUqpUKasMfRexBrvMmU1zR0Z3+d2r80omtpfzJIe8Ign1pmsjW5wv8tK6SPb29ixYsIBx48bRs2dPnj17RpEiRVi0aJHpA6FXa5sOHTqY9l+/fj2VKlWK0mwCyJ07N7Vr12bz5s2mOqlJkybs3LmTsWPH4unpydSpU6PMwdiwYUM6d+7Ms2fP8PHxifJvekxiq7GGDh3KsGHDaN26NY6OjhQqVIjx48fTt29fTpw4YfrgsGPHjvz8889MnjyZnDlzMmnSpCjzTkWaMmUKU6ZMYdiwYTx48IDcuXOb5tFOKIsbTsWLF2fhwoVUrVrVrIB9/vw5S5YsoVgxzQsjIiJJQ4Qxgv5lBiTaue0MljUA9uzZE+v61q1b07p162jX/f3339y4ccPs06t///2XkSNHmiaZHjt2rNk+n3/+OZ9//nm0x3N0dGTu3Llmy16eKNvV1ZVvvvkm2kvFUqVKxYABAxgwIHFe+zfNmrXRpUuXGD16NO+99x5dunQhODjYtM7JyYk2bdrQpEkTJk6cSJMmTdi3bx/btm0jICDAqs9JJCHs0qczmzvyZUlxHsmYskLSzCuSEIlVG8WnLoKYa6Ps2bMTGBhotixTpkyMGzcuxmNFV9tEevXDvlfNmDHD7HHWrFkZNWpUrPu0atXKdHfhV738vJo2bRrnGsvDw4OFCxdGWR55d7xXjx2dV8/n7OzM119/zddffx3rfvFhccOpd+/etGnThlq1alGtWjXc3Ny4e/cuv/32G3fu3GHp0qVWDykiIhIf8Slskuu5nz59SufOnenfvz916tTh0aNHTJ06FU9PT4oXL/5Gs7xtrFkbbd++nbCwMHbu3MnOnTvN1jVp0oSxY8fy7bffMmHCBBYvXkyuXLmYMGGCaZ4HkaQkOc0dmZyyiiREYtVGiVmTSeKxuOFUokQJVq1axezZs9m7dy/379/H1dWVMmXK0L17dwoXLmyLnCIiIhILLy8vJk+ezOzZs5k+fTpOTk5UrFiRhQsXRntJnViPNWujrl270rVr11i3qVatGtWqVUtobBERERGbsrjhBC8mwZw+fbq1s4iIiEgC1KtXj3r16iV2jLeSaiMREZG3g6WXrL3NNK5NRERERERERESsSg0nERERERERERGxKjWcRERERERERFIYo9GY2BEkBbLkfaWGk4iIiIiIiEgKEXmzkJCQkEROIilR5PsqLjelsXjS8I4dO9KpUyfdfldEREQE1UYiIpK02NvbkyFDBm7dugWAi4sLBoMhkVNJcmc0GgkJCeHWrVtkyJABe3v71+5jccPpjz/+0JtVRERE5P9TbSQiIklNtmzZAExNJxFryZAhg+n99ToWN5yqVq3K5s2bKV26dJyGUImIiIikZKqNREQkqTEYDGTPnp0sWbIQFhaW2HEkhXBwcIjTyKZIFjecUqdOzebNm/npp5/w8vLCxcXFbL3BYGDx4sWWHlZERMTqIiIisLNLnOkK43NuHx8fADZv3kzatGnN1vn6+nL9+nWWLl1qtYxv8vgpmWojERFJquzt7S1qEIhYk8UNp5s3b1KyZEnT41dnKNdM+CIiklTY2dmxe/J+7l97+EbPm8EjHbX6VYnXvtevX2f8+PGMHDnSyqliN3jwYMLDw9/oOVMK1UYiIiIiUVnccNInnyIikpzcv/aQ4It3EztGnHl4eLBq1Srq1atHpUqV3th5XV1d39i5UhrVRiIiIiJRxfs6gwcPHrB7926+//577t69y8WLF/UJnoiISAI1atSIihUrMnjwYB4/fhztNgULFmT9+vUxLpsxYwatW7emb9++lCpVim+++Ybw8HAmTJhA9erVKVKkCPXq1eP777837e/r60ubNm0AOHToEN7e3uzbt48PPvjAtP2uXbtM24eHhzNlyhSqVKlCiRIl6N27N/7+/qZjvI1UG4mIiIj8n3g1nL777juqV69Ojx49GDlyJDdu3GDMmDF89NFHPHz4Zi9bEBERSUkMBgP+/v48ePCAcePGxfs4v//+O+7u7mzatIk2bdqwYsUKtm3bxpQpU9i+fTutW7dmxIgRBAYGRrt/ZINq8ODBbN26lQIFCvDVV1/x33//ATBx4kRWrVrF8OHDWbduHZkzZ36rR/qoNhIRERExZ3HDadmyZcyYMYP27duzevVq0yd3rVu35tq1a0ybNs3qIUVERN4mOXPm5KuvvmL16tXs378/3sfp3bs3Hh4eeHp6cvXqVVxcXMiVKxc5c+akdevWLFy4kDx58sS4/xdffEHFihXx9PSke/fuPH78mHPnzhEaGsqKFSv44osveO+99/Dy8mLIkCF4e3vHO2typtpIREREJCqLG05Lly6lc+fO9OnTh3fffde0vHr16nzxxRfs2bPHqgFFRETeRp988gmVK1dmyJAhMV5aFxs3NzezeZlatWrF48ePqV69Ok2bNmXSpElkypQJNze3GI+RN29e08+Rd80LCwsjKCiIJ0+eUKJECdN6g8FA6dKlLc6ZEqg2EhEREYnK4obTP//8Q7ly5aJdlzdvXoKDgxMcSkRERGDUqFE8evSIMWPGxLrd8+fPoyxzcnIye+zp6cmOHTsICAigQoUK7N27l8aNG7Nhw4YYj+vo6BhlmdFoJFWqVKafRbWRiIiISHQsbjhlz56do0ePRrvu5MmTZM+ePcGhREREBHLkyIGvry9r1641m2vJwcHBbNTTlStXXnusJUuWsGPHDipXrszAgQPZsmULFStW5Mcff7Q41zvvvIOTkxN//vmn2fJjx45ZfKyUQLWRiIiISFSpLN2hefPmzJgxAycnJ2rUqAFASEgI27dvZ86cObRv397aGUVERN5aH330Edu2bWP//v2mxkWJEiVYs2YNZcuWxWg0MmbMmGhHI73s7t27zJo1CycnJwoVKsTFixc5ffo0bdu2tTiTs7Mzbdq0Yfr06WTOnBkvLy9Wr17NsWPHYhzpk5KpNkr+DAYDzs7OGAyGxI4iIiKSYljccPr888/5+++/mThxIhMnTgQwFasNGzakS5cu1k0oIiKSABk80iX7c44aNYqGDRuaHo8YMYIRI0bw8ccfkyVLFvr06cPNmzdjPUbPnj0JCwtj1KhR3L59m8yZM9OiRYt4/7vdp08fwsLCGDJkCKGhodSsWZNatWrx9OnTeB0vOVNtlDyFRxixt3vRYHJ2do4y6f3L6yX5M4aHY7C3T+wYIiJvFYsbTgaDgZEjR9K+fXsOHTrE/fv3cXV1pWzZshQoUMAWGUVEROIlIiKCWv2qJNq57ewsu3I9psmls2fPbnZJXb58+Vi2bJnZNo0aNTL93KtXL3r16mW2PlWqVAwYMIABAwZEe46xY8eafi5fvjxnz541W58rVy6zZXv37qVLly4MGjTItKxDhw5ky5YtpqeXYqk2ipsIYwR2huh/J2JbZyv2dgaGrz3O5eCok/J7uqfFr3mxN5pHbMtgb8/dnr14fv6C2fLUNWuQ3verGPfLlNYxyb13Y5OcsopIymdxwylSnjx5MBgMPHjwgEyZMuHh4WHNXCIiIglmacMnpZz7TZg/fz4rVqxg4MCBpE2blt27d3Pw4EEWLFiQ2NESjWqj2NkZ7JgUOIFrj66ZLfdw9aB/megbobZ2OfgxZ288SpRzy5v3/PwFwk6eNFuWKp9XrPu4OjkkyfduTJJTVhFJ+eLVcFq5ciXffvstt2/fNi3LkSMH/fv3p379+lYLJyIiIknTxIkTGTt2LO3atePJkyfky5ePadOmUaFChcSOlihUG8XNtUfXuPggKLFjiFgsOb13k1NWEUnZLG44ff/99/j5+VGrVi3q1KmDm5sbwcHBbNu2jf79++Po6Ejt2rVtkVVERESSiFy5cjFz5szEjpEkqDYSERERicrihtOiRYto2bIlw4YNM1veuHFjhg0bxqxZs1RUiYiIyFtDtZGIiIhIVBZPMHHz5k1q1aoV7bq6dety8eLFBIcSERERSS5UG4mIiIhEZXHDqWjRovz666/Rrjt69CgFCxZMcCgRERGR5EK1kYiIiEhUFl9S161bN/r168d///3Hhx9+SNasWbl37x67d+9m4cKFfP311/z++++m7cuWLWvVwCIiIiJJiWojERERkagsbjh17NgRgDVr1rB27VrTcqPRCICfn5/pscFg4PTp09bIKSIiIpIkqTYSERERicrihtOSJUtskUNEREQkWVJtJCIiIhKVxQ2ncuXK2SKHiIiI/H+bN29m2bJlnDt3DoPBQN68efnoo4/49NNPEzvaa4WEhLBhwwZatWoFgK+vL9evX2fp0qWJnMx2VBuJiIiIRGVxw0lERCTZiAgHO/tkde61a9fi7+/P4MGDKV26NEajkd9++41Ro0YRHBxMz549bRDWehYsWMD69etNDafBgwcTHh6eyKlERERE5E1Tw0lERFIuO3tY1wqC3/CcOe6FodnyeO26YsUKmjVrRvPmzU3L8ubNy7///suSJUuSfMMpct6iSK6uromUREREREQSkxpOIiKSsgWfhhtHEztFnNnZ2XH06FEePHhA+vTpTcs7d+5Ms2bNAHj27BnTpk1j8+bNPH78mPz589O7d2+qVKkCwPr165k6dSo9e/Zk+vTp/Pfff1SqVIlhw4aRNWtWAP755x8mTJjAwYMHefjwIW5ubjRs2JD+/ftjZ2fH+vXr+e6776hevTobNmygfPnyfPvtt+zatYs5c+Zw/vx5wsPDyZ8/P3379qVq1arMmDGDmTNnAlCwYEF2797NzJkzzS6pCwoKYsKECRw9epTnz59TuXJlvvrqK3LmzAlAmzZtKF68OHfv3mXHjh1ERERQs2ZN/Pz8SJs27Rv77yAiIiIiCWOX2AFERETk/3Tq1Im//vqLatWq0blzZ+bOncvx48dxdXUlT548AAwaNIjffvuNiRMnsmHDBt5//326du3K3r17Tce5e/cuixcvZurUqSxevJgbN27QqVMnnj9/DkC3bt149OgRCxcuZNu2bXTo0IGAgAD27NljOsbVq1e5desWGzdupG/fvpw8eZJevXrRoEEDtmzZwurVq8mUKRMDBw7k2bNndOjQgQ4dOpAtWzb2799P9uzZzZ7b9evX+eSTT3B0dGTx4sUsWLCA27dv07p1ax4/fmzabtGiRbi7u7N27VomTJjA7t27WbRoke1edBERERGxOosbTsOGDePYsWO2yCIiIvLWq1evHt9//z21atXi2LFjTJo0iY8++oh69epx5MgRrly5wtatWxkzZgzly5fH09OT9u3b06BBA+bPn286TlhYGOPGjaNMmTIUK1aMCRMmcO7cOQ4cOMCTJ0/48MMP+eabbyhUqBAeHh60a9cOd3d3zp49a5ane/fueHh4kD9/fuzt7Rk6dCjt2rXDw8ODwoUL07ZtW+7evcudO3dIkyYNLi4u2NvbkzlzZuztzeewWrFiBS4uLkycOJFChQpRvHhxpk+fzp07d9i0aZNpu3z58tGvXz88PT2pVasWlStX5ujRpDtKTbWRiIiISFQWX1J38OBB1qxZg6enJ02bNuXDDz8kS5YstsgmIiLyVipRogQlSpQgIiKCM2fOsG/fPpYtW8bnn3/OqFGjAGjZsqXZPmFhYaRLl870OE2aNBQpUsT02MvLi/Tp03Pu3DmqVq1K69at2bZtG8ePH+fKlSucPXuW4OBgIiIizI7r6elp+rlw4cKkT5+euXPncvHiRa5cucKZM2cA4jQx+Llz5yhSpAiOjo6mZZkzZyZPnjycO3fOtCxv3rxm+7m6uvLw4cPXHj+xqDYSERERicrihtOOHTs4cuQIGzduZO7cuUydOpXKlSvTtGlTatWqhYODgy1yiohgMBhwdnbGYDAkdhQRm7h58yZz5syhS5cuZMuWDTs7O7y9vfH29qZ27dp88MEHpm2XL19OmjRpzPa3s/u/gcvR/XscHh6Ovb09ISEhtG7dmidPnlCvXj2aNGlCsWLFTHeWe5mTk5Pp58OHD9OxY0dq1KhB6dKladiwIaGhofTo0SNOz+/VCcUjRUREmOV9uSGVHKg2EhEREYkqXpOGly5dmtKlSzNkyBB27tzJtm3bGDRoEI6OjjRs2JCPP/6YAgUKWDuriLzFIowRODs74+3tbXpsZ9A0dJKyODo6smbNGrJnz07nzp3N1kWOXnJ3dwfg9u3bpt8HgClTpmBnZ0efPn0AuH//PteuXcPDwwOA8+fP8/jxY7y9vdm/fz+nTp3it99+Mx3v/v373LlzJ8amEMCCBQsoX748M2bMMC2LnAw8cr/YGsIFCxZk8+bNPHv2zNRUCg4O5sqVK1FGbCU3qo1E4iZTWsdY/w3Xv+8iIilHgv5vnjp1asqWLUuZMmXw8vLiwYMHbNmyhQ8//JDPP/+cf//997XHuHPnDgMGDKBChQqULFmSzp07ExQUZFp/+vRpWrduTYkSJfDx8WHJkiUJiSwiyZSdwY5JgRP44ufeTAqcoGJUUqRMmTLRqVMnpk2bxpQpUzh9+jTXrl3j559/pmfPnpQvX55y5cpRs2ZNhg8fzp49e7h27Rrz5s1jzpw55M6d2+x4AwYM4OTJk/z5558MHDiQkiVLUrZsWbJlywbA5s2buX79OoGBgXTv3p2wsDCePXsWY77s2bNz9uxZAgMD+fvvv1m3bh3Tpk0DMO3n4uLCgwcPuHTpEmFhYWb7t2jRgv/++48BAwZw5swZjh8/Tp8+fciYMSMNGjSw5kuZaKxRG4mkZK5ODmb/pr/8pX/fRURSlniNcAoJCWH79u1s2rSJw4cP4+Liwvvvv8/QoUMpXrw4x44do2/fvvTp04eVK1fGeqwePXoQERHB3LlzSZMmDdOmTaNdu3bs2LGDJ0+e0L59e3x8fPDz8+PPP//Ez8+PNGnSmG4NLSJvj2uPrnHxQdDrNxR5mXvhZHXOL774Ak9PT1avXs3y5ct58uQJOXLk4P3336dLly7Ai9FMU6ZMYdiwYTx48IDcuXPj7+9PkyZNzI7VsGFDOnfuzLNnz/Dx8WHw4MEYDAaKFSvGoEGDWLRoEVOnTiVr1qzUr1+f7Nmzc+LEiRiz9e7dm+DgYLp27Qq8mNx79OjRDBgwgBMnTuDl5UWdOnVYvXo1jRo1YtmyZWb758qVi2XLljFhwgTT3eoqV67MhAkTzOafSo6sWRuJvA30b7qISMpnccPpyy+/ZPfu3YSGhlK6dGn8/f2pV68ezs7Opm2KFy9O48aNX3sL4wcPHpAzZ066dOliGmbevXt3PvzwQ86fP8+BAwdwcHBg5MiRpEqVCi8vL65cucLcuXPVcBIRkdeLCIdmyxPv3Hb2r98uGo0bN6Zx48Yxrnd2dubrr7/m66+/jvU4rVq1inZeJoB27drRrl27GPdt2rQpTZs2NVuWMWNGs8vpItWpU8f0c65cudi2bZvpcfHixc229fb2ZuHChTGeN/ISvZeNHTs2xu2TAmvWRiIiIiIpRbzuUte6dWuaNWtmdueaV1WoUIGCBQvGeqz06dMzadIk0+O7d++yaNEismXLRr58+ZgxYwblypUjVar