research_amp/cc/notebooks/Master_crypto_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "42069126",
   "metadata": {},
   "source": [
    "# Description"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fe5775a",
   "metadata": {},
   "source": [
    "This notebook performs EDA on the crypto prices and returns."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e2e6718",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "dbd664a7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-26T20:58:36.940259Z",
     "start_time": "2022-01-26T20:58:36.937593Z"
    }
   },
   "outputs": [],
   "source": [
    "# %load_ext autoreload\n",
    "# %autoreload 2\n",
    "# %matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3c316ebd",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-27T16:49:22.929766Z",
     "start_time": "2022-01-27T16:49:20.061014Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/core/explore.py:39: TqdmExperimentalWarning: Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)\n",
      "  import tqdm.autonotebook as tauton\n"
     ]
    }
   ],
   "source": [
    "# TODO(Grisha): move to `core/dataflow_model/notebooks` in #205.\n",
    "\n",
    "import logging\n",
    "import os\n",
    "\n",
    "import pandas as pd\n",
    "import pytz\n",
    "\n",
    "import core.config.config_ as cconconf\n",
    "import core.explore as coexplor\n",
    "import core.plotting as coplotti\n",
    "import helpers.hdatetime as hdateti\n",
    "import helpers.hdbg as hdbg\n",
    "import helpers.henv as henv\n",
    "import helpers.hprint as hprint\n",
    "import helpers.hs3 as hs3\n",
    "import im_v2.ccxt.data.client as icdcl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "284a5350",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-27T16:49:24.163650Z",
     "start_time": "2022-01-27T16:49:22.931477Z"
    },
    "lines_to_next_cell": 1
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[0m\u001b[36mINFO\u001b[0m: > cmd='/venv/lib/python3.8/site-packages/ipykernel_launcher.py -f /home/.local/share/jupyter/runtime/kernel-7cd8282d-36f7-45bd-902e-e92ea3f669ee.json'\n",
      "WARNING: Running in Jupyter\n",
      ">>ENV<<: is_inside_container=True: code_version=1.0.6, container_version=1.0.6, is_inside_docker=True, is_inside_ci=False, CI_defined=True, CI=''\n",
      ">>ENV<<: AM_AWS_PROFILE=True AM_ECR_BASE_PATH=True AM_S3_BUCKET=True AM_TELEGRAM_TOKEN=True AWS_ACCESS_KEY_ID=False AWS_DEFAULT_REGION=False AWS_SECRET_ACCESS_KEY=False GH_ACTION_ACCESS_TOKEN=True\n",
      "INFO  # Git\n",
      "    branch_name='CMTask919_Refactor_CddClient_Fix_notebooks'\n",
      "    hash='caf6c7dc5'\n",
      "    # Last commits:\n",
      "      * caf6c7dc5 max-rsrch Fix dassert                                                       ( 5 minutes ago) Thu Jan 27 16:44:13 2022  (HEAD -> CMTask919_Refactor_CddClient_Fix_notebooks, origin/CMTask919_Refactor_CddClient_Fix_notebooks)\n",
      "      * d1df00a0b max-rsrch Fix _LOG issue                                                    (   5 hours ago) Thu Jan 27 12:09:13 2022           \n",
      "      *   646a7a324 max-rsrch Checkpoint                                                        (   5 hours ago) Thu Jan 27 12:07:41 2022           \n",
      "      |\\  \n",
      "# Machine info\n",
      "    system=Linux\n",
      "    node name=0b2aa5826e44\n",
      "    release=5.11.0-1025-aws\n",
      "    version=#27~20.04.1-Ubuntu SMP Fri Jan 7 13:09:56 UTC 2022\n",
      "    machine=x86_64\n",
      "    processor=x86_64\n",
      "    cpu count=8\n",
      "    cpu freq=scpufreq(current=2499.998, min=0.0, max=0.0)\n",
      "    memory=svmem(total=33296007168, available=4242927616, percent=87.3, used=28067766272, free=968105984, active=1695342592, inactive=29197398016, buffers=428544000, cached=3831590912, shared=504725504, slab=856391680)\n",
      "    disk usage=sdiskusage(total=104021790720, used=90430623744, free=13574389760, percent=86.9)\n",
      "# Packages\n",
      "    python: 3.8.10\n",
      "    gluonnlp: ?\n",
      "    gluonts: 0.6.7\n",
      "    joblib: 1.1.0\n",
      "    mxnet: 1.9.0\n",
      "    numpy: 1.22.0\n",
      "    pandas: 1.3.5\n",
      "    pyarrow: 6.0.1\n",
      "    scipy: 1.6.1\n",
      "    seaborn: 0.11.2\n",
      "    sklearn: 1.0.2\n",
      "    statsmodels: 0.13.1\n",
      "\u001b[33mWARNING\u001b[0m: Disabling annoying warnings\n"
     ]
    }
   ],
   "source": [
    "hdbg.init_logger(verbosity=logging.INFO)\n",
    "\n",
    "_LOG = logging.getLogger(__name__)\n",
    "\n",
    "_LOG.info(\"%s\", henv.get_system_signature()[0])\n",
    "\n",
    "hprint.config_notebook()\n",
    "\n",
    "AM_AWS_PROFILE = \"am\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "808d63c0",
   "metadata": {},
   "source": [
    "# Config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "07d59b95",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-27T16:49:29.434329Z",
     "start_time": "2022-01-27T16:49:29.402716Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "load:\n",
      "  aws_profile: am\n",
      "  data_dir: s3://alphamatic-data/data\n",
      "data:\n",
      "  close_price_col_name: close\n",
      "  frequency: T\n",
      "  vendor: CCXT\n",
      "  extension: csv.gz\n",
      "  timezone: US/Eastern\n",
      "stats:\n",
      "  z_score_boundary: 3\n",
      "  z_score_window: D\n"
     ]
    }
   ],
   "source": [
    "def get_eda_config() -> cconconf.Config:\n",
    "    \"\"\"\n",
    "    Get config that controls EDA parameters.\n",
    "    \"\"\"\n",
    "    config = cconconf.Config()\n",
    "    # Load parameters.\n",
    "    config.add_subconfig(\"load\")\n",
    "    config[\"load\"][\"aws_profile\"] = AM_AWS_PROFILE\n",
    "    config[\"load\"][\"data_dir\"] = os.path.join(\n",
    "        hs3.get_s3_bucket_path(AM_AWS_PROFILE), \"data\"\n",
    "    )\n",
    "    # Data parameters.\n",
    "    config.add_subconfig(\"data\")\n",
    "    config[\"data\"][\"close_price_col_name\"] = \"close\"\n",
    "    config[\"data\"][\"frequency\"] = \"T\"\n",