/i1mhQgXmzJlDcHCwad4JERGRaMWz4ZPszy1vlDVrIxEREZGUwuKGU40aNahdu3asBRVAuXLlLDru0KFDWb16NY6Ojnz33Xe4uLhw8+bNKBNsRt5m+MaNG2o4iYiISKKzVW0EMGfOHPbv32828mvIkCGsWbPGbLucOXOyZ88ei48vIiIiYisWN5y2bNlC/fr1rR7ks88+45NPPmH58uX06NGDFStW8OTJkyi3Rk6dOjUAT58+jfe5jEYjISEhCcr7qtDQULPvSVlyygrKaw0Gg8Hs0o5Ikb9HT58+jXY9vHgesd216k2IKX9SyBaTpPg+iI2t8hqNxljvWia2Ed3lcGI7tqqNli9fztSpUylTpozZ8rNnz9K1a1dat25tWmZvrxF1IiIikrRY3HAqWbIkBw8epFKlSlYNki9fPgD8/f05duwYy5Ytw8nJKcrdciL/QHZxcYn3ucLCwjh9+nT8w8bi8uXLNjmuLSSnrKC8CeHs7Gx2+/RI//zzj+l7hgwZot330qVLid40iSl/Usj2OknpfRAXtsj76gcHIimNtWujf//9l+HDh3Po0KEoo6aMRiMXLlygc+fOZM6c2SrnExEREbEFixtOBQsWZMGCBWzfvp1ChQpFafwYDAZGjx4dp2PdvXuXAwcOULduXdM8TXZ2duTLl49bt26RLVs2bt26ZbZP5OOsWbNaGt3EwcHB1OCyltDQUC5fvoynp2eMI0WSiuSUFZTXGmIaYZIjRw6CgoLIkSNHjPvmyZMn0UcRxZQ/KWSLSVJ8H8TGVnkvXLhgtWOJJFXWrI0ATp06hYODA5s3b2bWrFlcv37dtO7q1auEhISQN29eq+UXERERsQWLG047d+4kS5YshIWFRXvrZEsunQgODqZfv34EBARQtWpV4MXoo7/++gsfHx/c3d1ZuXIl4eHhpqHiBw8eJE+ePLi5uVka3SxjQkZIxcbZ2dlmx7a25JQVlNcWIi9RjfwenaTcLEnK2SIlh/fBy6yd901eTpdUm4+SuN7E+8KatRGAj48PPj4+0a47d+4c8OJufr/88gt2dnZUq1aNvn374urqanl4ERERERuxuOFkzQkpCxQoQLVq1Rg1ahSjRo0iffr0zJkzh4cPH9KuXTtSp05NQEAAgwcPplOnThw/fpxFixbh5+dntQwiIpK8OTg4ABASEpIsmpDyZkXO2Rj5PrGFNzlZ97lz57CzsyNLlizMnj2bq1evMn78eM6fP8/ixYuxs7OL13GtPb/lq/PCxTQX36v7vKnGcVzyQNLLFJknsecJjOvrF1/Wft2TW15Lzx35Pan9nkV37pe/J3Wa31Ik4SxuOEV6+PAhf/75J48ePSJTpkwULVqUtGnTWnycyZMnM2nSJPr27cujR48oU6YMy5cvN13iExAQgL+/P02aNCFz5swMHDiQJk2axDe2iIikMPb29mTIkMF0ybWLi4sKOTE1UG7dukWGDBneyKTa1qqNYtOtWzdatmxJxowZgRcf3mXOnJmPP/6YEydOULx48Xgd11bzW0bOCxfTXHwve5Pz8sUlDyS9TK/mSax5AuP6+sWXtV/35JY3Pi5fvpzkfs9iovktNb+lvD3i1XCaO3cu3377LU+ePDEtc3R0pEuXLvTo0cOiY7m6ujJixAhGjBgR7fpixYqxatWq+MQUEZG3RLZs2QCizPsnkiFDBtP7w5asWRvFxs7OztRsipQ/f34Abt68Ge+Gk7Xnt3x1Xri4NIHf5Lx8cW1KJ7VMkXkSe55AWzf1rf26J7e8lnj5vRCXS+KTStbkMCJZ81uKJJzFDad169YxefJkmjdvTqNGjXB3d+f27dts2rSJmTNnkiNHDo1AEhGRN8pgMJA9e3bTPDoi8KKJ8iZGNr3J2mjgwIHcunWLRYsWmZZFzhuVkIaRrea3tGReuKT4B2hSy/RqnuQ2T2BcJbXX/XWSQl5nZ+c45UgqWZPT+zY5z28pktgsbjgtWrSIFi1aMHz4cNOyvHnzUr58eZycnFiyZIkaTiIikijs7e3fSINB5GVvsjaqW7cu3bt3Z+bMmTRq1IhLly4xcuRIPvjgA7y8vKxyDhERERFrsHhmyStXrlC7du1o19WqVYuLFy8mOJSIiIhIcvEma6NatWoxdepUdu/eTcOGDRk8eDB16tRh9OjRVjuHiIiIiDVYPMIpa9as/PPPP9Gu+/vvv60+OaaIiIhIUmbL2mjs2LFRlr3//vu8//778T6miIiIyJtg8QgnHx8fpk2bxvHjx82WHzt2jBkzZuDj42O1cCIiIiJJnWojERERkagsHuHUq1cv/ve///HJJ5+QM2dO3N3dCQ4O5vr163h5edG/f39b5BQRERFJklQbiYiIiERlccMpbdq0rF27lnXr1vH777/z4MEDihYtSocOHWjatClOTk62yCkiIiKSJKk2SnkypXUkwhiBnSH6iwFiWyciIiIvWNxwAkidOjUtW7akZcuW1s4jIiIikuyoNkpZXJ0csDPYMSlwAtceXTNb5+HqQf8yAxIpmYiISPJhccNp0KBBMa6zs7PDxcUFT09P6tevT8aMGRMUTkRERCSpU22Ucl17dI2LD4ISO4aIiEiyZHHD6ebNm/zxxx88ffrUNE/BnTt3+Pvvv7G3tzc9/u677/j+++/x8PCwRW4RERGRJEG1kYiIiEhUFl98XrNmTVxdXVm5ciW7d+9m1apV7Nq1i/Xr15M1a1a6d+/OgQMHyJ07N5MnT7ZFZhEREZEkQ7WRiIiISFQWN5wWLVpE//79KVGihNlyb29v+vTpw5w5c3B1daV9+/YcOnTIWjlFREREkiTVRiIiIiJRWdxwunfvHpkyZYp2Xfr06blz5w4AGTNmJCQkJGHpRERERJI41UYiIiIiUVnccPL29iYgIIBnz56ZLX/27BkLFiygcOHCAJw6dYrs2bNbJ6WIiIhIEqXaSERERCQqiycN//LLL2nfvj21atWievXquLm5ERwczC+//MLjx48JCAggMDCQyZMn061bN1tkFhEREUkyVBuJiIiIRGVxw6lkyZKsX7+e2bNn8+uvv3L37l2yZctG1apV6dq1K7lz5+bAgQP07t2bjh072iKziIiISJKh2khEREQkKosbTgB58+Zl/PjxMa6vWLEiFStWjHcoERERkeREtZGIiIiIuXg1nJ49e8batWv53//+x+3btxk9ejSHDx/m3XffpVixYtbOKCIiIpKkqTYSERERMWfxpOF3796lWbNm+Pv7c+XKFY4fP86TJ0/4+eefadOmDUePHrVFThEREZEkSbWRiIiISFQWN5zGjx/Pf//9x48//siGDRswGo0AzJgxg6JFizJ9+nSrhxQRERFJqlQbiYiIiERlccPp559/pk+fPrzzzjsYDAbT8tSpU9OhQwdOnTpl1YAiIiIiSZlqIxEREZGoLG44PX36lAwZMkS7zt7enrCwsIRmSpYMBgPOzs5mhaaIiIikfKqNRERERKKyuOFUtGhRVqxYEe26LVu2UKRIkQSHSk4ijBEAODs74+3tjbOzs2mZiIiIpHyqjURERESisvgudX369KFdu3Z8+OGHVK9eHYPBwNatW5kxYwb79+8nICDAFjmTLDuDHZMCJ3Dt0TUAPFw96F9mQCKnEhERkTdFtZGIiIhIVBaPcCpTpgwLFy7E2dmZgIAAjEYjixYt4vbt28yZM4cKFSrYImeSdu3RNS4+COLigyBT40lERETeDqqNRERERKKyeIQTQNmyZVm5ciVPnjzhwYMHpE2bljRp0gAQHh6Ovb29VUOKiIiIJGWqjURERETMWTzCqVatWpw5cwYAJycnsmbNaiqojh8/TqVKlaybUERERCQJU20kIiIiElWcRjht3bqV58+fA3D9+nV27txpKqxeduDAAd2JRURERFI81UYiIiIisYtTw+nEiRMsXrwYAIPBwKxZs2Lctn379tZJJiIiIpJEqTYSERERiV2cGk79+/enbdu2GI1GateuzcyZMylcuLDZNvb29qRNm5a0adPaJKiIiIhIUqHaSERERCR2cWo4OTo6kjNnTgB2795NlixZcHBwsGkwERERkaRKtZGIiIhI7Cy+S13OnDk5fvw4hw4d4tmzZxiNRgCMRiMhISEcOXKE1atXWz2oiIiISFKk2khEREQkKosbTsuXL2fUqFGmYupldnZ2VKlSxSrBRERERJID1UYiIiIiUdlZusOyZcuoVq0ahw4dokOHDnz88cf8+eefTJs2jdSpU9OoUSNb5BQRERFJklQbiYiIiERlccPp77//pmXLlqRPn54iRYpw5MgRnJycqFu3Lp07d2bJkiW2yCkiIiKSJKk2EhEREYnK4oaTg4MDTk5OALzzzjtcuXKFsLAwAEqXLs3ly5etGlBEREQkKVNtJCKJxWAw4OzsjMFgSOwoIiJRWNxwKly4MD///DMAefLkISIigmPHjgFw8+ZN66YTERERSeJUG4nImxQe8X/zxTk7O+Pt7Y2zs3MiJhIRiZ7Fk4a3b9+enj178vDhQ0aPHk2tWrUYOHAgderUYcuWLZQuXdoWOUVERESSJNVGIvIm2dsZGL72OJeDH5str5gvM11r50+kVCIiUVk8wql27drMnj0bLy8vAEaOHImnpycrV64kb968DBs2zOohRURERJIq1UYi8qZdDn7M2RuPzL7+uR+S2LFERMxYPMIJoEaNGlStWhWAjBkzMmvWLJ4/f46rq6tVw4mIiIgkB6qNRERERMxZPMIpLCyM4cOH8/HHH5uWHT16lIoVKzJu3DgiIiKsGlBEREQkKVNtJCJvM2N4eILWi0jKZfEIpxkzZrB582Z69+5tWubt7c2XX37JjBkzyJgxI507d7ZqSBEREZGkSrWRiLzNDPb23O3Zi+fnL0RZlyp/PjLNnJEIqUQkKbC44bRlyxa++uorPv30U9OyDBky0K5dO1KlSsWSJUtUVImIiMhbQ7WRiLztnp+/QNjJk4kdQ0SSGIsvqbt37x4eHh7RrsubN69u/ysiIiJvFdVGIiIiIlFZ3HDKmzcv27dvj3bdnj17eOeddyw63v379xk2bBjVqlWjVKlStGjRgsDAQNP6AwcO0LRpU4oXL069evX44YcfLI0sIiIiYjPWro1EREREUgKLL6lr27Ytvr6+3L9/n9q1a+Pm5sbdu3f5+eef+emnnxgzZoxFx+vXrx+3b99m8uTJuLm5sXTpUjp27MiGDRswGo106dKF9u3bM2HCBPbu3cvAgQPJlCkTFStWtDS6iIiIiNVZuzYSERERSQksbjg1btyY//77j2+//ZYdO3aYlmfMmJGhQ4fSuHHjOB/rypUr/Pbbb6xYsYLSpUsDMHToUH799Ve2bNnCnTt3KFiwIH379gXAy8uLv/76i4CAADWcREREJEmwZm0kIiIiklJY3HACaNWqFS1btuTSpUvcv3+fdOnSkTdvXuzsLLtCL2PGjMydO5eiRYualhkMBgwGAw8fPiQwMJDatWub7VOhQgX8/f0xGo0YDIb4xBcRERGxKmvVRq+aM2cO+/fvZ+nSpaZlp0+fxt/fn5MnT5IpUybatWtH27ZtE/oURERERKwq3lWQwWAgb968lCpVinz58sWroEqXLh3Vq1fH0dHRtGz79u1cuXKFqlWrcvPmTbJly2a2T5YsWQgNDeXevXvxjS4iSYwxPDzWxyIiyYE1aqOXLV++nKlTp5otu3fvHu3btyd37tysW7eOHj16MHHiRNatW5egc4mIiIhYW7xGONnKH3/8waBBg6hTpw41atTgyZMnZs0owPT42bNn8T6P0WgkJCQkQVnhRWHp7Owc7brQ0FCMRmOCz2ELoaGhZt+TutjyxjTKLTFf+5fzRpcvMbLF9F59+vSp6XtivZcjs93t2Yvn5y+QKn8+Ms2cYXbemPLr98x6bJVXo1FFLPfvv/8yfPhwDh06hKenp9m61atX4+DgwMiRI0mVKhVeXl5cuXKFuXPn0qxZs8QJLCIiIhKNJNNw2rVrF19++SWlSpVi4sSJAKROnTpKYynycUx/HMdFWFgYp0+fjn/Y/8/Z2Rlvb+9o1126dCnJ/6F5+fLlxI5gkVfzOjg4UMS7MHapHMyWRzwP4+RfpwkLC3uD6aK6fv0673q/i30qe9Oy8OfhnPrr1BvPFtN79Z9//jF9z5AhQ7T72vq9HJnt+fkLhJ08Ge15Y8qv3zPrs0XeVz84EJHYnTp1CgcHBzZv3sysWbO4fv26aV1gYCDlypUjVar/K+EqVKjAnDlzCA4Oxt3dPTEii4iIiESRJBpOy5Ytw9/fn3r16jFu3DjTHyfZs2fn1q1bZtveunULFxcXXF1d430+BwcH8uXLl6DMEPPoGoA8efIk6ZEXly9fxtPTM0GNuzclprwGg+FFs2ldKwj+/w1E98LYNVtO/vz5E+31j8ybO3du7FPZs3vyfu5fe0gGj3TU6lclUbLF9F7NkSMHQUFB5MiRI8Z9bf1ejinby+eNyzZJTUr5PUuoCxcuWO1YIm8LHx8ffHx8ol138+ZNChQoYLYsS5YsANy4cSPeDSdrjf6O9OqoydhGhb+8z5v6f3pc8ryOtfNa8hol9ihaa7x+sUmM1zYhktN7N7Fe29DQ0ER/31pKo79FEi5ODacPPviAcePG8e6777Jx40aqV69OxowZrRJgxYoVfPPNN7Rp04bBgweb/fKVKVOGw4cPm21/8OBBSpUqlaB5EQwGAy4uLvHePy6Swx+Yzs7ONn8drCnGvMGn4cbRKNsmttSpUwNw/9pDgi/eNS1PCtkiRWaM/B6dxMobl/MmpdcyJinm9yyeVFBJSmXL2ig20U03EPn/8MjLpOPDWqO/XxU5ajK2UeGR3uSo1bjkeR1r543Pa5RYo2it8frFJjFe24RITu/dxHptXz6vRn9r9Le8PeLUcLpy5Qp37774g3nQoEGsWrXKKkXVpUuXGD16NO+99x5dunQhODjYtM7JyYk2bdrQpEkTJk6cSJMmTdi3bx/btm0jICAgwecWERERiS9b1Uav4+TkFGW6gchGU0KaxdYa/R3p1VGTcWk+v8lRq9Zohls7ryWvUWKPorX1hwmJ8domRHJ67ybWa5snTx5CQkI0+huN/pa3S5waTl5eXnz55ZcUKFAAo9HIiBEjSJs2bbTbGgwGFi9eHKeTb9++nbCwMHbu3MnOnTvN1jVp0oSxY8fy7bffMmHCBBYvXkyuXLmYMGECFStWjNPxRURERGzBVrXR62TLli3a6QYAsmbNGu/j2mr0tyWjJpPDH6AvS4y8r54zuY2ijSu9F2wnMUeuRza6ktv7VqO/ReIvTg2n8ePHM3PmTO7fv2/6BYmpM25Jx7xr16507do11m2qVatGtWrV4nxMEREREVuzVW30OmXLlmXlypWEh4djb//iphQHDx4kT548uLm5We08IiIiIgkVp4ZTgQIFmD59OgCFChVixIgRFCtWzKbBRESSs8hJNPUplkjKlFi1UbNmzQgICGDw4MF06tSJ48ePs2jRIvz8/Gx+bhERERFLWHyXujNnzph+DgoK4tGjR2TMmJF33nnHqsGSCmN4OAZ7+xgfy9tD7wWJqwhjRJRJNCOMEdgZ4n+zAxFJut5kbeTm5kZAQAD+/v40adKEzJkzM3DgQJo0aWL1c4mIiIgkhMUNJ4CtW7cybtw4s0m+3d3d6d+/P40bN7ZWtiTBYG/P3Z69eH7+Aqny5yPTzBmJHUkSid4LEld2BjsmBU7g2qNrAHi4etC/zIBETiUitmSr2mjs2LFRlhUrVoxVq1bF+5iS/GkUrYjt6fdMJOEsbjjt2bOHAQMGUKFCBfr164e7uzu3bt1i8+bNDBo0iAwZMlCjRg0bRE08z89fIOzkycSOIUmA3gsSV9ceXePig6DEjiEib8DbWBvJm2WXOTPhEUbs7V784fvqKNqX14lI/On3TMS6LG44fffdd9SrV48pU6aYLW/WrBl9+/Zlzpw5KqpERETkraHaSGzNLn067O0MDF97nMvBj83Webqnxa+55lYVsQb9nolYl8UTipw7dy7GeQKaNGliNo+BiIiISEqn2kjelMvBjzl745HZFub11QAAKLZJREFU16t/GItIwuj3TMR6LG44ZcyYkQcPHkS77v79+zg6OiY4lEhK5JzBCSLCzRe++lhERJId1UYiIiIiUVnccKpYsSIzZ87k5s2bZstv3LjBrFmzqFy5stXCiaQkqdM6gp09rGsFc0q9+G6nu9yJiCR3qo1EREREorJ4Dqd+/frRrFkz6tSpQ8mSJXF3dyc4OJijR4+SPn16+vfvb4ucIilH8Gm4cTSxU4iIiJWoNkq6jOHhGOz14Y6IiEhisLjhlDlzZjZs2MCCBQv4/fffOXnyJOnTp6dNmza0b98ed3d3W+QUERERSZJUGyVdBnt77vbsxfPzF8yWp65Zg/S+XyVOKBERkbeExQ0nADc3NwYMGGDtLCIiIiLJkmqjpOv5+QuEnTxptixVPq9ESiMiIvL2sHgOJ0m6jOFRJ6CObpmIiIiIiIiIiC3Fa4STJE2vDhtPlT8fmWbOSORUIiIiIiIiIvK2UcMphYlu2LjIWyEiPOpd/6JbJiIiIiIiIjanhpOIpAx29rCu1Yu7AAK4F4ZmyxM3k4jIW8pgMODs7IzBYEjsKCIiIpJI1HASkZQj+DTcOJrYKURE3lrhEUbs7V40m7y9vRM7joiIiCQiixtOT5484bvvvuPnn38mNDSUiIgIs/UGg4Fdu3ZZLaCIiIhIUqba6P/Y2xkYvvY4l4Mfmy2vmC8zXWvnT6RUIiIikhgsbjj5+/uzdu1aypUrR+HChbGz043uRERE5O2l2sjc5eDHnL3xyGzZO+5pEimNiIiIJBaLG047duygb9++dO7c2RZ5RERERJIV1UYiIiIiUVn8EVxYWBjFihWzRRYRERGRZEe1kYiIiEhUFjecqlSpwi+//GKLLCLJhl3mzIRHGM2WvfpYRETeDqqNRERERKKy+JK6+vXrM3z4cO7evUvx4sVxdnaOsk3jxo2tkU0kybJLn85sYlRP97T4Nden2yIibyPVRiIiIiJRWdxw+uKLLwDYuHEjGzdujLLeYDCoqJK3RnQTo4qIyNtFtZGIiIhIVBY3nHbv3m2LHCIiIiLJkmojSUyZ0joSYYzAzhDzTBmvWy8iImILFjeccubMafo5NDSUx48fkyFDBhwcHKwaTERERCQ5UG0kicnVyQE7gx2TAidw7dG1KOs9XD3oX2ZAIiQTEZG3ncUNJ4DAwEDGjx/PyZMnMRpfTJRcrFgx+vbtS4UKFawaUERERCSpU20kie3ao2tcfBCU2DFERERMLG44/fHHH7Rr1w4PDw+6d++Ou7s7t27d4ocffqBTp04sXbqUkiVL2iKriIiISJKj2khEREQkKosbTlOnTqVMmTLMnz8fe3t70/KePXvSsWNHZsyYwYIFC6waUkRErCM8woi9nSHKMoPBEMMeIvI6qo1EREREorJ49sATJ07Qtm1bs4IKwM7OjtatW3P8+HGrhRMREeuytzMwfO1xPpv9Pz6b/T+Grz0epQElIpZRbSQiIiISlcUjnNKkScPz58+jXff8+XPTvAUiIpI0XQ5+zNkbjxI7hkiKodpIREREJCqLRziVKlWKuXPnEhoaarY8JCSEuXPnUqZMGauFExEREUnqVBuJiIiIRGXxCKf+/fvTtGlTatWqRY0aNcicOTO3b99m7969PHnyBH9/f1vkFBEREUmSVBuJiIiIRGVxw+mdd95h1apVzJw5k3379vHgwQPSp09PuXLl6NmzJ/ny5bNFThEREZEkSbWRiIiISFQWN5wA8uXLx9SpU60cRUTkLRIRDnb2MT8WkWRFtZGISPJgDA/HYK+aS+RNiFPDaePGjVSvXp2MGTOycePG127fuHHjBMYSEUnh7OxhXSsIPg3uhaHZ8sROJCIWUG0kIpI8GeztuduzF8/PXzBbnrpmDdL7fpU4oURSqDg1nHx9fVm9ejUZM2bE19c31m0NBoOKqiTCLnNmwiOMZrc8D48wYjDoFugiSULwabhxNLFTiEg8qDYSEUm+np+/QNjJk2bLUuXzSqQ0IilXnBpOu3fvJnPmzKafJXmwS58OezsDw9ce53LwYzzd0+LXvNjrd9SlPiIiIrFSbSQiIiISO7u4bJQzZ04cHR0B+P3333FxcSFnzpxRvhwdHfnxxx9tGlgsdzn4MWdvPOJy8OO47RB5qc+cUi++q9kkIiJiRrWRiIiISOzi1HB62aBBg7h27Vq0606fPs306dMTHEqSgMhLfYJPJ3YSE4PBQERERGLHEBERMaPaSERERCSqOF1S17lzZ4KCggAwGo306NHD9Kney+7cuUPu3Lmtm1DkJXZ2duyevJ/71x7iUTo75VqXTOxIIiLyFlJtJCIiIhK7ODWcunbtypo1awDYsGED3t7eZMqUyWwbOzs70qVLR9OmTa2fUuQl9689JPjiXTLkSpfYUURE5C2l2khEREQkdnFqOJUqVYpSpUoBEBoaSv/+/fHw8LBpMBEREZGkSrWRiIiISOwsnsPpwIEDHDlyxBZZRERERJId1UYiIiIiUVnccHJwcCBjxoy2yCIiIiKS7Kg2EhEREYkqTpfUvaxPnz6MHz+eR48eUahQIVxcXKJskyNHjniFmTNnDvv372fp0qWmZadPn8bf35+TJ0+SKVMm2rVrR9u2beN1fBERERFrs2VtJCIiIpJcWdxwGjFiBOHh4QwYMCDGbU6fPm1xkOXLlzN16lTKlCljWnbv3j3at2+Pj48Pfn5+/Pnnn/j5+ZEmTRqaNWtm8TlERERErM1WtVFs/v33X6pVqxZl+ZgxYzRJuYiIiCQJFjecRo0aZdUA//77L8OHD+fQoUN4enqarVu9ejUODg6MHDmSVKlS4eXlxZUrV5g7d64aTiIiIpIkWLs2ioszZ86QOnVqdu3ahcFgMC13dXV941lEREREomNxw6lJkyZWDXDq1CkcHBzYvHkzs2bN4vr166Z1gYGBlCtXjlSp/i9mhQoVmDNnDsHBwbi7u1s1i4iIiIilrF0bxcW5c+fw9PQkS5Ysb/zcIiIiInFhccMJ4O7duyxYsIDDhw/z8OFDMmbMSJkyZWjXrh1ubm4WHcvHxwcfH59o1928eZMCBQqYLYssrG7cuBHvhpPRaCQkJOS12xkMBpydnaMsDw0NxWg0xrj+5W3elNiyvOrp06fAi4xxPc6bfj6vnhte5I7rc7R23ri+vqGhoRbntUVWJ0fH/9fevUdHUd9vHH92EzYJCQkJl4S7kCAQIRJKRLTckdN6aUE8pQqtFxAkEVRAQRQVFaQS7igX5aICByko4kGPoG1toRZBfhAxRSIETwQhBEwCBHLd3x82K8tuLktmM7vJ+3UOh7Pf2Z19MsxsPnz2OzOyBARU+dzyfaGyrN7+t6/OPldXjsWq8pq1raXKPxdqovznBeo6I2uj6vj2228VGxtr+HoBAACM4nHD6dSpUxoxYoTOnTun7t27Kz4+XmfOnNGaNWu0detWbd68WdHR0YaEu3z5smw2m9NYUFCQpF/+c3QtiouLq3UthZCQEMXHx7uMZ2Zm6tKlSxUuv/I5taWyLFc7efKkJOn48ePVXk9t/zzunDx5Uo0bN67Wc43OW93te+X7Vjevt7Kee3SCSjK+kyQFDeiviGlTXZ5bvi9UltXb//bV2efqyrFYVV6ztrVU+edCTV39OQ7UNbVZG5U7cuSIIiMjNXLkSGVmZqpdu3YaP3682+s6AQAAmMHjhtPcuXMVGBiojz76SG3atHGMZ2Vl6aGHHtKCBQs0Z84cQ8IFBwerqKjIaay80eTuDjDV1aBBA8XFxVX5vIq+lW/fvn2V39qXP6e2eDKDoGXLljp69Kiuu+46l9kOVf3MZrh06ZKOHz/u0R1+jM5b3e3bvn17FRQUeJTXW1lLMr5T8aFDkqTAOPffgpfvC5Vl9fa/fXX2ubpyLFaV16xtLVX+uVAT3333nWHrAnxVbdZGklRSUqJjx44pLi5O06ZNU1hYmLZv366xY8dqzZo16t27t8frrO7s78p4MtvaHbNmJ1+r+pzXn7JK/pe3Mv56nF15FkD53xaLRbYgmwKs7mfll5aVqqiwyKf2BSO2H7O/UZ943HDatWuXpk+f7lRQSVKbNm2UkpKiV1991bBwMTExys7Odhorf1yTbwotFkuNGlbV+ZDy5i+1miqfJRYSElLt7eALP0957uowK29ISIjjl1B185q5bcszVpbVzG1ZW8+pLVVl8YV9wZPPheqgoEJ9UJu1kSQFBgZqz549CggIUHBwsCSpa9euysjI0KpVq66p4VTd2d+V8WS2tTtmzU6+VvU5rz9llfwvb2X89Ti78n3LZ1OXv3bevrnKOp/l9Pw2jdpocs8nfW5fMCoPs79RX3jccCotLVVkZKTbZVFRUbpw4UKNQ5VLSkrSxo0bVVpaqoD/XY/mP//5j9q3b++V6yEAAAB4qjZro3KhoaEuYx07dtSuXbuuaX3Vnf1dmZo2mM2anXyt6nNef8oq+V/eyvjrcXblWQDls6nLX5t1PkvH8o76VN6KGJGH2d+oTzxuOHXq1Ekffvih22sEfPDBBy4X+a6J4cOH680339QzzzyjMWPGKC0tTWvXrtXMmTMNew8AAICaqM3aSJIyMjI0YsQILVu2TL169XKMHzp06JqbRjWd/W0EX5qRWh3k9R5/yir5V15fOAvA386yuJIReZj9jfrE44ZTcnKyRo8erby8PN1+++1q1qyZzpw5o+3bt2vXrl1avHixYeGaNGmiN998U7NmzdKwYcPUrFkzPfXUU6bcfhgAAMCd2qyNJCk2NlYdOnTQiy++qJkzZyoyMlKbNm3SgQMHtGXLFkPfCwAA4Fp53HC69dZbNWfOHKWmpuqf//ynY7xp06aaPXu2brvttmsO4+6CmgkJCXr33XeveZ0AAADe5M3ayB2r1arly5dr3rx5evzxx5Wfn6/4+HitWbPG8NlUAAAA18rjhpMkDR06VL///e917Ngx5eXlKSIiQh06dGB6YEXKSqWr777gbgwAAPil2q6NmjZtqldeecUr6wYAADDCNTWcJGn37t3au3evcnNz1aRJE91yyy3q2bOnkdnqDmuAtGWklPO/u7807SINX29uJgAAYChqIwAAgF943HDKy8vTuHHjdODAAQUGBqpx48bKzc3VsmXL1LdvXy1ZsoTbPLqT81/px/8zOwUAADAYtREAAIArq6cvmD17tjIzM7V06VJ9/fXX2rVrl9LS0rRo0SIdOHBACxYs8EZOAICZykorfwzUY9RGAFA1i8WikJAQv7wMS1SYTWX2skqfU9VyoD7yeIbTP/7xD02ZMkWDBw92jFmtVg0ZMkTnzp3T0qVLNXXqVENDAgBMduWpwZwWDDihNgIA96zNmqm0zK4A68/Npvj4eLMjXZNGwQ1ktVg1b99cZZ3PclneplEbTe75pAnJAN/mccPJbreradOmbpe1aNFCBQUFNQ4FADCHxWJRWVmZrFY3E2A5NRhwi9oIANyzRoQrwGrR85vTdDzngtOy3nHN9MjgjiYluzZZ57N0LO+o2TEAv+Fxw2nYsGFatmyZbrrpJoWGhjrGS0pKtG7dOg0bNszQgACA2mW1WvXZ/F3KzcqXJLX5VQvdNCrR5FSA76I2AoDKHc+5oG9/PO801q5paAXPBlBXeNxwCgkJ0fHjxzVo0CANGjRI0dHR+umnn/T555/r1KlTioiI0NNPPy3p52/KZ8+ebXho1E3l022vHvPH87wBf5ebla+cY+ckSY1bh5ucBvBt1EYAAACuPG44bdu2TWFhYZKkL774wmlZTEyM9u/f73hMowCeuHq67XVNwzTzngSTUwEAUDlqIwAAAFceN5z+9re/eSMHIMn9dFsAAHwZtREAAIArjxtO5fLz83XgwAGdP39eUVFR6tatm+PbPQAAgPqG2ggAAOAX19RwWrlypV5//XVdvnzZMWaz2TRu3DilpKQYFg51m720VJaAALNjAABQY9RGAAAAzjxuOG3ZskXz58/XPffco9/97ndq2rSpzpw5ow8++EBLly5Vy5YtuRsLqsUSEKBzj05QScZ3ChrQXxHTppodCQAAj1EbAQAAuPK44bR27Vrde++9ev755x1jHTp0UK9evRQcHKy3336bosrHWSwWhYSE+MSFS0syvlPxoUMKjIut8Dm+lLc6/CmvP2UFAF9FbQQAAODK6ukLvv/+ew0ePNjtskGDBunYsWM1DgXviAqzqcxepuDgYMXHxyskJERl9jKzY1Xo6rzBwcFmR6pQeVbp59tj+0tef9i2AODrqI0AAABceTzDKTo6WidPnnS77IcffuDimD6sUXADWS1Wzds3V1nns9SmURtN7vmk2bEqdHXeX0X/Sn+Kv9/sWG5dnVWSX+X15awA4OuojQAAAFx5PMNp4MCBWrRokdLS0pzGDx48qCVLlmjgwIGGhYN3ZJ3P0rG8o47GiK8rz3v64mmzo1SpPKu/5fWHrADgq6iNAAAAXHk8w2nChAn697//rREjRqhVq1Zq2rSpcnJydOLECcXGxmry5MneyAkAAOCTqI0AAABcedxwCgsL0+bNm7Vlyxbt3btXeXl56tatmx566CHdfffdXAsGAADUK9RGAAAArjxuOI0ePVpjxozRfffdp/vuu88bmQAAAPwGtREAAIArj6/htH//fm6hDgAA8D/URgAAAK48bjj16dNH27ZtU3FxsTfy+DRrs2YqLbObHQMAqs1eWmp2BKDOq8+1EQAAQEU8PqUuKChI27Zt08cff6zY2Fg1bNjQabnFYtFbb71lWEBfYo0IV4DVouc3p+l4zgX1jmumRwZ3NDsWAFTIEhCgc49OUEnGdwoa0F8R06aaHQmoc+pzbQQAAFARjxtOp06dUmJiouOx3e484+fqx3XR8ZwL+vbH82rXNNTsKABQpZKM71R86JAC42LNjgLUSdRGAAAArjxuOL3zzjveyAEAAOCXqI0AAABcedRwSktL04kTJ9SuXTvFx8d7KxMAAIBfqG+1kb20VJaAALNjAHUexxqAuqBaDaf8/HyNGzdOBw4ckN1ul8ViUWJioubNm6cWLVp4OyMAAIBPqa+10ZXXhbsa14kDjFPRscZxBsCfVKvhtHDhQqWnp2vChAnq2rWrjh07puXLl+u5557TG2+84e2MfqVxUKTKyspktXp8A8Ba509ZAQDwJfW5Niq/LtzVuE4cYCx3xxrHGQB/Uq2G09///ndNmjRJ999/vySpb9++io6O1pQpU1RQUOByN5b6LKxBqKxWqz6bv0u5Wflq86sWumlUYtUvNMHVWSX5dF7UD9ZmzVRaZleA1WJ2FACoELURAABA5arVcDpz5oxuuOEGp7FevXqptLRUP/74o2Jj6bRfLTcrXznHzqlx63Czo1SpPKskv8iLus0aEa4Aq0XPb07T8ZwLkqTecc30yOCOJicDgF9QGwEAAFSuWg2nkpIS2Ww2p7GIiAhJUmFhofGpAJgupHGwVFYqWa+4YOXVj73oeM4FffvjeUlSu6ahtfKe1XX1hTy5sCdQ/1AbAQAAVM6ju9S5Y7fbjcgBwMcEhdl+bi5tGSnl/Fdq2kUavt7sWD7hygt5BnaMU9TSJWZHAuBDqI0AAAAMaDhZLFxnBajTcv4r/fh/ZqfwORVdNBcAqI0AAAA8aDi98MILCgsLczwu//ZuxowZCg395XQXi8Wit956y8CIAAAAvofaCAAAoGLVajglJSVJcp0i7m6caeQAAKCuozYCAACoXLUaTu+88463cwCAX7I2a6bSMrsCrJxCA9Qn1EYAAACVq/E1nACgPrNGhCvAatHzm9N0POeCesc10yODO5odCwAAAABMZTU7AADUBcdzLujbH8/rZG6B2VEAAAAAwHQ0nAAAAAAAAGAoGk4AAAAAAAAwFA0nAAAAAAAAGIqGEwAAAAAAAAxFwwkAAAAAAACGouEEAAAAAAAAQ/lFw6msrEyLFy9Wnz591L17dz388MPKysoyOxYAAIApqI0AAICv84uG0+uvv64NGzbopZde0saNG1VWVqYxY8aoqKjI7GgAAAC1jtoIAAD4Op9vOBUVFWn16tWaOHGi+vfvr86dO2vBggU6deqUduzYYXY8AACAWkVtBAAA/IHPN5wOHz6sixcvqnfv3o6x8PBwxcfHa+/evSYmAwAAqH3URgAAwB9Y7Ha73ewQldmxY4cmTJiggwcPKjg42DH+2GOP6fLly1qxYoVH69u/f7/sdrsaNGhQredbLBaVnT0re3GxLMHBsjZurJ8uFqmk1K6gBlaFhzRQXmGeSuwlkqSggCCFNQjTpbzLKispU2BQoILCbNLFbKms+OeVWhtIoc1l9Ka/MqukKvNenVWSa14vZb0679VZJVWZtza3bVV5r2lfqKWsUt3bFwItgYoIinDKYrFYfHZfqDKvids2MMCiyFCb7Ha7SkpKFBgYKIvFYti+UFxcLIvFoh49ehj+swH1lZm10dW/X5yWufmMKefus7Gcu89Io1SUt7Ks/pa3sqzezMu+4Ht5fS2rv+WtreOM2gj1SaDZAapy6dIlSZLNZnMaDwoKUl5ensfrs1gsTn9Xh7VJE6fHkaHOWSKCIlxeExIR7DwQ2rzCLEa6OqtUdV6XrJJLXm9klaretlI18tbStpW8sy/UVlap7u0LLll8fF+oKq+Z29ZisTj9Z9OofcFisXjt5wLqK7NrI3e/X67k7jOmnLvPxqtzGK2yvJVllfwrb2VZpdqrO6/EvlAz3tgXzMgq+Vdebx9n1EaoT3y+4VT+zV1RUZHTt3iFhYUKCQnxeH2JiYmGZQMAAKht1EYAAMAf+Pw1nFq0aCFJys7OdhrPzs5WdHS0GZEAAABMQ20EAAD8gc83nDp37qywsDDt2bPHMZafn6/09HQlJSWZmAwAAKD2URsBAAB/4POn1NlsNo0aNUqpqamKiopSq1atNHfuXMXExGjIkCFmxwMAAKhV1EYAAMAf+HzDSZImTpyokpISPfvss7p8+bKSkpK0atWqat9pDgAAoC6hNgIAAL7OYvfGvSYBAAAAAABQb/n8NZwAAAAAAADgX2g4AQAAAAAAwFA0nAAAAAAAAGAoGk4AAAAAAAAwFA0nAAAAAAAAGIqGEwAAAAAAAAxFw6mGysrKtHjxYvXp00fdu3fXww8/rKysLLNjVcuKFSv0pz/9yewYFcrNzdVzzz2nvn37qkePHrr33nu1b98+s2NV6OzZs3ryySd18803KzExUWPHjtXRo0fNjlWlzMxMJSYm6r333jM7SoVOnz6tTp06ufzx5cxbt27V7bffrm7duumOO+7Qxx9/bHYkF3v27HG7XTt16qRBgwaZHQ+An/LX2sjX6yKJ2qi2UBsZzx/qIonaCDBaoNkB/N3rr7+uDRs2aM6cOYqJidHcuXM1ZswYffjhh7LZbGbHq9D69eu1cOFC9ezZ0+woFZo0aZLOnDmj+fPnq0mTJnrnnXc0evRovf/+++rQoYPZ8VykpKSorKxMK1euVGhoqBYtWqQHHnhAO3bsUEhIiNnx3CouLtaUKVNUUFBgdpRKHT58WEFBQfr0009lsVgc440aNTIxVcU++OADPfPMM5o+fbr69Omj7du3a9KkSYqJiVFiYqLZ8RwSExO1a9cup7EDBw5owoQJSk5ONikVAH/nj7WRP9RFErVRbaA2Mp6/1EUStRFgNGY41UBRUZFWr16tiRMnqn///urcubMWLFigU6dOaceOHWbHc+v06dN65JFHlJqaquuuu87sOBX6/vvvtXv3br3wwgvq2bOn2rdvrxkzZqh58+b68MMPzY7nIi8vT61atdLLL7+shIQExcbGKjk5WdnZ2crIyDA7XoWWLFmisLAws2NU6ciRI7ruuuvUvHlzNWvWzPEnODjY7Ggu7Ha7Fi1apD//+c8aOXKk2rZtq/Hjx+uWW27Rl19+aXY8JzabzWl7hoaG6pVXXtGwYcM0fPhws+MB8EP+Vhv5S10kURvVFmojY/lTXSRRGwFGo+FUA4cPH9bFixfVu3dvx1h4eLji4+O1d+9eE5NV7JtvvlGDBg20bds23XjjjWbHqVBkZKRWrlypbt26OcYsFossFovy8/NNTOZeRESE5s2bp+uvv16SdO7cOa1du1YxMTGKi4szOZ17e/fu1bvvvqs5c+aYHaVK3377rWJjY82OUS2ZmZk6ceKE7rrrLqfxVatWady4cSalqp7ly5fr0qVLmjp1qtlRAPgpf6uN/KUukqiNagO1kfH8uS6SqI2AmuKUuho4deqUJKlFixZO482bN3cs8zUDBw7UwIEDzY5RpfDwcPXr189p7JNPPtH333+v6dOnm5SqembMmKFNmzbJZrNp2bJlatiwodmRXOTn5+upp57Ss88+67L/+qIjR44oMjJSI0eOVGZmptq1a6fx48erb9++ZkdzkZmZKUkqKCjQ6NGjlZ6ertatW2v8+PE+feyV/0dg8uTJaty4sdlxAPgpf6uN/KUukqiNvI3ayDv8tS6SqI0AIzDDqQYuXbokSS7XIwgKClJhYaEZkeqs/fv36+mnn9aQIUPUv39/s+NU6v7779eWLVt05513KiUlRd98843ZkVy88MILSkxMdPm2yReVlJTo2LFjysvL04QJE7Ry5Up1795dY8eO1RdffGF2PBcXLlyQJE2dOlV33nmnVq9erVtvvVXJyck+mbfchg0b1KhRI40YMcLsKAD8GLVR7aE2Mha1kXf4a10kURsBRmCGUw2UnyNdVFTkdL50YWGhz14I0R99+umnmjJlinr06KHU1FSz41SpfJr4rFmzdPDgQa1bt06vvPKKyal+sXXrVu3bt88nr/fgTmBgoPbs2aOAgADHcda1a1dlZGRo1apVTqdt+IIGDRpIkkaPHq1hw4ZJkrp06aL09HStWbPG5/KW27p1q4YOHepz134A4F+ojWoHtZGxqI28x1/rIonaCDACM5xqoHy6bXZ2ttN4dna2oqOjzYhU56xbt04TJkzQgAEDtHz5cgUFBZkdya1z585p+/btKikpcYxZrVbFxcW57B9m27Jli86ePav+/fsrMTHRcXeQ559/XmPGjDE5nXuhoaEuv+w7duyo06dPm5SoYuXHfvk1K8rFxcXphx9+MCNSlQ4fPqysrCy/+FYXgG+jNvI+aiPjURt5jz/WRRK1EWAUGk410LlzZ4WFhWnPnj2Osfz8fKWnpyspKcnEZHXDhg0b9NJLL2nkyJGaP3++z95KWZJycnI0adIkp6nBxcXFSk9P97kLOqampuqjjz7S1q1bHX8kaeLEiZo1a5a54dzIyMhQjx49nI4zSTp06JBPXnT0hhtuUGhoqA4ePOg0fuTIEbVt29akVJXbt2+fmjRpos6dO5sdBYCfozbyLmoj76A28h5/rIskaiPAKJxSVwM2m02jRo1SamqqoqKi1KpVK82dO1cxMTEaMmSI2fH8WmZmpmbPnq3bbrtN48aNU05OjmNZcHCwGjVqZGI6V9dff7369u2rl19+WS+//LIiIiK0YsUK5efn64EHHjA7npOKvmFu0qSJT377HBsbqw4dOujFF1/UzJkzFRkZqU2bNunAgQPasmWL2fFcBAcHa8yYMXrttdcUHR2thIQEbd++Xbt379batWvNjudWenq6OnXqZHYMAHUAtZH3UBt5D7WR9/hjXSRRGwFGoeFUQxMnTlRJSYmeffZZXb58WUlJSVq1apXjfGVcm08++UTFxcXauXOndu7c6bRs2LBhPnm72vnz52vevHl64okndP78efXs2VPr169Xy5YtzY7m16xWq5YvX6558+bp8ccfV35+vuLj47VmzRqX6dm+Ijk5WSEhIVqwYIFOnz6t2NhYLVmyRL169TI7mltnzpzh7isADENt5B3URijnb7WRv9VFErURYBSL3W63mx0CAAAAAAAAdQfXcAIAAAAAAIChaDgBAAAAAADAUDScAAAAAAAAYCgaTgAAAAAAADAUDScAAAAAAAAYioYTAAAAAAAADEXDCQAAAAAAAIai4QTUIXa73ewIAAAAPoPaCADMQ8MJMNjXX3+tJ598Uv3791dCQoIGDx6sGTNmKCsry/GcTp06acmSJYa+71dffaWxY8cauk4AAICaojYCgPqJhhNgoPXr1+uPf/yjzp49q8mTJ+uNN97Q2LFj9eWXX+qee+7R4cOHvfbef/3rX3X06FGvrR8AAMBT1EYAUH8Fmh0AqCu++uorzZo1SyNHjtQzzzzjGO/Vq5cGDx6soUOHavr06XrvvfdMTAkAAFA7qI0AoH5jhhNgkFWrVqlRo0aaNGmSy7KoqChNmzZNgwYNUkFBgdOy9957T506ddIPP/zgND5w4EBNmzbN8Xj37t36wx/+oMTERCUlJWn8+PGOb+2mTZum999/XydOnFCnTp0chVthYaFeffVV9evXT127dtVdd92ljz76yOV9Zs+erfvvv18JCQlOBSEAAMC1ojYCgPqNGU6AAex2u3bt2qWBAwcqJCTE7XNuv/32a15/VlaWkpOTNXz4cE2aNEn5+fmaP3++xo4dq507dyo5OVnnzp1Tenq6li5dqrZt28putyslJUX79+/XxIkTFRsbq507d+qJJ55QUVGRhg4d6lj/+vXr9eCDD+rhhx9WaGjoNecEAACQqI0AADScAEP89NNPKiwsVOvWrb2y/rS0NF2+fFnjxo1TdHS0JCkmJkafffaZCgoK1LZtW0VFRclms6l79+6Sfv7W71//+pcWLFjgKOj69OmjS5cuKTU1VXfeeacCA3/+CGjZsqWmTJnilewAAKD+oTYCANBwAgwQEBAgSSotLfXK+m+88UYFBQXpnnvu0W9+8xv17dtXvXr1UkJCQoWv+eKLL2SxWNSvXz+VlJQ4xgcOHKht27YpIyNDXbp0kSTH3wAAAEagNgIA0HACDBAREaHQ0FCdPHmywucUFBSouLhYERERHq+/devWWrdunVauXKnNmzfr7bffVnh4uO677z49/vjjslgsLq/Jzc2V3W5Xjx493K4zOzvbUUw1bNjQ40wAAAAVoTYCANBwAgzy61//Wnv27FFhYaGCgoJclm/atEl/+ctftHnzZqfx8oKorKzMafzixYtOjxMSErR06VIVFRXpq6++0rvvvqvly5erc+fO+u1vf+vyfo0aNVLDhg319ttvu83brl07j34+AAAAT1AbAUD9xl3qAIM89NBDys3N1cKFC12WnTlzRqtXr1ZcXJxuuOEGp2VhYWGSpFOnTjnGjh49qtzcXMfjtWvXasCAASoqKpLNZlPv3r310ksvSZLjm0Or1flwvummm1RQUCC73a5u3bo5/hw5ckSvvfaa01RyAAAAo1EbAUD9xgwnwCDdu3fXY489poULF+ro0aMaOnSoIiMjlZGRoVWrVqmwsNBtwdWrVy8FBwdrzpw5euyxx3Tx4kUtXrxYjRs3djzn5ptvVmpqqlJSUjRq1CgFBARo48aNstlsGjBggCQpPDxcOTk5+vzzz9WlSxf169dPSUlJSk5OVnJysmJjY5WWlqbFixerT58+ioqKqqUtAwAA6iNqIwCo35jhBBho/PjxWrlypSRp9uzZGjt2rNatW6f+/ftr69atio2NdXlNeHi4lixZotLSUqWkpGjRokVKSUlR165dHc/p3Lmzli9frgsXLmjSpEl69NFHlZubq9WrV6tDhw6SpLvvvlutWrVSSkqKtm7dKqvVqpUrV+qOO+7QihUrNHr0aG3cuFEPPvigFixYUDsbBAAA1GvURgBQf1nsdrvd7BAAAAAAAACoO5jhBAAAAAAAAEPRcAIAAAAAAIChaDgBAAAAAADAUDScAAAAAAAAYCgaTgAAAAAAADAUDScAAAAAAAAYioYTAAAAAAAADEXDCQAAAAAAAIai4QQAAAAAAABD0XACAAAAAACAoWg4AQAAAAAAwFA0nAAAAAAAAGCo/wcR1+kSkErHGwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1200x800 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set the seaborn style and color palette\n",
    "sns.set_style(\"whitegrid\")\n",
    "sns.set_palette(\"Set1\")\n",
    "\n",
    "# Set up the figure and axes for subplots\n",
    "fig, axes = plt.subplots(2, 2, figsize=(12, 8))\n",
    "\n",
    "# Define the variables to be plotted\n",
    "variables = ['Age', 'Sex', 'Final_Category', 'Distress']\n",
    "\n",
    "# Define custom titles for each graph\n",
    "titles = ['Age', 'Sex', 'Behaviour', 'Distress']\n",
    "\n",