    "    config[\"data\"][\"vendor\"] = \"CCXT\"\n",
    "    config[\"data\"][\"extension\"] = \"csv.gz\"\n",
    "    # TODO(Grisha): use `hdateti.get_ET_tz()` once it is fixed.\n",
    "    config[\"data\"][\"timezone\"] = pytz.timezone(\"US/Eastern\")\n",
    "    # Statistics parameters.\n",
    "    config.add_subconfig(\"stats\")\n",
    "    config[\"stats\"][\"z_score_boundary\"] = 3\n",
    "    config[\"stats\"][\"z_score_window\"] = \"D\"\n",
    "    return config\n",
    "\n",
    "\n",
    "config = get_eda_config()\n",
    "print(config)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42055791",
   "metadata": {},
   "source": [
    "# Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a2303831",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-27T16:49:48.140898Z",
     "start_time": "2022-01-27T16:49:32.421311Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  Reading data for vendor=`CCXT`, exchange id='binance', currencies='BTC_USDT' from file='s3://alphamatic-data/data/ccxt/20210924/binance/BTC_USDT.csv.gz'...\n",
      "WARNING Removed 210 / 1615122 = 0.01% rows\n",
      "INFO  Index length increased by 4729 = 1619641 - 1614912\n",
      "INFO  shape=1619641\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>full_symbol</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>currency_pair</th>\n",
       "      <th>exchange_id</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</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>2018-08-17 00:00:00+00:00</th>\n",
       "      <td>binance::BTC_USDT</td>\n",
       "      <td>6316.00</td>\n",
       "      <td>6319.04</td>\n",
       "      <td>6310.32</td>\n",
       "      <td>6311.64</td>\n",
       "      <td>9.967395</td>\n",
       "      <td>BTC_USDT</td>\n",
       "      <td>binance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-17 00:01:00+00:00</th>\n",
       "      <td>binance::BTC_USDT</td>\n",
       "      <td>6311.64</td>\n",
       "      <td>6311.77</td>\n",
       "      <td>6302.81</td>\n",
       "      <td>6302.81</td>\n",
       "      <td>16.781206</td>\n",
       "      <td>BTC_USDT</td>\n",
       "      <td>binance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-17 00:02:00+00:00</th>\n",
       "      <td>binance::BTC_USDT</td>\n",
       "      <td>6302.81</td>\n",
       "      <td>6306.00</td>\n",
       "      <td>6292.79</td>\n",
       "      <td>6297.26</td>\n",
       "      <td>55.373226</td>\n",
       "      <td>BTC_USDT</td>\n",
       "      <td>binance</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 full_symbol     open     high      low    close     volume currency_pair exchange_id\n",
       "timestamp                                                                                                            \n",
       "2018-08-17 00:00:00+00:00  binance::BTC_USDT  6316.00  6319.04  6310.32  6311.64   9.967395      BTC_USDT     binance\n",
       "2018-08-17 00:01:00+00:00  binance::BTC_USDT  6311.64  6311.77  6302.81  6302.81  16.781206      BTC_USDT     binance\n",
       "2018-08-17 00:02:00+00:00  binance::BTC_USDT  6302.81  6306.00  6292.79  6297.26  55.373226      BTC_USDT     binance"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vendor = config[\"data\"][\"vendor\"]\n",
    "universe_version = \"v3\"\n",
    "resample_1min = True\n",
    "root_dir = config[\"load\"][\"data_dir\"]\n",
    "extension = config[\"data\"][\"extension\"]\n",
    "aws_profile = config[\"load\"][\"aws_profile\"]\n",
    "ccxt_csv_client = icdcl.CcxtCddCsvParquetByAssetClient(\n",
    "    vendor,\n",
    "    universe_version,\n",
    "    resample_1min,\n",
    "    root_dir,\n",
    "    extension,\n",
    "    aws_profile=aws_profile,\n",
    ")\n",
    "start_ts = None\n",
    "end_ts = None\n",
    "ccxt_data = ccxt_csv_client.read_data(\n",
    "    [\"binance::BTC_USDT\"],\n",
    "    start_ts,\n",
    "    end_ts,\n",
    ")\n",
    "_LOG.info(\"shape=%s\", ccxt_data.shape[0])\n",
    "ccxt_data.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6919e835",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:04.539398Z",
     "start_time": "2022-01-14T09:58:04.205524Z"
    }
   },
   "outputs": [
    {
     "ename": "AssertionError",
     "evalue": "\n################################################################################\n* Failed assertion *\n'UTC'\n==\n'US/Eastern'\n################################################################################\n",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAssertionError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_365/3557880064.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;31m# Check the timezone info.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m hdbg.dassert_eq(\n\u001b[0m\u001b[1;32m      3\u001b[0m     \u001b[0mccxt_data\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtzinfo\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m     \u001b[0mconfig\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"data\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"timezone\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m )\n",
      "\u001b[0;32m/app/helpers/hdbg.py\u001b[0m in \u001b[0;36mdassert_eq\u001b[0;34m(val1, val2, msg, *args)\u001b[0m\n\u001b[1;32m    134\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mcond\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    135\u001b[0m         \u001b[0mtxt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"'%s'\\n==\\n'%s'\"\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mval1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 136\u001b[0;31m         \u001b[0m_dfatal\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtxt\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmsg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    137\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    138\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/app/helpers/hdbg.py\u001b[0m in \u001b[0;36m_dfatal\u001b[0;34m(txt, msg, *args)\u001b[0m\n\u001b[1;32m    113\u001b[0m             \u001b[0mdfatal_txt\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;34m\"\\n\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    114\u001b[0m         \u001b[0mdfatal_txt\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mmsg\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 115\u001b[0;31m     \u001b[0mdfatal\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdfatal_txt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    116\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    117\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/app/helpers/hdbg.py\u001b[0m in \u001b[0;36mdfatal\u001b[0;34m(message, assertion_type)\u001b[0m\n\u001b[1;32m     59\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0massertion_type\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     60\u001b[0m         \u001b[0massertion_type\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mAssertionError\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 61\u001b[0;31m     \u001b[0;32mraise\u001b[0m \u001b[0massertion_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mret\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     62\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     63\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAssertionError\u001b[0m: \n################################################################################\n* Failed assertion *\n'UTC'\n==\n'US/Eastern'\n################################################################################\n"
     ]
    }
   ],
   "source": [
    "# Check the timezone info.\n",
    "hdbg.dassert_eq(\n",
    "    ccxt_data.index.tzinfo,\n",
    "    config[\"data\"][\"timezone\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d5093fd1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:07.784044Z",
     "start_time": "2022-01-14T09:58:07.778831Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<DstTzInfo 'US/Eastern' LMT-1 day, 19:04:00 STD>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# TODO(Grisha): change tz in `CcxtLoader` #217.\n",
    "ccxt_data.index = ccxt_data.index.tz_convert(config[\"data\"][\"timezone\"])\n",
    "ccxt_data.index.tzinfo"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "435a1986",
   "metadata": {},
   "source": [
    "# Select subset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6b45d4f4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:09.851277Z",
     "start_time": "2022-01-14T09:58:09.725730Z"
    }
   },
   "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>close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:00:00-04:00</th>\n",
       "      <td>6311.64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:01:00-04:00</th>\n",
       "      <td>6302.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:02:00-04:00</th>\n",
       "      <td>6297.26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             close\n",
       "timestamp                         \n",
       "2018-08-16 20:00:00-04:00  6311.64\n",
       "2018-08-16 20:01:00-04:00  6302.81\n",
       "2018-08-16 20:02:00-04:00  6297.26"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ccxt_data_subset = ccxt_data[[config[\"data\"][\"close_price_col_name\"]]]\n",
    "ccxt_data_subset.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "184f4b2c",
   "metadata": {},
   "source": [
    "# Resample index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1a08df11",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:12.780024Z",
     "start_time": "2022-01-14T09:58:12.752124Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape=1619641\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>close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:00:00-04:00</th>\n",
       "      <td>6311.64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:01:00-04:00</th>\n",
       "      <td>6302.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-16 20:02:00-04:00</th>\n",
       "      <td>6297.26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             close\n",
       "2018-08-16 20:00:00-04:00  6311.64\n",
       "2018-08-16 20:01:00-04:00  6302.81\n",
       "2018-08-16 20:02:00-04:00  6297.26"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# TODO(Grisha): do we want to merge it with `core.pandas_helpers.resample_index`?\n",
    "# The problem with `resample_index` in `pandas_helpers` is that it does not\n",
    "# generate empty rows for missing timestamps.\n",
    "def resample_index(index: pd.DatetimeIndex, frequency: str) -> pd.DatetimeIndex:\n",
    "    \"\"\"\n",
    "    Resample `DatetimeIndex`.\n",
    "\n",
    "    :param index: `DatetimeIndex` to resample\n",
    "    :param frequency: frequency from `pd.date_range()` to resample to\n",
    "    :return: resampled `DatetimeIndex`\n",
    "    \"\"\"\n",
    "    hdbg.dassert_isinstance(index, pd.DatetimeIndex)\n",
    "    min_date = index.min()\n",
    "    max_date = index.max()\n",
    "    resampled_index = pd.date_range(\n",
    "        start=min_date,\n",
    "        end=max_date,\n",
    "        freq=frequency,\n",
    "    )\n",
    "    return resampled_index\n",
    "\n",
    "\n",
    "resampled_index = resample_index(\n",
    "    ccxt_data_subset.index, config[\"data\"][\"frequency\"]\n",
    ")\n",
    "ccxt_data_reindex = ccxt_data_subset.reindex(resampled_index)\n",
    "_LOG.info(\"shape=%s\", ccxt_data_reindex.shape[0])\n",
    "ccxt_data_reindex.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c36ce7f5",
   "metadata": {},
   "source": [
    "# Filter data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7a77280c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:17.713029Z",
     "start_time": "2022-01-14T09:58:17.710453Z"
    }
   },
   "outputs": [],
   "source": [
    "# TODO(Grisha): add support for filtering by exchange, currency, asset class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "acaf70a0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:20.116439Z",