    "# Iterate over each variable and plot a bar chart in the corresponding subplot\n",
    "for i, (variable, title) in enumerate(zip(variables, titles)):\n",
    "    df_filtered = cluster_df.copy()  # Make a copy of the DataFrame\n",
    "    \n",
    "    # Exclude 'Unknown' values for the 'Distress' variable\n",
    "    if variable == 'Distress':\n",
    "        df_filtered = df_filtered[df_filtered[variable] != 'Unknown']\n",
    "    \n",
    "    # Exclude 'Anti-Predator' and 'Sexual' levels for the 'Final_Category' variable\n",
    "    if variable == 'Final_Category':\n",
    "        df_filtered = df_filtered[(df_filtered[variable] != 'Anti-Predator') & (df_filtered[variable] != 'Sexual') & (df_filtered[variable] != 'Unknown')]\n",
    "    \n",
    "    # Exclude 'Unknown' values for other variables\n",
    "    if variable in ['Age', 'Sex']:\n",
    "        df_filtered = df_filtered[df_filtered[variable] != 'Unknown']\n",
    "    \n",
    "    # Calculate counts for each category in the variable, grouped by Cluster\n",
    "    counts = df_filtered.groupby(['Cluster', variable]).size().unstack().fillna(0)\n",
    "    \n",
    "    # Calculate proportions\n",
    "    proportions = counts.div(counts.sum(axis=0), axis=1) * 100\n",
    "    \n",
    "    # Plot the bar chart in the i-th subplot\n",
    "    ax = proportions.plot(kind='bar', stacked=False, ax=axes[i//2, i%2])\n",
    "    ax.set_title(title)\n",
    "    ax.set_xlabel('Cluster')\n",
    "    ax.set_ylabel('Proportion of category per cluster (%)')\n",
    "    \n",
    "    # Remove 'Unknown' and 'Anti-Predator' from the legend\n",
    "    handles, labels = ax.get_legend_handles_labels()\n",
    "    filtered_handles_labels = [(h, l) for h, l in zip(handles, labels) if l not in ['Unknown', 'Anti-Predator', 'Sexual']]\n",
    "    if filtered_handles_labels:\n",
    "        handles, labels = zip(*filtered_handles_labels)\n",
    "        ax.legend(handles, labels, bbox_to_anchor=(1.05, 1), loc='upper left', title=\"Category\")\n",
    "    else:\n",
    "        ax.legend().remove()\n",
    "    \n",
    "    # Rotate cluster labels to be horizontal\n",
    "    ax.set_xticklabels(ax.get_xticklabels(), rotation=0)\n",
    "\n",
    "# Adjust layout\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6fb5556",
   "metadata": {},
   "source": [
    "## 5. Behavioural and Demographic Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "674b9f23",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x1400 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise the associations between the variables using boxplots\n",
    "create_boxplots(cluster_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "d3686264",
   "metadata": {
    "metadata": {}
   },
   "outputs": [],
   "source": [
    "# Create new dataframe, remove unknowns and convert categorical variables to category type\n",
    "filtered_df = cluster_df.copy()\n",
    "filtered_df = filtered_df[(filtered_df['Final_Category'] != 'Unknown') & (filtered_df['Age'] != 'Unknown') &(filtered_df['Sex']!= 'Unknown')]\n",
    "cat_vars = ['Age', 'Sex', 'Final_Category', 'Distress']\n",
    "filtered_df[cat_vars] = filtered_df[cat_vars].astype('category')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23775191",
   "metadata": {},
   "source": [
    "Specify and run GLM to test statistical significance of the variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "6ddd241a",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                 Generalized Linear Model Regression Results                  \n",
      "==============================================================================\n",
      "Dep. Variable:                  UMAP1   No. Observations:                  284\n",
      "Model:                            GLM   Df Residuals:                      271\n",
      "Model Family:                Gaussian   Df Model:                           12\n",
      "Link Function:               Identity   Scale:                          24.771\n",
      "Method:                          IRLS   Log-Likelihood:                -852.10\n",
      "Date:                Mon, 16 Sep 2024   Deviance:                       6713.1\n",
      "Time:                        15:17:12   Pearson chi2:                 6.71e+03\n",
      "No. Iterations:                     3   Pseudo R-squ. (CS):             0.3741\n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================================================\n",
      "                                                                  coef    std err          z      P>|z|      [0.025      0.975]\n",
      "-------------------------------------------------------------------------------------------------------------------------------\n",
      "Intercept                                                      14.7483      0.836     17.633      0.000      13.109      16.388\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Affiliation]       0.3900      1.109      0.352      0.725      -1.784       2.564\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Anti-Predator]     4.5185      4.187      1.079      0.280      -3.687      12.724\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Competition]      -3.5396      0.803     -4.409      0.000      -5.113      -1.966\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Logistics]         0.2991      1.408      0.212      0.832      -2.461       3.059\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Nursing]          -4.8791      1.482     -3.293      0.001      -7.783      -1.975\n",
      "C(Final_Category, Treatment(\"Separation\"))[T.Sexual]           -1.0306      1.983     -0.520      0.603      -4.917       2.856\n",
      "C(Distress, Treatment(\"No distress\"))[T.Distress]              -2.8584      1.184     -2.415      0.016      -5.178      -0.538\n",
      "C(Distress, Treatment(\"No distress\"))[T.Not Applicable]        -0.7021      1.113     -0.631      0.528      -2.884       1.480\n",
      "C(Distress, Treatment(\"No distress\"))[T.Unknown]                1.3616      0.785      1.734      0.083      -0.177       2.900\n",
      "C(Age)[T.Infant]                                               -7.1349      2.431     -2.935      0.003     -11.899      -2.370\n",
      "C(Age)[T.Juvenile]                                             -4.9996      0.997     -5.016      0.000      -6.953      -3.046\n",
      "C(Age)[T.Sub-Adult]                                            -4.7339      0.781     -6.063      0.000      -6.264      -3.204\n",
      "C(Sex)[T.Male]                                                 -0.2294      0.736     -0.312      0.755      -1.672       1.213\n",
      "===============================================================================================================================\n"
     ]
    }
   ],
   "source": [
    "def fit_glm_model(df):\n",
    "    # Specify the formula \n",
    "    formula = 'UMAP1 ~ C(Final_Category,Treatment(\"Separation\")) + C(Distress,Treatment(\"No distress\")) +C(Age) +C(Sex)'\n",
    "\n",
    "    # Specify the family and link for the GLM\n",
    "    glm_model = sm.GLM.from_formula(formula, family=sm.families.Gaussian(), data=filtered_df)\n",
    "    glm_model_result = glm_model.fit()\n",
    "\n",
    "    # Print the summary of the model\n",
    "    print(glm_model_result.summary())\n",
    "\n",
    "    return glm_model_result\n",
    "\n",
    "behaviour_context_model = fit_glm_model(filtered_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a231fee",
   "metadata": {},
   "source": [
    "Calculate model performance metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "27206b6d",
   "metadata": {
    "metadata": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 4.1926478036045225\n",
      "Root Mean Squared Error: 4.861844555663967\n",
      "R-squared Score: 0.3287912199089097\n",
      "Normalised RMSE: 0.2662276385477146\n"
     ]
    }
   ],
   "source": [
    "def analyze_model_performance(df, behaviour_model_result):\n",
    "    # Extract UMAP1 values\n",
    "    behaviour_model_performance = df[['UMAP1']].copy()\n",
    "    \n",
    "    # Calculate residuals \n",
    "    predicted_values = behaviour_model_result.fittedvalues\n",
    "    actual_values = df['UMAP1']\n",
    "    residuals = actual_values - predicted_values\n",
    "    \n",
    "    # Add residuals to the dataframe\n",
    "    behaviour_model_performance[\"residuals\"] = residuals\n",
    "    behaviour_model_performance[\"Final_Category\"] = df[\"Final_Category\"]\n",
    "    behaviour_model_performance[\"Predicted_UMAP\"] = predicted_values\n",
    "    \n",
    "    # Calculate and print evaluation metrics\n",
    "    mae = mean_absolute_error(behaviour_model_performance['UMAP1'], behaviour_model_performance['Predicted_UMAP'])\n",
    "    print(\"Mean Absolute Error:\", mae)\n",
    "\n",
    "    rmse = mean_squared_error(behaviour_model_performance['UMAP1'], behaviour_model_performance['Predicted_UMAP'], squared=False)\n",
    "    print(\"Root Mean Squared Error:\", rmse)\n",
    "\n",
    "    r2 = r2_score(behaviour_model_performance['UMAP1'], behaviour_model_performance['Predicted_UMAP'])\n",
    "    print(\"R-squared Score:\", r2)\n",
    "\n",
    "    nrmse = rmse / (max(df.UMAP1) - min(df.UMAP1))\n",
    "    print(\"Normalised RMSE:\", nrmse)\n",
    "\n",
    "analyze_model_performance(filtered_df,behaviour_context_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe10c952",
   "metadata": {},
   "source": [
    "Perform Tukey pairwise analysis of each level in each category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "1fa8ff86",
   "metadata": {
    "metadata": {},
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tukey Results for Final_Category levels ('Logistics', 'Sexual'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "  group1  group2 meandiff p-adj   lower  upper reject\n",
      "-----------------------------------------------------\n",
      "Logistics Sexual  -0.0032 0.9991 -5.8445 5.838  False\n",
      "-----------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Logistics', 'Separation'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "=========================================================\n",
      "  group1    group2   meandiff p-adj  lower  upper  reject\n",
      "---------------------------------------------------------\n",
      "Logistics Separation  -1.8574 0.194 -4.6718 0.9571  False\n",
      "---------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Logistics', 'Competition'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "==========================================================\n",
      "   group1     group2  meandiff p-adj  lower  upper  reject\n",
      "----------------------------------------------------------\n",
      "Competition Logistics   5.4542 0.0008 2.3228 8.5857   True\n",
      "----------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Logistics', 'Affiliation'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "===========================================================\n",
      "   group1     group2  meandiff p-adj   lower  upper  reject\n",
      "-----------------------------------------------------------\n",
      "Affiliation Logistics   0.9705 0.5557 -2.2988 4.2398  False\n",
      "-----------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Logistics', 'Anti-Predator'):\n",
      "     Multiple Comparison of Means - Tukey HSD, FWER=0.05     \n",
      "=============================================================\n",
      "    group1      group2  meandiff p-adj  lower   upper  reject\n",
      "-------------------------------------------------------------\n",
      "Anti-Predator Logistics  -5.6987 0.329 -17.7383 6.3409  False\n",
      "-------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Logistics', 'Nursing'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05  \n",
      "======================================================\n",
      "  group1   group2 meandiff p-adj lower   upper  reject\n",
      "------------------------------------------------------\n",
      "Logistics Nursing  -8.8851   0.0 -12.15 -5.6201   True\n",
      "------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Sexual', 'Separation'):\n",
      "  Multiple Comparison of Means - Tukey HSD, FWER=0.05  \n",
      "=======================================================\n",
      "  group1   group2 meandiff p-adj   lower  upper  reject\n",
      "-------------------------------------------------------\n",
      "Separation Sexual   1.8541 0.4131 -2.6149 6.3232  False\n",
      "-------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Sexual', 'Competition'):\n",
      "  Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "========================================================\n",
      "   group1   group2 meandiff p-adj  lower   upper  reject\n",
      "--------------------------------------------------------\n",
      "Competition Sexual    5.451 0.0304 0.5311 10.3709   True\n",
      "--------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Sexual', 'Affiliation'):\n",
      "  Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "========================================================\n",
      "   group1   group2 meandiff p-adj   lower  upper  reject\n",
      "--------------------------------------------------------\n",
      "Affiliation Sexual   0.9673 0.7064 -4.1453 6.0798  False\n",
      "--------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Sexual', 'Anti-Predator'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05     \n",
      "============================================================\n",
      "    group1    group2 meandiff p-adj   lower    upper  reject\n",
      "------------------------------------------------------------\n",
      "Anti-Predator Sexual   -5.702 0.4751 -24.6924 13.2885  False\n",
      "------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Sexual', 'Nursing'):\n",
      "Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "====================================================\n",
      " group1 group2 meandiff p-adj  lower   upper  reject\n",
      "----------------------------------------------------\n",
      "Nursing Sexual   8.8818 0.0005 4.3704 13.3932   True\n",
      "----------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Separation', 'Competition'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "==========================================================\n",
      "   group1     group2   meandiff p-adj lower  upper  reject\n",
      "----------------------------------------------------------\n",
      "Competition Separation   3.5969   0.0 1.9634 5.2303   True\n",
      "----------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Separation', 'Affiliation'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05     \n",
      "============================================================\n",