     "start_time": "2022-01-14T09:58:20.060279Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filtering between 2019-01-01 00:00:00-05:00 and 2020-01-01 00:00:00-05:00 with inclusive=`left`, selected rows=525600 / 1619641 = 32.45%\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>close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-01-01 00:00:00-05:00</th>\n",
       "      <td>3697.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-01 00:01:00-05:00</th>\n",
       "      <td>3697.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-01 00:02:00-05:00</th>\n",
       "      <td>3696.32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             close\n",
       "2019-01-01 00:00:00-05:00  3697.35\n",
       "2019-01-01 00:01:00-05:00  3697.29\n",
       "2019-01-01 00:02:00-05:00  3696.32"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the inputs.\n",
    "# TODO(Grisha): pass tz to `hdateti.to_datetime` once it is fixed.\n",
    "lower_bound = hdateti.to_datetime(\"2019-01-01\")\n",
    "lower_bound_ET = config[\"data\"][\"timezone\"].localize(lower_bound)\n",
    "upper_bound = hdateti.to_datetime(\"2020-01-01\")\n",
    "upper_bound_ET = config[\"data\"][\"timezone\"].localize(upper_bound)\n",
    "# Fiter data.\n",
    "ccxt_data_filtered = coexplor.filter_by_time(\n",
    "    df=ccxt_data_reindex,\n",
    "    lower_bound=lower_bound_ET,\n",
    "    upper_bound=upper_bound_ET,\n",
    "    inclusive=\"left\",\n",
    "    ts_col_name=None,\n",
    "    log_level=logging.INFO,\n",
    ")\n",
    "ccxt_data_filtered.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3468919",
   "metadata": {},
   "source": [
    "# Statistics"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79db714e",
   "metadata": {},
   "source": [
    "## Plot timeseries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b66a5499",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:27.400930Z",
     "start_time": "2022-01-14T09:58:23.100836Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJYAAAE9CAYAAABZdOj5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAACA/0lEQVR4nO3dd3xTVRsH8F+S7j1oS2nZ0FL2KEt2UUFkKaA4cBThRWWqQB0v4C6giAqIKFVRxusAEVBBLHtvKKuMMkr33ivJ+0fIbdKMZrRJ2v6+n48fc+859+bJaWnTJ+c8RySXy+UgIiIiIiIiIiIyktjaARARERERERERUd3ExBIREREREREREZmEiSUiIiIiIiIiIjIJE0tERERERERERGQSJpaIiIiIiIiIiMgkTCwREREREREREZFJmFgiIiIiIiIiIiKT2Fk7gNqQnV0ImUxu7TAaFF9fN2RmFlg7jHqH42o9HHvL4nhbB8fdsjje1sOxtx6OveVwrK2HY29ZHG/rEItF8PZ21dpWLxNLMpmciSUr4JjXDo6r9XDsLYvjbR0cd8vieFsPx956OPaWw7G2Ho69ZXG8bQuXwhERERERERERkUmYWCIiIiIiIiIiIpMwsURERERERERERCZhYomIiIiIiIiIiEzCxBIREREREREREZnEoMTS4sWLERERgdDQUMTHx2u0r1ixQqPt7NmzGD16NIYNG4bIyEhkZmaa3UZERERERERERLbDoMTS0KFDsX79egQFBWm0Xbx4EWfPnlVrk8lkmDt3LhYsWICdO3ciPDwcn3zyiVltRERERERERERkWwxKLIWHhyMwMFDjfFlZGd577z0sWrRI7XxcXBwcHR0RHh4OAJg4cSL+/vtvs9qIiIiIiIiIiMi22Jlz8eeff47Ro0cjODhY7XxycjKaNGkiHPv4+EAmkyEnJ8fkNi8vL4Pj8vV1M/1Fkcn8/NytHUK9xHG1Ho69ZXG8rYPjbrzz19Ph5+WCwEauRl/L8bYejr31cOwth2NtPRx7y+J42xaTE0tnzpxBXFwc3njjjZqMp0ZkZhZAJpNbO4wGxc/PHenp+dYOo97huFoPx96yON7WwXE3zdtfHQYAxERFGHUdx9t6OPbWw7G3HI619XDsLYvjbR1isUjnJB6TE0snTpzAjRs3MHToUABASkoKJk+ejI8//hiBgYFISkoS+mZlZUEsFsPLy8vkNiIiIiIiIiIisi0G1VjSZurUqTh48CBiY2MRGxuLxo0bY+3atejfvz86duyIkpISnDx5EgCwadMmDB8+HABMbiMiIiIiIiIiItti0IylDz74ALt27UJGRgZefPFFeHl5YceOHTr7i8ViLFmyBAsXLkRpaSmCgoKwdOlSs9qIiIiIiIiIiMi2iORyeb0rRsQaS5bHda61g+NqPRx7y+J4WwfH3XgymRwvLdkDgDWW6hKOvfVw7C2HY209HHvL4nhbh74aSyYvhSMiIiJqaApKyq0dAhEREZFNYWKJiIiIyEB/HEywdghERERENoWJJSIiIiIDlZXLrB0CERERkU1hYomIiIjIQLL6V5qSiIiIyCxMLBEREREZiIklIiIiInVMLBEREREZiLvOEhEREaljYomIiIjIQKp5pdIyqfUCISIiIrIRTCwRERERGUiuklk6djnVipEQERER2QYmloiIiIgMFNjIVXh84HySFSMhIiIisg1MLBEREREZqEVjd+FxebnMipEQERER2QYmloiIiIgMJFfZFU7KHeKIiIiImFgiIiIiMpRqLok7xBERERExsURERERkEiaWiIiIiJhYIiIiIjKYaiopNbvYanEQERER2QomloiIiIgMJK8DdZVkcjmOXUpFZHQslmw4be1wiIiIqJ6zs3YARERERHVFHcgrYf/ZJKzbeRUAcOVOjnWDISIionqPM5aIiIiIDCSH7WeWNv57zdohEBERUQPCGUtEREREBtj07zXsOnHX2mFUq7xCZu0QiIiIqAHhjCUiIiIiA9SFpBIRERGRpTGxRERERFSPBPq6WDsEIiIiakCYWCIiIiKqR5Izi4THrZt4WDESIiIiagiYWCIiIiKqRwZ1bSI8dnO2t2IkRERE1BCweDcRERFRPbB6axy83Bzh4lT59s7297AjIiKiuo6JJSIiIqI67vLtbBy/nAYAGNojGM6OEvh7uUDOzBIRERHVMi6FIyIiIqrjjlxMER5LpTLYS8SQSESQM7NEREREtYyJJSIiIqI67uD5ZOHxvYxClJRJIQKYWCIiIqJax6VwRERERPXItcRcAIBIJGKNJSIiIqp1nLFEREREZKIKqczaIQAAxg5oqXFOJAJrLBEREVGtY2KJiIiIyEQZuSXWDgEA4O7ioHFOJGKNJSIiIqp9TCwRERERmUg1cZOQnIepS/egqKTC4nHIZJoJJBEALaeJiIiIahQTS0REREQmylSZsfT+DydRIZXj9wM3LR6HVEsG6erdHMTfzbF4LERERNSwMLFEREREZKK8ojKNc7tPJVo8DqnMNmo9ERERUcPDxBIRERGRiXIKNBNL1nDpVra1QyAiIqIGioklIiIiIhP9uveGtUMAAFxMyFI7HtilifD4RlKupcOpV/45eRdpOcXWDoOIiMhmMbFEREREVM+M7tdCeHzxZpbujqTX1TvZ2Lj7GqJWH7F2KERERDaLiSUiIiIiMyVnFqodr98Vj/e+P2GR5750SzNx5OPhJDz+/WCCReKoj/KLyq0dAhERkc1jYomIiIjIQA522t86vf3NMbXjf08n4lZKPtb8cbHWY0rNKqr152ioriVyGSEREVF1DEosLV68GBEREQgNDUV8fDwAIDs7G1OmTMGwYcMwatQoTJ8+HVlZlZ+YnT17FqNHj8awYcMQGRmJzMxMs9uIiIiIrEG561pZhebua3K5XOd1Ry+l1lpMSrmFtlFAvD5ycbJTO5bKZDh6KcVK0RAREdkmgxJLQ4cOxfr16xEUFCScE4lEeOmll7Bz505s27YNTZs2xSeffAIAkMlkmDt3LhYsWICdO3ciPDzc7DYiIiIiaykt00woKWXmllgwEk2qy94A4PGBrQAAzo4S4VyFVHf8pFuHlj5qx5//ch5r/riEpRvPWCkiIiIi22NQYik8PByBgYFq57y8vNC7d2/huGvXrkhKSgIAxMXFwdHREeHh4QCAiRMn4u+//zarjYiIiMhaikp119o5d8O6s6t97yeWlIkk5Y5wzz4UKvQ5dTXd8oHVA1Vno8Xd333v8u1soa5WZHQslmw4bfHYiIiIbEWN1FiSyWTYuHEjIiIiAADJyclo0qRym1sfHx/IZDLk5OSY3EZERERkLSl66hhl5emfsZRXZNhStb+O3kZCcp5RcQGA/f26T6+M7YSYqAh4uDoAACQSkdDn0IVko+9LQNVVjq2DPITHqnW1rtzJAaCYGaZcNklERNRQ2FXfpXrvv/8+XFxc8Oyzz9bE7czm6+tm7RAaJD8/d2uHUC9xXK2HY29ZHG/r4LgbZll0rM625Oxivdem5ZWhdXNfAPrH+5e9NwAA2z4dY1RsqXmlAABvLxe1+7cqqJxlJZaIG/zX2tjXL5XJ4ZFbqnb9g72a48aWC8K5nScThce/HUjAjkOKHfiM/RrWdw39e8+SONbWw7G3LI63bTE7sbR48WLcvn0bq1evhlis+MQsMDBQWBYHAFlZWRCLxfDy8jK5zRiZmQWQyXQX0qSa5+fnjvT0fGuHUe9wXK2HY29ZHG/r4LjXjHPX9C8z++nPS2jT2E3veKvOcjHmayKVyfDTX5cBALl5xWrXuqi8y7uUkNmgv9amfK9HRsfC0b6yTlV6ej4KCkrV+vxvd7zwWJlUAoC4q6kI8HExMdr6hT9nLIdjbT0ce8vieFuHWCzSOYnHrKVwy5YtQ1xcHFauXAkHBwfhfMeOHVFSUoKTJ08CADZt2oThw4eb1UZERERkCyY9HIJGnpUFs/VsCgcAuJNWUO09r92t3NY+MjoWiQZcAwALY07g4v26P6nZ6sv1RKLKpXAP9Wyq1rbhn3jEnk4EaZd/f/liablU7bzKkOr15pqjkFX3jUFERFRPGDRj6YMPPsCuXbuQkZGBF198EV5eXli+fDm+/vprtGjRAhMnTgQABAcHY+XKlRCLxViyZAkWLlyI0tJSBAUFYenSpQBgchsRERGRLRjSPRjuLg5Y9Xtcjd2zapHoBTHHERMVUe11SRmFwuPCYt0Fxn3c1XeO231KkVSK6B5sTJgNwr2MQvz322Ma509eSYOBeSUAQHm5DI4Okuo7EhER1XEGJZbeeecdvPPOOxrnr169qvOa7t27Y9u2bTXaRkRERGRN70X2AmD4zBVDubs4aJwrKC6Hm7O9wffoEeqvsy2noBTxd3MQ0tRL7XxpmZTJjyp0zRZb9XscWgYaXtOjXCqDIzi2RERU/9XIrnBEREREDUGTRq61cl/VHdyUSsukWnrq1lhPTZ8dR24jev1pyORyHL+cKpy/cDPTqOdoCPTt4peQbHhNj7Jy475+REREdRUTS0REREQGEosVCaCcAt3JB1No23TkRlKulp7qeoT6AQDefLa7RptYy7SqlxbvweqtF4XjO2ksflrV1Ts5BvXTNZvsxRHtAIA1loiIqMFgYomIiIjISDeT8qrtMzGijcH305aDUE0AAcD+c0k4E6++A523uyOcHe3QNthL43pDlrhtP3zb4BgbimuJOQb1C7+f1AMgLDFcOWegkNDjBsVERNRQMLFEREREZLTqswYRPSoLY1dIZXr7GjK75fu/ruDLzRfUo5AD4hqu99TQ5RfpLoKu6qkH2wqP27fwRkxUBJwd7YTEklwls1S1ODsREVF9wsQSERERkdE0szmO9pUzhN6e1AN2ksq3WdqWuqkyddmUXC6HyMxK4pduZZl1fUNlb1f59W4d5Ck8Ft3/sqt+TScv3oPI6FiLxUZERGRJTCwRERER6aFttonqbBWlUpVizc0bq+8eduVOts57p2YVaV0Kp9Ynu0jtuLxCih/+voKi0gq9O9S9MbGr7sb7Ptl0tto+DdHIB1rgwym98fHUPnhpZJhaW79OjdWOWzb2EB5XXQpXXsEi3kREVL8xsURERESkQ1FJOSYv3qNx3s3ZHrPGd1Y7Z28nRq8wf3i5OQizlYZ2VyyHW/7LeUi1LIeLPX0Pb645iuuJmoW6HxvQEgDw9/E7ePPro8L5exmF2H0yEfvOJuHoxVS9M5baNfc24FWSNoO7NkGgrysCfFwgqjJDbfKj7QEA/30+HG9P6gEXJzuhLSVTkQTcd/YeKqQyzP7ykNCWW1CKG/dyq10aSUREVJfYVd+FiIiIqGE6HZ+hs61qPic81A9TRnVQO+fj4Sg8FmsphrT+n3gAwMZ/rwFQJKwKitVr/Ow+mah2/MnGM8hTqQOUV6h7hzpDF8lFRsfiqQfb4qHwpgZeUf/5eDgJj8tUZh2NfKCF8LhloAeqSs8tBqD4ulX92s1ZUZlkiomKqKlQiYiIrIozloiIiIjuu5mUp7bsLObPyzr7Vp0p5OHqoKVT5UN9CSCl54aF4tv5QwAoCnOfu56B7PxStT55BhaX1hajqpnj1Gdcbdx9zeD7NjSxp+8Jjx8f2EpvX2cHwz63/finU2bFREREZCuYWCIiIiK674N1J9WWnelTNWXTv1Oglj6VvQwpsn0nLV+4IrugFJ//et6gWIw154kuCPJzrZV71wfdQ/zUjrPySgy+1t7OsLfX17QsfyQiIqqLmFgiIiIiMkDn1r7qJ6rkiYL83DSu2XnijvBYKqu+ro4IIohEinTUvrNJpoRpkE6tfGHaPnQNg0OV5FCrJopd3xp5OmnrrkYiMfztdV5R9bPYiIiIbB0TS0REREQGCKtSCNuQGUi5BZWJgwNn7unpqdAiULGbnDFJn5VzBhrRu5K23e5I4eilVLXjvh0CAACNfVyqvXZEn2Ya52KiIjCmv6IY+5DuQcL52V8cNCdMIiIim8DEEhEREZEWd9MK1I4f6qle2NrQwthK32yNq7ZP++Y+OtvcnO01zsVERcDZ0bS9WBzsJCZd1xApx97L3bGanoCTgx1ioiLgXaXv4G5BaOTphAd7BNdKjERERNbCxBIRERGRFlWLZourzFCSyYyf8bPtUILedjs77ekqB3sxpj/eyejnA4CoZ7pjybS+GuerJj4AoFxl97OGSNcsro6tfDH/6W54YXg7g+/16av91I49XR2w5OUHEOjriuUz+gMAhjLJRERE9QATS0RERNQgVUhliIyOxZ7Tii3hi0srhLa0nGIs/+Wc3utvp+Yb/ZxbDuhPLEnE2t+alZXLkFNQqrWtOiFNvdDIy1lr29ynuqkdr92hexe8hkCqJ1kY2swbYrGx89S083B1gJODRCNZSUREVBcxsUREREQNUlm5opj2j7viAQDlFZXFtaNWH1HrO69KAkb1ektxtFdfuqZRTNxAzfwri4xX3cHs+OU0k+5ZX0ilisRSTS1Xe2tSD7w4Qvssp5IyKf45ebdGnoeIiMiamFgiIiKiBqnqZBFds1XGDmiJdlUKdwOApIZmrxiqXbPKGL56fRBmT+hi0n3mPl2ZJGuqZSc7pbtpBYiMjjVpyR8A/Ljrap3b9Uy5c58hu78Zok2QJwZ0blIj9yIiIrJVTCwRERFRg5SSVaR2XC7VPgMporv22Su6+qt65qEQve2qS6GqWxXl6CDBR1P7YPXrgzRmLxmqkacTXJ0qi4A72Ot+K7gw5jgA4Os/Lhr9PFfvZGPP6Xt1btezohLFckiRhZOGREREdRkTS0RERNQguarssvbjrquoqNCeKNK2GxsA3EqprLH0UHhTrX2qm/kiUykW/fUbg4XHn7zygNb+jX1c4GBiUin6P32w6MVeaudEBtT4OXHF+OVx8Ym5AIBmAbpnRNmajNxizLu/BPLoxRSLPe/B88k1cp+SsorqOxEREdUCJpaIiIiowbl+L1etjtKe0/eQkVti1D1UUzJPPdhWa59WTTx0Xr/r+B21YztJ5dsy1VlFNcXf2wUuTnY1fl9ttuy/CQC4k1pgkeczh0wuh0wmx/kbmcK5hGTjC7ObKuZP8wum7zhyC68s24+fY6/XQERERETGYWKJiIiIGpwz8eka53RtNa+TAaul3F0cMG1MB61tm1SSAPOfVi8Ozs3CLOelxXvw0pI92HW8spD2hMGtrRiR8X7bp0jk/V0lWVnXLV5/WlHny9h/m0REZFFMLBEREVGDo60+0ue/njfqHgVF5Qb1q1rkW1sCK7SZenFwW0ssXbqVZdJ1HVv61HAkNWfaJ3sRGR0rHKflFAuPH+nT3BohmWxQ18oC4cmZhZi76rDxiVIbdPVuDgBg5vID1g2EiIj0YmKJiIiIGpzjlw2rG/Tt/CE62wytdSSuklgyZPZF1WvCQ/0Mei5TPNq3OcK07Hqn6pNNZ026d1yCaQkpSyjTUVMLMOxrZC7VwunmPF9hSTn2nU0Sjt/+5hgy80qwU2UGVl1w4WYmbqXkCcequxEWlbJ+FBGRLWNiiYiIiBqcvMKyavusnDNQbde2qgzdmU0iVn+7VSGtPokgEYsRExWBb+cPwesTu+KVxzoZ9FymGDeoNZ4Y0sbk64tKKrD1YAIycou1tiem236dJVXDezXT+3WvKQtf6Ck8LjYjcTJDx2yen/fYdr2lQxeSsXprnHD82c/n8N73J4Xjl5bsER7XtaWJREQNDRNLRERERFo4O+ovdN2nQ4BB92ns46x2vP3wLRSVGLaMTiwSoUOL2l9O5upc+Vojo2MNXkYVezoR05fvx9aDCVix+YLWPgvWHq+RGC2lf+dAizxPoK+r8DjTyMLxSofjamZHOWtYu+Myjl9OU1uOqJONLQ0lIiJ1TCwRERERmaBXmL9B/fy9XfDFrAHC8Y4jtzFdZZZJBxuoQ+Rgpz77ytCVWT/tihcel5XLUFYuxRSVmSZKxy6lIjmz0KwYa0pRSQW+/uOizvZyPUvkasui704g9nSi0dd9u938HeWIiIjMxcQSERERURWhTb2q7VN1iZs+bs72Ots6t/I1+D61xd1FPb7955J09KyUnqO+9C0lqwiH4lIglWlmpb7+4yLe/uYY3v7mqHmB1oD/xV7DsUupwvHH/+mj1u7oYNgSx5qmmqQzlKHJTWMUlZSjvEJa4/fVZ8by/cLjyOhY7DpRpT5U3a9DTkRUrzGxRERERA1OqyYeWs8rdzGbOb6zxWLpbeCSutokqlJTaN3Oq9Veo2253I/VXJecWYScglLjgqtBUpkMB86rLx8L8HZRO27so35cm5o0cq2+kx7K2W4OdmL07dBYrc3QpZpVTV9+AB+sO2VWXMYqLFGvMbXp32s6+569noE/DibUdkhERGQEJpaIiIiowfF2c9R6/qkH2yImKqLa+ko1ycPFwWLPZY6UrCK1Y1MLXP9TdTaKBW07dEvt2NVJ8XUOCfYEAMRERVg0nkFdm+htLy6twG/7bqBCqn15nnLntI//0xfPDQsVznu6OUBqQJF4Xe6m2VbBdTmACqkMc1YcxBe/nsfvBxPw7ynjlw4SEVHtYGKJiIiIGhxty51ioiLUCiob4tG+zTHygeYG9R3ULdioe9uaM/HpascSiWlvI/86dqcmwjHJH1USS8qZMvOe7o6vXh9k8Xge7KH/e+Kzn89hx5HbmLp0r9r564m5+OfkXWHZoVgsgqODBDFREYiJioCbk72QdLJVrYO0zxpUpfo1yS8qR25B5W6O6/9RLB2UyeWQGVoUjIiIagUTS0RERNTgHI5LqZH7jBvUGo8PNGwr9H5dLLPbWG3JMHLnsq9es3yiRp9F32nuTvfWpB4A7idm7C1fW6nqEkSlQxeS8fexO7h+L1c4Fxkdi4JixW6CH/10Cht3X8PPsdcBABKx+n3EYpHWWldKN5JysffsPb27/5WUVehsM4dyFtaNe3nV9lW+KrlcjmuJOVr7vPf9Cby0WLNgfFVyuRzrd8XjXrptzcYiIqoPLDfPm4iIiKgBKy61bEFkczRp5IqkjEK4OtkJs3riq/xhfzctX+f108Z0gKODBK+M7YhVv8eptXm6Wmfp351UzYRCi8buVohEnXKsAUXyQyQSYe0O7bu9zfz8gNpyvbL7O9hVXZYoFot0zuK5dCsLn2w6CwC4cCMTM8Zpryf2yrL9GNO/JVoHeaBjy5orMP/qZ/vVjv28nJCeo0havhvZC65OdpCIRXB3cYBUpnh9v+27qfVekdGxBj9vYnoh/j2diH9PJ6qN4a2UPLz3/Um8MrYjwtvVfDF0IqKGgDOWiIiIqMFRzvB4uGdTDOkWhAUvhFs1DlujTHS0b+EjnFMWNlfKUVmWpCo81A+9whSFo/29nYXz85/uBgDILSzD3FWHazReU9mZuJyvJs2Z0EV4rEwU6VOupY+To/psKzs9M5aUSSUAOHMtQ++spa0HE7Dsf+eqjckci6c9IDz2dHOAj4cTPN0cITby30ZkdKzeRNPCmMoZa5dvZ+Ps9Qxcvp2N974/CQBCAjQjp1hv8XAiItLEGUtERETU4Izu1wJbDiRg3KDWsLezTHLhyu0sjXPK4tG26qWRYbhwMxMlZVKENVdPLCl3T+vQ0gcXEypf2yuPdRIee7tXFklv6u8mPM7MM25ZXX3m6+lkVP9ELUu5qs5YupFU/TIzpS9/u4CZ4ztDLpdr7M5mKVHPdMftlHyDCtkvntYX81cf0dmunPWl6sSVNLXjpRvPaL1WNTG168RdeLo6YPzg1ujXSfcy1r1n7iG0mZfR9dnI+m4m5aFFoLvJGxEQUSXrf0xDREREZGHKORqWnDHUMlCzWLGlklqGiOgepHHO3k6CVx9XJIrsJOpj5e5iDwDo16lym/vwUL8qfRzQpbUvxvRvicqKOQqpVXaZsyRlkWtb8diAlgCAuJuayceqotefNvi+unaTU3X2egYARYJk5ucHqr1f9E+ncOZaut5+hlImHkOaeuGhnk012u3t1GdiLZveD35ezhr9VE3WUm+poEj77Lrq5BaWYe2Oy5DKZFiz7aLW3fLW7byKt785ZtL9yXou3crCB+tOYrcVd6kkqk9s590MERERkYUod8yy5AfVzk72GudsYSmW0sShbbUnl+7HWHXBlHKlleqn/cN7a+6QN2tCF4zp31JjrLceTEBcQqZZMZvClhJKSlsOJAAAVm65UG3fqkvhZo3XXiMJAL749bzw+Ma9XCQk657JtOl+IXBt1my7iMu3snDjXi7iE3Px5W/Vx1kdF0c7vbFX9dn0fvByUySiZk/ojDlPdNHZ95ttF9WSalfv5ui9d8D92Xe6TFmyF0cvpmJhzHFsP3xLa59lP59FRk6x3vuQ7VBuRpB4f9kvEZnHtudfExEREdUC5R/nunblqg3aJkfZ0owlO4kYWXmlujtUySzJlVvdq4yhvtdTtZj00UupOHopFS+OaIcBnZsYH3A9dbLKsi1dJj8ahlZNPPQuwYpLyMJv+25g3KDW+PDHUxrXK4uEb95/QyNh9WREG/zvfrLp6MVUHL2YaszL0EpZjBsAVswZaNA1zo52KC6tgKdb5bLKzq0bAQBefawjVm6J07jmyMVUlFXI8OpjnfD+DyeRla9IIjzzUAjaNfPCf9cq6i01C3CDi6MdZk/oAgd7CeLv5sDPyxnlFVJEfX1Uazyb99/E5v2KYuJr5w8RzsfdzMK81UfQI8RPmOVHtqukVLHs05ozJ4nqE9t5N0NERERkIX8du2Px59RWjFjflvDWoFwWpUqZN5JXySwpE0UikUhY0qQvseRoL9F6/rs/rwAAvvo9DpHRsdh5/I7aVvf7zyXh7W+0/5FfHZlMjtsplbvXDeisu1aOrai6ix4ATBnZHgO7qCffiksrDKrrs+PIba1f18Yqs3S2H76t0a5taZq5Lt/KNvqaZdP7YaWOJJS7Sk2m2RO6oG+HymWZp66mY8eRW0hIzkPu/ULzPdv5I8ivstbX0B7BmPd0dzjc/94MaeoFb3dHg/9dpmVrzlA6FZ+OZf87a9D1ZD0X7teFu5aYa+VIiOqHahNLixcvRkREBEJDQxEfHy+cT0hIwJNPPolhw4bhySefxK1bt2q1jYiIiKgu0zY7ypaWwukiqswsqZELS+GApx8MgbOjHXw9HKFLda9VWWD5f7HX8c22S8L57/+6guTMIqzeqplw0Ucul2Pj7mt49/sTiL+TbVAM1qJa5Fybvh0bq+2wBwB7zybp7N9MpVA6oL4kTqlVE82aX6rEIpFawfWacFPPUjxdHO0lcHbUvshCtQh859a+mDKqvVr7b/tuqh17uCoSUV/MGoAHewSrJaJUuWspIj798U6YNCxU7ZxyOVVVcQnV18oi69K3GyIRGa/a365Dhw7F+vXrERSkvuZ+4cKFePrpp7Fz5048/fTTWLBgQa22EREREdVlLlr+OK5aENsWpd+vG3P8svoSLeUMJpFIhB6hflg5Z6BGseWq/jO6g0HPeeaa5gyb45fTDCpGrbTjyG38ezpR8fiQoobRnjP3DL7ekmaMq37pVNXkk7LgtzZvTupR7f0MWQb6+pNdq+1jDF0JIlN1ub8k7rGBrYy6zs3ZHk8/FKIz0ejmXFkP7cvZAxATFYHuIX7o11GRiAoJ9gQAfP3HRQCKmXCP9G6mdo+lG8/g8q0sREbHYunGM4iMjkVRSblRcVLt4S5+RDWr2sRSeHg4AgPVpw1nZmbi0qVLGDlyJABg5MiRuHTpErKysmqljYiIiKiu6xrip3HOVUtBb1vx2Yz+AIDE+zthHbusXmNH+YG/MWWqercPwKev9lM7N7xXMx29FcKaewuPjZkJoqyDAwCxJxU7Pzk56E98WUuLxvpnDwHqsb89qQd6hPrr7Ktr2aFS/06GLQn0cHXAW88qklTvPBcOoHI5oSkzPrLuzzCa/GiY0ddq4+Jkh5ioCIx6oEWN3E+VcudA1X+jDvYSxERF4Nn7M5cKihWJogPnkzFhSBu1wvCXb2dj6aazwmMA2PjvtRqPk0zT6v4unX3aB1g5EqL6waSPDZKTkxEQEACJRPFLSyKRwN/fH8nJyZDL5TXe5uPjUxOvlYiIiMhqRCIRlrzcF+t3xePcDcVuaM0bu1s5Kt087y8bUtaGkkorEwkVUhlO3J/BZGwBdOV9lZT1bFoGuiMhubIeUmp2EQK8XeDr6SScC/DWv9V8dYyd2WJLOrf2xfBezTCsdzONMTTUuEGt0K9ToLC7miHaBHsKCZOYqAj8cVAx+0suN35XxdZNPAHctfj3ffR/+qh9H5nL1HmGhy6k4LEBreDjUXOxkIm0r/AlIhPVy13hfH1rdj04GcbPz3bfHNdlHFfr4dhbFsfbOhriuOcVlgmPLf36w9r4o/OtHCGx1KaZj019Ddo09cL1+1uzK+PydFf8ESyTy+Hn546XPvxHbSclby8Xs16DvYMER6+k415GERzsJSgrlwIA1u64jM/mDIadyvI6b29Xs55r7JC2cLHRWWKd2zTCeS1FtoHKr8WrT3Yz6zleGK17yd13/30YL76/CwAQHhagc5xd7yelGvm5Q6Jtq0MtlPdaFR0LAPDwcK617/txQ9rgtz3X1c51CKnZWSnF95OsTQPccDe1AFuXjhYSsBvefwRP//cvnde+seowfvn4UTg51M6fYbb088SWebjnAAAcHe1qbMw49pbF8bYtJv1ECwwMRGpqKqRSKSQSCaRSKdLS0hAYGAi5XF7jbcbKzCyAzMZ2Wanv/PzckZ6eX31HMgrH1Xo49pbF8baOhjjucrkckxfvEY4t+fqV4/3jX5eFc019nW3qaxDg6YTrilVjQlylpZV1YdLT8zW2587LLzbrNfx5+Jbw2MfDEVn3E0vXE3ORlJyD1MxCoT0zqxBOZtTfLswvQWG+9oLL1jb9sY4or5BhYcxxdG3bCLtPJgptpoxvi8buuJWifp2u+3w8tQ/k5RWY/3Q3LN5wBo/2bqazb2Fh6f175UEirv6Loe3nTHJqHjwca2dZ4qO9myE02BMf/XgKAPDMQyE1/m8sO1vxb+BuqmKZaGZmgVHXX7iaatDyR2M1xJ/ppsq7/3OgtLRCbcxkMjk2/XsND/VsCj8vw2dIcuwti+NtHWKxSOckHpN+Nfv6+iIsLAzbt28HAGzfvh1hYWHw8fGplTYiIiKimiCzsZ2AbG1JzHPDQ+HhYo+P/9NHOKeaPNBWV8eUN5MfT+2j9XxWXinC21XWDpr5+UGhPo2peoUp7qesGWWr7CRiODvaYcnLD+DpB0OE84bOCqrqjYndEOxXWaB45vjOGn0WT+uL5TP7I8DHBQAQ2swbMVERepeqKaMx559S6yBP0y82gM/9QueThoViaI/gGr9/dUsA35vcC+MHtxaOY6Ii1Ar17z2je0c/sgxdX8JbKfnYfSoRa+4XZiciw1T7XuCDDz7AwIEDkZKSghdffBGPPvooAGDRokX46aefMGzYMPz000949913hWtqo42IiIjIHEUlFfjhr6vCcfMATqOvyt5OguUzByDA20U4p/oH8UUtxbONrbEEAAE+LmqFjlWVlFYIj0vvz14SGJjNKFTZfWvamI7Y9ukYk+sSWZsx9ZBUuTjZ4b3JvYXjrm0aafTx83KGh4uR42LC1zs7vxQJyXnCsa7d2GqKj4cT1swdjCHdgqrvbILqFkYE+7lhRJ/mGN67GaY/rlh++PUbg4X2/eeScC+jUMfVZElVk+XK3S5t7UMIIltX7VK4d955B++8847G+datW+OXX37Rek1ttBERERGZ46vfL+DircrZLwtf7GnFaOoO1STAn0dva7SbkGcQNAtww51U9WVEM8d3xtSle7X2N/RPvRnLD5gelI2wk4hRIZVh/tPm1VWqacpyE/F3c9C+hWErC15feag2Q9KqNpNXZSoJz1WvDdTZ74khbYTHVROw//32GB7t2xzjBrWuehlZwv0vR1ZeqXBKKpMJSyhVNxIgourV7scFRERERDZCNalkTYun9bV2CMZR+Xs4V6XwudBsRmapalIJqCYhoJJZksnluJGUq9ZcVFKBCzczTY7HlqycMxBfvzEYjYyo86LN3Ke64SMdSw9NsfX+rnCfbDpr0vVvPdujxmKxFmViqam/m1FFuAN9XdSOdxzRTNSSZSh3ubx+r/JnyP6zSWYt8SRqyJhYIiIiIrIgPy9nfDilN15/squ1QzGI8g8wAEjOLNJoF5szZclAr4ztCEB9xtLuk4n4cN0pfPHreeHcez+cwGc/nxOOp43pUOux1RZ7OzHs7cx/qx7W3BuNfVyq72gC1SWHcrkc+88lofj+UsYjcSm4eCsL5RXqyxnbBNdufSVLKJfKAAAeLsbtMPjGRNuafdaQ5WlJkp+5pn1XRiKqXu3sc0lERERkw96N7GXV5w/0dUWgr2v1HW2Av7f+GTO1nVeKiYrAqatpANTroWw7pJg5c/Z6BrYdSsDIB1ogLbtY7dpeYTW7zTypO3AuGcN7N0NpmRR/HE7AX0fv4Pu/ruCJIW3w857rAIBP789s8vdyRnRdm62nQ6CP4t9uv07G7V7t7W5avSyqedpmWt5O5fI3IlNxxhIRERHVe6qFWJ96sC2a+mvfLpc02VWzK5nYxF3L9FkxewAWvdhTpcC3+nPcSc1HYUllke8tBxJQXKo+M8ahBmb7kKbnh4cKj5XJo7U7LuGvo3c0zqtKyynWOFdX+Xo6ISYqAn06NDb62kkPh6gdZ+eX6uhJtUl1t0Xl74f8onJd3YmoGvyNS0RERPXer3tuCI8fCm9qxUjqHvda3E1t5ZyBGNiliXD8zEOKP7pdnOzRTGXXvqL7S672n0tCZHQsFn13QuNe05fvVzueMa5zbYTc4A3qqr7TWmR0LE5eTbdSNHXP4G5BeH1iV+HYGoXNSX0mpkzHNn9yuRzFpRXIyiuxVFhEdRYTS0RERFTvlVap80KG01afRzUZVHG/3owpnB3t8MIj7TBpWCiGdA9CRHft28P/c/IuACD29D2D792+hbfJcZF+y2f2N/qa+rIMzlwikQgdDNxNj2qP6kxLXYklqUyO/649hjdWHUaRygxJc2Tnl2Lb4Vtqy3qJ6gPWWCIiIqJ6z9Ol9mbd1Hfadml75qEQ7D+XBAAoLTM/aTekm/aEkqnen9zLrN3qSD8PPf+ehvYIRocWPvjiN0VR9W2fjkF6OmvXVOVoL0FpORPetkBXjmfq0r3C4+nL9+OVsR0R3s7frOf6+KdTyMgtgZ+nk0lLKYlsFWcsERERUZ237fAt3E7R/cfr7/e3SKeaoVqfpKCk9uuSPDGkjcF9XxnbEUF+rKFlDS0DPTCyb3N0aeMLQPtsN1Lo2raRtUNo0FSTSS8v22fQNat+j8Pu+7MnTZWRq1hWt2bbJRSWlOP45VS8+tk+bPr3mln3JbI2JpaIiIioTsstLMOW/Tfx/g8nq+07Y1wnC0RU/6kuIyksrpklIvo0Vam3BADPDassIL1yzkChNpNELDJ7RgGZ5otZA/Df58Ph6eYIkUiE+U93Q9Qz3a0dls0a2iMYANCNCSarqLoULTW7yKDrNuyuTADF3czEL3s1C9XLZHIkJOehvEKG4lLdPx9nLD+A1VsvorhUil0n7mLv2Xs4dinVwFdAZFu4FI6IiIjqDJlcjpNX0hDezh9ikQjFpRWY8+VBoa2qf08lYv0/8cJxe9Y2McnjA1th8/6bWts8a7G4t67nGNwtCOt2XgWgqNM0tEcwwtv5w93ZvtZjIXUikWL2h1uVsQ9txhpX+rQJ8oSrkx183J2sHUqDVPW3xZtfHzX6Hst+PgcAmDBYfUblW2uOqu2CWLm7pX7r/lb8TOvdPsDoWIisjYklIiIiqjP2n0tSvPneehExURH4dd8Nvf1Vk0oAt6A31cgHWmgklloGuiMhOR/e7o4WieG9yb1QWFyOpv7uWtstkeCiSlNGtYejvQTtmnkhK6/U2uHUWXKNFAdZhJ5hH9o9GP+eTjT4VmnZRfDzU/xcKiguV0sqAcDNpDy0auKBO6mG1RqLjI41OBlFZCuYWCIiIqI6IymjUHj8694b2FNll7CPfjqFt57tofN6FnSuOd1D/JCQnA8PCyV0glk3yab0VSk87OLEmWKmEIlETCtZib6EXu8OAXoTS9fv5aJNkKdwHPX1UfTu0BjHLqZo7f/BupP4dt4QLPruhMHxRUbHAgD++3w4WgZ6GHwdkbUwsURERER1RmJagfD4z6O3NdqvJ+Yip6AU73xzDIsie1oytAZnRJ/m6N8pEJ5ulpmxRFTfKJcRkhXoGfc2QZ54e1IP/LjzKu6o/M5R+ujHUxrndCWVlF5askfr+f6dAtEzzB+uTvbIyivBqt/j1Nrf/+EkZy9RncD54ERERFRnODtW/5nYaysOoai0AvO+OqJ2/u1JumcyUfUWvqCeqBOJREwqEZlBxMyS1egadfv7y6VbB3liyqj2AID3InuZ9Bxr5g7Wev6TVx4QHr84oh06tfJFqyYe3HiA6jTOWCIiIqI6o3f7AJy5lmH0ddPGdEBrlaULZLzmjbXXNrKWGeM6oULKP8qp7hIBkPFb2Cqq7gqn9OpjHYXHQX5uJs0WWjF7IFycdP+Z7ePhhI+m9kEjTyeN5dnNA9xx24BaTHmFZcgvKkMQlwiTjWBiiYiIqAE4dTUdwX6uCPBxsXYoZjHlw/1+nRqjVxh32alvurX1s3YIROYRAXrXZFGt0fW7pHPrRmbfW19SaeWcgQCAxjp+Fy98sSfKyqX4X+x17DlzT2sfAJj/9RGUlkkBAGvnD2H9QLI6LoUjIiJqAFZuuYC31hi/nbKtkcpkGuc+m9Efa+cP0XnN5Efb12ZIDYpEzD9eiGqKWCTiSjgb0jpId5HsL2cPwBezBlR7jxdHtFM7jomKwGfT+6FP+wB89dogg5ZzO9hLMGlYqHB8Oj5do48yqQQALy3WXr+JyJI4Y4mIiKiBqA9/v2TklGic07XNfOSIMLRtyuVvNenL2QP4hzBRDeK/J+vQNuxvTwrX2d9Vz86HMx7vhOScEozo1VRru6ebI6aO7mBsiIIVmy/oXZLHbyGyBZyxRERERHXG7wcTDO7bv3MgArzr9tI/W+PkYGfQJ+5EVD2RSP+291R7lDWWHhvQ0qjrYqIisGx6P3yrMku2W4gfnq+FmbHtW3gDAB7o2FijrZ+Wc0TWxMQSERFRHbd+VzxSs4usHYZVqJaViImKwHuRvdCqiQc+mtrHekERERlABBGnm1iJcqZYj1DFTmyueuoiVeXl5gixSITJj4Zh1vjOtREeAMWsW0CxFK6guFytTSIRw0PHbF0ia+BHTkRERHXYpVtZ+Pd0Iv49nWjS7jV1yc2kPI1zH1dJIAX7u+Gd53QvZyAishUiEXeFsxblTDGRCPhwSm+4uxifpOnXKbCmw1JjZ6eYA1JSJsXMzw8AANoGe+JaYq7Oa87Ep6NVEw94ujnWamxEVTGxREREVAfJZHK8vvIQcgvLhHMZucVo5Ols0v3kcjlOXU1H17aNYCexzQnNH6w7KTwOa+6Ny7ez4WgvsWJERETmYmbJGpQzlsQikc3ulqptqwRdSaXs/FK4Otnhy80XAADOjnYY3qsp+nduAkd7id6d6ohqAr/DiIiI6qCSMqlaUgkA5n11BAO7NIGvpxNGPdBCOH/5dna19zt3IxOrfo/D6H4tMHZAq5oO12yyKhVuZ4zrhISkPH4qS0R1Vn3ZFa6opALlFdI69fNYWWNJZMMbXVZIDf/meH3lIbXj4tIKbDmQgC0HFHUJp45qjz4dGkMmk+PghWT069QYErFtfohEdRMTS0RERHVQUUm51vP7zyUBAJwcJNi4+xqaB7jjdmp+tff74tfzAIA/Dt3CH4du2dyyOtWtlQFFEemwFj5WioaIyHxpOcVIyynGQz2bomWg7q3ubV3U10dQUFxuc783dJHJ5Dh6MVVxYMOZJQ9X3TvRAcALj7TD939dMehea7ZdQp8OjbH/XBLW7byKopIKDO/drCbCJALAxBIREVGd9M32S3rbN+6+BgAGJZW0ySsss5nCoLtP3sWG+68HAL6dN0RPbyKiuuX9H07WmaSMNlULS9u6f08lIi4hCwBQUlph5Wh0U51R9NSDbeHh4oAtB27i6QfbolMrX4hEIoMTSwAQGR0rLB/feiiBiSWqUUwsERER1UH6infWhNlfHrSJP3TOxKerJZUAQCy23U+YiYhMIZfLkVNQBm93215OlppVhDfXHMWkYaEQARjUtYm1QzJaYnqB8DivqExPT+uLeqY7fD2c4OvpBADo3T5AZ99+HRvjUFwKAOCTVx6Ak4MdXJzsEBkdK/QpLVfM/q06C5jIXEwsERERNSByuRwikQjZ+aV4Zdk+lJRJEeCtWfDbw9VB6GtNykKkSs0bu1spEiKi2rPjyG1s3n8T0dP6wt/LtE0YLOHNNUcBAD/uvAoAKK+QWTMckyQkV+4w2rqJpxUjqV5IUy+D+r09qQdaB3kKiSUfDyeTnm/D7njcSy/E3Ke6mXQ9NVxMLBEREdVxvdsHoF+nxtiyP0HtDbM2UpkcErF6oc/U7GKNfnmFZZi8eA/6dWyMySPb13jMhvJ0c0BuQeUnys8+FGK1WIiIasu56xkAgNyCUptOLFW18d9r1XeyMZl5JcJjZ8e6/efwmrmDkZpdjKBGrjr7THo4BD/uitd7n5KyCryybH9Nh0cNSN3+l0RERNTAKT+lBID07GKNxFLPdv64mZQnvJEuK5fi31OJOu8XExWhNm3+UFyKVRNLjw9she/+vIL+nQPx1NC2df6PACIibZT18OztuFNXbSsurT/LwOwkYrWkUvR/+iAzt0Stz+BuQegZFgB7OzFe/nSfcP7L387jzDVFQtOzSk3Fk1fSEN7OvxYjp/qGP7mIiIjqMGVSCVC8eaxa2PrlsR2x9JUHhOPT8Rn4+/hdrfdaO9+2imLL5XIUlygKqz4+sBWTSkRUbym3ls/OK9W566c+2w/fwqjXt2LD7nis+j2upsMT6JsZQ9bn7+2isWOqSCSCm7M9HO0lagW7lUklAMgtVK81ter3OLUPmYiqw8QSERFRPSESiSAWi6CsihQe6qfR51ZKHoq17ILzzbzBVq+nVNXfx+9gU+x1AIpPZYmI6rsvN1/A9OUHjL5u8/6bAIDdJxNx8kpaTYcluJdRCAD4fGZ/jTaZXK5x7sa9XERGx9pkkqIh/l55Ykgbo/pv3F33ljqSdTS8f01ERET13PxnugMAxg1uLZxT7jQUe/qecG7C0LZY9GJPrJwzUG1bY1vxy54bwmPlFslERA1BhdS8otj593c7O38jE6fj02siJABAjxDFBxZuzvbo1rYRJCq7dH647hQqpDJk5BRj8uJY5BWV4cMfT6ldfyslD29/cxTnb2TWWEzGcnO2BwD07xxotRisyUNl2VuAtzPmPtUNfTsodpsb27+lWt9/Tmqf4UxUFeeUExER1TMhTb0QExWhdu65YaH4/NfzaucmPRKGjIwCVNXU3w130xTn2zXzqrU4jcG6I0TUkOQWlAlbzFenqERzFuqsLw4iJioCy385BwCY/GgY+nUyP5Hi6+kEJwcJRCIRZozrDADCbKSE5DxMXbpX6Dv7i4Nq15aUVeC9708CAJb/cg7PPBSCTq184O/tYnZcxlDuZOdo3zB/r3w0pTeyC8rUljX6eznjRlIe+ncOxOj+LTHt070oK5fhEZWlc0T6NMx/TURERAaKjI7Fyipb3n+47iQio2Mh1zLt31aVlmsWK9W19O3dyF5oFuAGwDpLBZIyCvHBupMWf14iIlshh+bvlwqpDF/9HofkzEK189OXa9/NKymjst/aHZeNjuHQhWRERsfiwPkk4VxZuRQOVWaQThjSuuqlWlXddWz9P/GI+vqo0XEp3UnNh1wuxz8n7uL45VQAiqV6Upn+2V5l938f2ts1zJmwLk72GrWyfD2dEP2fvvDxUCQzV8weqGiwrRXyZMOYWCIiIqrGKZVlBDK5HDeSFDuvxd/NMej6zftv1moxVUOkZhWpHQf7uentv+jFXmgd5GGV5Nk73x7DzaTK3e2ip/W1eAxERNakXCpWIZXh+OVUyOVy3EzKw4krafj+rysG3eOdb4+pHU9duhdzvjyoo7e6k1fShGTUd39eEWb5lJbL4FBlBmn3EM16frUt/m4OFn13ApMX78HGf69h9daL+GnXVfz322NYGHNC77XK32pVXwdVspOI4ewoQX6h8YXkqWHivyYiIiIj/Hy/mDSg/mmwPtsP36rVYqqGKJeqJ4jmPd2t2mtEIhFkFsorZeUptkfWVlfE38vZMkEQEVnRe5G9hMcV9xM532y7hNVbL2Ly4j3CBxTXEnNNun+FVIbcwjLIZHJ8/9dljZlPqqp+GPKfT/bi1703cORiCjKqbGcf4O2C6Gl94eGiqF00tEcwYqIisHha31rbbfSElt+pyhqChv5urjrzitQVl0px8EIyvtl2Ue381TvZSEjO03EVNVRMLBERERlAuePOrhOVhSyPXEq1VjhGc3KofAPdK8xfKF6qj1gkssiMpSNxKXhj1WFcupWFncfv1PrzERHZkkkPh+D9l3oj2L9yJumWgwkA1BMoeVW2hAc0d2JbO38IGvvor1n00pI92H8uGW9/o5jRtHH3NRw8n1xtnH8eva2zzd/LGctnDkD0f/rgmYdCAAB+Xs5qS67XzB2MloEeeP3Jrvj6jcHC+cjoWFy+nV3t86u6nZKvtz1q9ZFq7+HQQGssGevIxVQcupCM8zcyUVomxeINZ/D+D1yuTurM/te0Z88ejB07FmPGjMHo0aOxa9cuAEBCQgKefPJJDBs2DE8++SRu3bolXGNqGxERkbVsP3wLpWXqdYpu3DPtU2NDyeVyXEvMqZHkjlRl6lGrJp4GXSMWwSIzlr7ZfgkA8Mmms/ht3021ttWvD6r9AIiIrGhI92CNmjelZVL8feyOzh0xE9MLkJVXgjup6gkWkUiEVx/rKBz/9/lwtSROVZHRsfjn5F3E/Hm52tpEhtBWiPvRvs3h6mQHO4kY/30+HB1a+sBOol68Z+nGM8LjguJyFBSXay1KrnS9mt+/7i7Vf3jCpXCGW7vjMpb/cg4vL9snnIuMjtVIbFLDZdaucHK5HPPmzcP69esREhKCK1eu4KmnnsKDDz6IhQsX4umnn8aYMWOwdetWLFiwAOvWrQMAk9uIiIgspbRcitVVlgLM/eqw2nF176cqpDLEnkoUjnccuYUOLX3g5GBX7SfKALDvbBLW7byKnu384WAvRrtm3ibv6iNTyRC1CTIssSQSiSA38A+N5MxCvP3NMSye1hd+Ri5dC/R1QXJmkcb56Gl9uVSBiBqsn/dc19m2YO1xjXPbPh2D9PR8BPm54b/PhyOokatRP0PvpReiWYC72rmoZ7ojev1ptXNr5g42+J4AMG5Qa4wbpF7gWyQSIdjPFYnplcvWpDIZpizZq9ZPdYdTuVyusenEF7MGICO3GL/uvYFZ4zujoLgCr688JNRC1MehgRbvrkkvLd6jsQstNUxmp2nFYjHy8xWZ8vz8fPj7+yM7OxuXLl3CyJEjAQAjR47EpUuXkJWVhczMTJPaiIiILOliQhbO3S+eqlRQrFnEMiOnWOc9/jl5F5tUajL9tu8m3vv+JN5ao9gF5/yNDI1Pm1Wt23kVgGIpxKELKVi74zIio2OF3W+MEdrUCwAwa3xntGriYdA1l29n41piLvadvYd71dSsUG4hPX/1EbUkVrXPcSsLnVr5apxfO38IaysRUYPT1F//xgq6jB3QUu24ZaCH3qTSrPGdNc4t+u4Ezl7PQF6RYsmdh6sDQpp6ISYqAu9P7oVFL/bEyjkDa2y30Pcm91ZLSuQXaf6O3XHkFgDF79rJi/fgSFyKWrubsz1aNPbAGxO7wd5OAtW8U3W/ixwdmFjSZ/KjYZj7VPX1GIkAMxNLIpEIy5cvxyuvvIIhQ4bg1VdfxeLFi5GcnIyAgABIJIp/rBKJBP7+/khOTja5jYiIyBbN01PHIf5Ojs42uVyO5b+cx6LvTuDSrSx88et5g5e8rd56sfpOVbRr7o2v3xiMLm0aGX3tD38rdtrRJS2nGKXllcsEX1qyR2sS7v0fTmDpxjNYvTUOJWWKJQ5LN51Vq1ulVPVTaSKihmDyo2Faz08Y0lrreaXqlobFREVgyct9hefo0qaR1iVyX/x6Hpt2XwOgXtMpyM8NzQLc4exo1oIXvV5bcUjjnI+7EwDg5v1i0cql07p4uTkKj19asgdp2ZqzYZW4FE6/fp0CEdbcWzj+7/PhCLg/2/rDKb0BAGP7t9R6LTU8Zv1kqKiowNdff41Vq1ahR48eOHXqFGbPno0lS5bUVHwm8fU1LdNP5vHzc6++ExmN42o9HHvLsrXxdk/WXxhUVYkMaBqgGX/VGU+qJi/eIzz+ZNNZAICHlwucHBS/mpMN3NXG3HEz5npdfSOjYzXOzfz8AL5f8DB8PStnHSWojqlIjLnP9lC7ZsUbQ9A0wB1icf1NKtna93lDwrG3Ho69fqrjk16gWaAbAJ4b2RG+Xi7o1yUIkxb9rdH+xIOhGvfS9jzbPh2jdm7GE13h6myP6B9OCOeOqmxMYa2v3eevDcasZXvxzfZLGskk1d852uIL9ndDYloBACDq66Mar1nJ38