      "   group1     group2   meandiff p-adj   lower  upper  reject\n",
      "------------------------------------------------------------\n",
      "Affiliation Separation  -0.8869 0.3213 -2.6468 0.8731  False\n",
      "------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Separation', 'Anti-Predator'):\n",
      "      Multiple Comparison of Means - Tukey HSD, FWER=0.05      \n",
      "===============================================================\n",
      "    group1      group2   meandiff p-adj   lower   upper  reject\n",
      "---------------------------------------------------------------\n",
      "Anti-Predator Separation  -7.5561 0.1604 -18.1475 3.0353  False\n",
      "---------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Separation', 'Nursing'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05  \n",
      "======================================================\n",
      " group1   group2   meandiff p-adj lower  upper  reject\n",
      "------------------------------------------------------\n",
      "Nursing Separation   7.0277   0.0 4.3876 9.6679   True\n",
      "------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Competition', 'Affiliation'):\n",
      "     Multiple Comparison of Means - Tukey HSD, FWER=0.05     \n",
      "=============================================================\n",
      "   group1      group2   meandiff p-adj  lower   upper  reject\n",
      "-------------------------------------------------------------\n",
      "Affiliation Competition  -4.4837   0.0 -6.5423 -2.4252   True\n",
      "-------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Competition', 'Anti-Predator'):\n",
      "      Multiple Comparison of Means - Tukey HSD, FWER=0.05       \n",
      "================================================================\n",
      "    group1       group2   meandiff p-adj   lower   upper  reject\n",
      "----------------------------------------------------------------\n",
      "Anti-Predator Competition  -11.153 0.0569 -22.6414 0.3355  False\n",
      "----------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Competition', 'Nursing'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "=========================================================\n",
      "   group1    group2 meandiff p-adj   lower  upper  reject\n",
      "---------------------------------------------------------\n",
      "Competition Nursing  -3.4308 0.0207 -6.3246 -0.537   True\n",
      "---------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Affiliation', 'Anti-Predator'):\n",
      "      Multiple Comparison of Means - Tukey HSD, FWER=0.05       \n",
      "================================================================\n",
      "   group1       group2    meandiff p-adj   lower   upper  reject\n",
      "----------------------------------------------------------------\n",
      "Affiliation Anti-Predator   6.6692 0.2644 -5.1839 18.5223  False\n",
      "----------------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Affiliation', 'Nursing'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "==========================================================\n",
      "   group1    group2 meandiff p-adj  lower    upper  reject\n",
      "----------------------------------------------------------\n",
      "Affiliation Nursing  -7.9146   0.0 -10.9136 -4.9155   True\n",
      "----------------------------------------------------------\n",
      "Tukey Results for Final_Category levels ('Anti-Predator', 'Nursing'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "===========================================================\n",
      "    group1     group2 meandiff p-adj  lower   upper  reject\n",
      "-----------------------------------------------------------\n",
      "Anti-Predator Nursing -14.5838 0.0011 -22.38 -6.7876   True\n",
      "-----------------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 21 entries, 0 to 20\n",
      "Data columns (total 5 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   Variable                21 non-null     object \n",
      " 1   Level_1                 21 non-null     object \n",
      " 2   Level_2                 21 non-null     object \n",
      " 3   Reject_Null_Hypothesis  21 non-null     bool   \n",
      " 4   P-Value                 21 non-null     float64\n",
      "dtypes: bool(1), float64(1), object(3)\n",
      "memory usage: 825.0+ bytes\n",
      "DataFrame structure for Final_Category:\n",
      " None\n",
      "         Variable    Level_1        Level_2  Reject_Null_Hypothesis   P-Value\n",
      "0  Final_Category  Logistics         Sexual                   False  0.999089\n",
      "1  Final_Category  Logistics     Separation                   False  0.194031\n",
      "2  Final_Category  Logistics    Competition                    True  0.000844\n",
      "3  Final_Category  Logistics    Affiliation                   False  0.555726\n",
      "4  Final_Category  Logistics  Anti-Predator                   False  0.329030\n",
      "Tukey Results for Distress levels ('Not Applicable', 'No distress'):\n",
      "      Multiple Comparison of Means - Tukey HSD, FWER=0.05       \n",
      "================================================================\n",
      "   group1       group2     meandiff p-adj   lower  upper  reject\n",
      "----------------------------------------------------------------\n",
      "No distress Not Applicable   1.3668 0.1403 -0.4542 3.1877  False\n",
      "----------------------------------------------------------------\n",
      "Tukey Results for Distress levels ('Not Applicable', 'Unknown'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05     \n",
      "============================================================\n",
      "    group1      group2 meandiff p-adj   lower  upper  reject\n",
      "------------------------------------------------------------\n",
      "Not Applicable Unknown   0.1407 0.8765 -1.6434 1.9248  False\n",
      "------------------------------------------------------------\n",
      "Tukey Results for Distress levels ('Not Applicable', 'Distress'):\n",
      "    Multiple Comparison of Means - Tukey HSD, FWER=0.05    \n",
      "===========================================================\n",
      " group1      group2     meandiff p-adj lower  upper  reject\n",
      "-----------------------------------------------------------\n",
      "Distress Not Applicable   4.2232 0.002 1.5746 6.8718   True\n",
      "-----------------------------------------------------------\n",
      "Tukey Results for Distress levels ('No distress', 'Unknown'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "=========================================================\n",
      "   group1    group2 meandiff p-adj   lower  upper  reject\n",
      "---------------------------------------------------------\n",
      "No distress Unknown   1.5075 0.0963 -0.2721 3.2871  False\n",
      "---------------------------------------------------------\n",
      "Tukey Results for Distress levels ('No distress', 'Distress'):\n",
      "   Multiple Comparison of Means - Tukey HSD, FWER=0.05   \n",
      "=========================================================\n",
      " group1     group2   meandiff p-adj  lower  upper  reject\n",
      "---------------------------------------------------------\n",
      "Distress No distress   2.8564 0.0271 0.3292 5.3837   True\n",
      "---------------------------------------------------------\n",
      "Tukey Results for Distress levels ('Unknown', 'Distress'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      " group1   group2 meandiff p-adj  lower  upper  reject\n",
      "-----------------------------------------------------\n",
      "Distress Unknown   4.3639 0.0005 1.9582 6.7696   True\n",
      "-----------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 6 entries, 0 to 5\n",
      "Data columns (total 5 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   Variable                6 non-null      object \n",
      " 1   Level_1                 6 non-null      object \n",
      " 2   Level_2                 6 non-null      object \n",
      " 3   Reject_Null_Hypothesis  6 non-null      bool   \n",
      " 4   P-Value                 6 non-null      float64\n",
      "dtypes: bool(1), float64(1), object(3)\n",
      "memory usage: 330.0+ bytes\n",
      "DataFrame structure for Distress:\n",
      " None\n",
      "   Variable         Level_1      Level_2  Reject_Null_Hypothesis   P-Value\n",
      "0  Distress  Not Applicable  No distress                   False  0.140313\n",
      "1  Distress  Not Applicable      Unknown                   False  0.876504\n",
      "2  Distress  Not Applicable     Distress                    True  0.002018\n",
      "3  Distress     No distress      Unknown                   False  0.096299\n",
      "4  Distress     No distress     Distress                    True  0.027126\n",
      "Tukey Results for Age levels ('Adult', 'Sub-Adult'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "group1   group2  meandiff p-adj lower   upper  reject\n",
      "-----------------------------------------------------\n",
      " Adult Sub-Adult  -4.8731   0.0 -6.393 -3.3532   True\n",
      "-----------------------------------------------------\n",
      "Tukey Results for Age levels ('Adult', 'Juvenile'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "group1  group2  meandiff p-adj  lower   upper  reject\n",
      "-----------------------------------------------------\n",
      " Adult Juvenile  -6.8215   0.0 -8.4096 -5.2334   True\n",
      "-----------------------------------------------------\n",
      "Tukey Results for Age levels ('Adult', 'Infant'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "group1 group2 meandiff p-adj   lower    upper  reject\n",
      "-----------------------------------------------------\n",
      " Adult Infant  -9.4911 0.0001 -14.0634 -4.9188   True\n",
      "-----------------------------------------------------\n",
      "Tukey Results for Age levels ('Sub-Adult', 'Juvenile'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05  \n",
      "======================================================\n",
      " group1    group2  meandiff p-adj  lower upper  reject\n",
      "------------------------------------------------------\n",
      "Juvenile Sub-Adult   1.9484 0.0159 0.368 3.5289   True\n",
      "------------------------------------------------------\n",
      "Tukey Results for Age levels ('Sub-Adult', 'Infant'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "group1   group2  meandiff p-adj   lower  upper reject\n",
      "-----------------------------------------------------\n",
      "Infant Sub-Adult    4.618 0.0731 -0.4389 9.675  False\n",
      "-----------------------------------------------------\n",
      "Tukey Results for Age levels ('Juvenile', 'Infant'):\n",
      " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
      "=====================================================\n",
      "group1  group2  meandiff p-adj   lower  upper  reject\n",
      "-----------------------------------------------------\n",
      "Infant Juvenile   2.6696 0.2393 -1.8133 7.1526  False\n",
      "-----------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 6 entries, 0 to 5\n",
      "Data columns (total 5 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   Variable                6 non-null      object \n",
      " 1   Level_1                 6 non-null      object \n",
      " 2   Level_2                 6 non-null      object \n",
      " 3   Reject_Null_Hypothesis  6 non-null      bool   \n",
      " 4   P-Value                 6 non-null      float64\n",
      "dtypes: bool(1), float64(1), object(3)\n",
      "memory usage: 330.0+ bytes\n",
      "DataFrame structure for Age:\n",
      " None\n",
      "  Variable    Level_1    Level_2  Reject_Null_Hypothesis       P-Value\n",
      "0      Age      Adult  Sub-Adult                    True  1.602981e-09\n",
      "1      Age      Adult   Juvenile                    True  0.000000e+00\n",
      "2      Age      Adult     Infant                    True  8.430577e-05\n",
      "3      Age  Sub-Adult   Juvenile                    True  1.594777e-02\n",
      "4      Age  Sub-Adult     Infant                   False  7.310966e-02\n",
      "Tukey Results for Sex levels ('Female', 'Male'):\n",
      "Multiple Comparison of Means - Tukey HSD, FWER=0.05\n",
      "===================================================\n",
      "group1 group2 meandiff p-adj  lower   upper  reject\n",
      "---------------------------------------------------\n",
      "Female   Male  -3.3863   0.0 -4.7911 -1.9816   True\n",
      "---------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1 entries, 0 to 0\n",
      "Data columns (total 5 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   Variable                1 non-null      object \n",
      " 1   Level_1                 1 non-null      object \n",
      " 2   Level_2                 1 non-null      object \n",
      " 3   Reject_Null_Hypothesis  1 non-null      bool   \n",
      " 4   P-Value                 1 non-null      float64\n",
      "dtypes: bool(1), float64(1), object(3)\n",
      "memory usage: 165.0+ bytes\n",
      "DataFrame structure for Sex:\n",
      " None\n",
      "  Variable Level_1 Level_2  Reject_Null_Hypothesis   P-Value\n",
      "0      Sex  Female    Male                    True  0.000003\n"
     ]
    }
   ],
   "source": [
    "def run_pairwise_tukey(df, var):\n",
    "    levels = df[var].unique()\n",
    "    combinations = list(itertools.combinations(levels, 2))\n",
    "    tukey_results_all = []\n",
    "\n",
    "    for combo in combinations:\n",
    "        subset_df = df[df[var].isin(combo)]\n",
    "        tukey_results_var = pairwise_tukeyhsd(subset_df['UMAP1'], subset_df[var])\n",
    "\n",
    "        print(f\"Tukey Results for {var} levels {combo}:\")\n",
    "        print(tukey_results_var)\n",
    "\n",
    "        # Extract the relevant data\n",
    "        for i in range(len(tukey_results_var.reject)):\n",
    "            tukey_results_all.append({\n",
    "                'Variable': var,\n",
    "                'Level_1': combo[0],\n",
    "                'Level_2': combo[1],\n",
    "                'Reject_Null_Hypothesis': bool(tukey_results_var.reject[i]), \n",
    "                'P-Value': float(tukey_results_var.pvalues[i])\n",
    "            })\n",
    "\n",
    "    # Convert to DataFrame \n",
    "    mixed_tukey_results_df = pd.DataFrame(tukey_results_all)\n",
    "    print(f\"DataFrame structure for {var}:\\n\", mixed_tukey_results_df.info())\n",
    "    print(mixed_tukey_results_df.head())  \n",
    "\n",
    "# Run pairwise Tukey tests for each variable\n",
    "for fixed_effect in ['Final_Category', 'Distress', 'Age', 'Sex']:\n",
    "    run_pairwise_tukey(filtered_df, fixed_effect)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}