+9Rl9fff0+3/jBCKRnF6FlE0/06hyk1nY1Mddqr7u+jnddZVZi6fLly0hLS0OPHoo3hj169ICzszMcHR2RmpoKqVQKiUQCqVSKtLQ0BAYGQi6Xm9RmjMzMAqOm4ZP5/PzckZ5u+B9hZBiOq/Vw7C3LFsc7J1f3EreqXlkSq1FjoKxcqqO3bu99cxRznuiCd78/Ue2ON0rmjJux427sc73wnmJDj9WvD9JYknH8UgomvLVD7VxmViFc7Op3UsnWvs8bCo699XDsq6c6PolJ2mcepafno1eoH8pLyjC6Xwv8ceiWWnuwj7PGvQzRrZWPzrZgP1eLfO20vZ7cHN0zjVRpi2/G450wX2U2sa7XUFhQUmOvr75/n7vZi7W+vsu3sqzyuuv7eNsqsVikcxKPWfP/GjdujJSUFNy8qdjB5caNG8jMzETz5s0RFhaG7du3AwC2b9+OsLAw+Pj4wNfX16Q2IiIiS/qqSuFuY529nmH0NRduKmY46UoqrZwzEO9F9hKOv50/xLTgTHQrJQ+R0bGYsmQPfj9w0+Dk2bRP91XfCdDYFYmIqKFp7Fu5scOkh0O09hndryU8DNj1zFxe7o7Vd6oBY/q3RFhzb4x6oAUmDm2LtfOHoMSA3y+6dg3183LGuEGtAADtW3hr7QNwKVxNMaXuI9U/Zs1Y8vPzw6JFizBr1iyhFsJHH30ELy8vLFq0CFFRUVi1ahU8PDywePFi4TpT24iIiGzBkxFt8L9Y3bv1AEBatuEznlQlVVkC9/zwUPRs5w8XJ8UfEcEqhV3FFq5DpCzQLZXJ8cehW/jj0C2MUamv8PLYjth26BYS0wuMvvfciV3r9RI4IiJD2N8vjC0SAQ90DMSPu+Lx3LBQtT5isQjLZw7QugzZVK9P7IpP7y/LVpo8Qnu9p5omEok0ikSrftDw6av9kFdYhrPXMyAWAVsOJOCz6f30Ficf0ac5ftt3U+8uqNx1tGaUlBk/Q9tQOQWlSM0qQmgz3QlCsg1mV18bPXo0Ro8erXG+devW+OWXX7ReY2obERGRtX02oz88XOw1EktSmQwSceWnn3fTDEuuNG/srjZD6Z0qRbIf6NgY9lW2RP701X4GF/o2x9vP9cCH607p7bP1YILwuGc7f/Rs56/xx46y+KeHiz3at/DB1NEdNPu04OxkImpY3ovsBVmVn+V292fRuLs4wNFBYrGt3Du08MGMxzvhy80XhHOebpaZsaSNs6MdZo7rDIgAb3dHeLs7onljRU2dUf2qLxgtEokgEkFjfFU52DGxZI6oZ7ojev1peLk51NpzKAu6fzilN/45mYgJg1vXagF5Mh3n/xERERnB09UBIpEIPUL81M6XVvnE7sSVNOHxm892x9MPttW41wuPtMO8arbyrZpUAhRvsn08nIwJ2yStm3hi/tPGbzWs3CVGuZXz5dvZiLuZibyicq1vCL+YNcC8QImI6qBgf8VOa6rcnO1hJxFh/CD9u8ABQE3P8ezStnLX0Ef7Nq/huxuva9tG6GrCTqZKYpEIMpnudnsuhTOLg71i/KQWqG389jfHsPfMPXyzTf+ugGQ9/NdERERkgpfHdkSHlpWzbMortL97DfR1QdtgLzwY3lTtfExUBAZ2aaL3k7dl0/vVTLBmMHT6+RND2giPR/dvifdf6o2vXqusf7Hs53MAgD1n7mlc6+Zc+7VCiIjqAjuJGGvmDkH/ztVvXvTJq4rfEU9GtKmmp2HEIhFioiIQExWBcQYktmydWCzSO2NJIuHya3MoZ2nfSy9EZHQs/j52p9af05j6lXdS83H0YkotRmM7EpLzINWXRbUAJpaIiIgM5Kky3VssFuG1J7qgxf2p+VU/sGvdxAMA0L3KzCYAGNZLM8lU1RsTu8LLissQjDW8dzO14+oKcX84pTfsJCKsmTu4FqMiIqq/vN0dERMVgWG9mlXfuQFSzFjSnViydJ3C+qagqAwAsHm/YiOvn/forz2pj0wuR05BqbDMvyaSJIu+O4E1ZsxwysorqRM7zSemF+D9H07it303rRoHFygSERFVofoJZ7MAN3i7OeLcjUyN5W8ikQgDuzbBrb+vatxjQJcmuJGUh0Fdm2i0jR9c/SfB7et5zaFAX1esmWvZXe2IiKjhEIv111gi85TqmKltip92XsXes0l4uGdTZOeX4sSVNLUSAvOf7obFG87U2PNVJzu/FG+sOoxHejfDhCE1MyOwphUUl2Pm5wegzI8mJOVZNR7OWCIiIqpC+QlVWHNvvDGxG8YOUGxb3K2t5uwj5eedVYtpK++hWtBbSds5W+bhqlmYU9c22FWFh6qPWXUzmYiIiGqCWCSC3Lqrg+o1uxpcSrj3bBIAYNeJu0KNyg27rwntqsvyF68/bfbz7Tp+B1OW7EFkdKzW2VGFxeUAgPM3Ms1+rtpyOj4dAKB8+3n1bo71ggFnLBEREQEAMnKL4eHiAAd7iVCIsmNLH7g528PN2R5r5w+BSMu0eW3ngMpPSSXiynYRgJ5h/jUffC17ckgbfLNdMZ18/tPdkJ1fij4dGuPHXfHVXvvy2I6YvHgPAKBPhwBMHdWhVmMlIiICgMKSCvx7OhGj+rUQPiApLZdWcxUZyl5inQ/Jrt7NMXtn3E0qO/tOWbIXAPBI72b4q0qdqMKScrOepzaduppe4/eUymQ4dCEF/TsFQiw2LnHIxBIRETV4Mrkc8746AkBR70g520j1l6quBJLybNXp9lIt91hr4LbR/To2NqifpfTpECAkllQ/NRzSPQh7Tt9D3w664xWJRLCTiFAhlQszv4iIiCxl9pcH0T3ED8WlFXgwPNja4dQbdlbcVe/KrWzIysvh7+1Sbd/I6FisnT8Ecuivq1U1qQQAOQVl5oRZq/KKaj62f04k4uc91yGTyTG4W5BR1zKxREREDV5adrHasTIpVGZA/YALCVkAgP3nkvBQeFO8vvIwJg5tg4IixadchhYHXTF7AO6kFmDJxjMY0t223viKRCJMejgEQX5uaueHdFUklh7qqT/eVa8NQlm5FC5O3P2NiIgsT7ls6PLtbCtHUn+0CvTAc8ND4eXmiC9+PW/WvXqF+eP45TStbe4uivcOqjOK5q04AAB4e1IPtA7yrPb+32y7hKOXUtHM3w3Tx3UyK9baIpXJIJXK4WAvMaj/7ZR8o+5fIZWhtFwKVz3vxfafUyxJvJWSB8C4xFLdKvJARERUC6pO972XXgAA2LK/+h02kjMKAQBJGUV4bcUhVEhl+GlXPLYdvgVA8YvcEC5O9mjX3BsxURFodX9HOVsypHswQpp6qZ0L9ndDTFQEWjTWH6+dRMykEhER2Yy5E7taO4Q6TyQSYXDXIPh5Opl9L11JJQBo3USRONJWRPvDH0/h+7+uQCaT48SVNJ1L5I5eSgUA3EkrEGaoA8CauYOxdr76RiILX+ipdrxkw2mzl94ZYtWWOEz7dJ9Z94iMjkVkdCySMws12t755hhmLD+g81qZTI6UrCIAwP5zydh79p5Rz80ZS0RE1OBdvpWldlxarkgGdWvbqNprcwsVU5GVn4ZW5ezIX7VERETW1jLQHQnJilkeYfV851VLUp3NHBkdi+j/9IG9nQR2EhHcXdQ3/7idkg+RCGgW4K52vnNrX5y/kYk5T3RBWnYx1v+jqOG44IVwBPpUbvrh5myPgmL1ukf7zyUJM22U/jO6g3BPEQBdaSG7KnWi3nq2B5o3dkdMVAQio2MBAFfu5GDy4j346vVBcDRwNpEpzlzLMOm62RO6YPkv59TOXbiRiUBf9c1S0nLUZ+cDwB+HErDjyG2UV8jQvLH612Td31fRv1Ogxhjpwne7RETUoBUUl+OHv6+qnfvnhGKq9d20AoOu16eObQBHRERULwzq2gT7zibB080BuQVl+O/zPau/iMwW9fVR4bGHqwOWz+gvHL/7/Qnh8YDOgXhxRBiAyt3XOrXyBQB4ujqgpEyqMSP6zWe74+1vjlUbw9d/XET/zoEAdCeV3ovsJTxeNr0fcgpK9c7AfvnTfYjRUyuzQiqDVCY3O/mUllMMfy9nvX3KVIrQd27tq9Fub0D9K2XiTEnb0rqpS/ciJioCB84lwc/LGe1b6k7IMrFEREQNWnZ+qca5i7cUNRgyckvMvr+uot9ERERUe54f3g7PD2+HkrIKoXYi1Y52zbxw5U6Oxvm8wjLcTMrTusT/wPlkvDgiDGnZRRpt4e2076Dr6eqg9bw2B88n62zzdHNAsH/lTCsvN0d4uTlWe891f19BYx8XZOSVIKiRKwZ1DcLFhCyENvPC1KV7hX5jB7TE6H4tDY5VVZkBOxfmFeov3P3jrnj8uCseox5ogYKSckx6OFRoKy6twO8HEnReO7x3M/ytUshcNQH1/VtDdV7HxBIRETVoR+JS1I5VZyBFPdO92uubBbjhTqrumU2GFu8mIiKimufkwD95a5u+D9E+WHcSMVERWmfEyORypBvxIZ6jg+mzgRp5OmHm+M4IrrIRiS4r5wzEuesZCGvujTkrDgEA9p5VX3L3y54bKCqt0Lj29wMJJieW/jxyG13bNkKvsABUSGW4fDsb3u6OanHHJaiXcJj8aBjW7riscS9lvc89pyvrJb362X6Nfi0au+NWSj7WzB0MO4kYjw9spZYoU5r31WGsfedhrXHzXxkRETVo5VV2fjt2v8AjADRp5Fq1u4ZAX1ediaWubaqv0URERERUlyUk5+ltP3QhWSgMrWrJ+tPwq2bZlyqJWIxnHgrBL3uuI3paX7x2P+Hj6mSHL2YNwD8n7uKvY3eE+pdqz/XyAwY/D6CokdmnQ2O9fbQllZSu3slGi8YeKJfK4OZs+AYmRy+l4uilVIhFIqz6PU44P2NcJzTzd8fFW1lYt1O9hEO/ToHoFRaABWuPITVbs5aSPtqW9xlaV0ntGqOvICIiqk+qfMimLBgJAI721f9i1bZTSHg7f7w8pgOXwREREVG9V1Kmf/nW2h2X8XDPphrn4xNzEZ+YCwAGJ1+G9gjGxOFhSE+vnAG1bHo/iEQiPNyrGW6l5uPoxVQ9d6gd4we3xiO9m2Hy4j0AgMUbzqi1d23TCCIR4OPhhGceCqn2fqpJJQD48rcLGn1mT+giPLa3E+Pj//TVqJ2kj76aUTFREcjOL4VMJsfcrw5Xey8mloiIqEHr0NIH/55K1Npmyic2APDK2I7mhERERERUp70X2QsLYo4Lx/pm9wDAlFHtjX6OT1/th8LictjbVS6R698pUC2xZCcR491I8wq3qyZglv9yTig2DgBPDGmD4b2bCccrZg/E9OWay83OXq/c9e2hnk2rLdBtCG2Fu6On9YUIgJ+Xs5BkWjt/CABg07/X8c/JuwCAhS9UPybe7oq6U+88F44P1p3EBy/10dmXiSUiImrQ7MS6ZxUZMuNIy4QlIiIiogYt2N8N70b2wsL7ySV9xbQBQKzn/Zgu3u6OQvJDqX0LHwzs0gT7zyXh/Zd6I8iAsgbGmD2hC95acxQpWUWYMa4TurX1U2t3cao+xXIvrcDgxNLHU/vAz8sZLy3Zg3bNvJCVX4rHB7ZCr7AArf1V7/vtvCFq4/rUg23RvoU3Pv/1PJr6G1ZrCgBaNfFATFSE3q8RE0tERNSgmZsXCvJzxYkrlcecrUREREQNUZfWvnj18U7Ch266kjoTBrfGL3tvqJ2T1GD5gBceaYcXHmlXY/er6sMpvSGXG54M690+QK2Gp7OjZhrG2VGC5gHuuHInB/06Nsaofi3g6eooFCzXt2xNF23xdWnTyKR7VYeJJSIiarDyCsuw8/id6jvqMbJvC+QXlQvL6Tq28qmJ0IiIiIjqFLFYpFZGQFtiI9jPDY/0aY7uoX5wsJNg/uojqJDKTJqxZC0ikQj68mBPDGmDn/dcB6BIQgX6uuKFR9rh5r1cLN10Vus1UpkcLRp7YN7T1e9IbIuYWCIiogbp/I1MLP/lnM72T1/tZ9B9xGIRQpt64d9TiWjR2J3bGhMREVGDMqxXU+w8fteg5JBUptiNN8DbBQDg6myH3IIyVEhl+i6rU4b3bqZWdwkAHO0lQomFi7ey8Ou+Gxg7oCU6tlTUSZJK5XUquVaVaVVJiYiI6rird7P1tldds69Pi0B3AMAjfZqbFRMRERFRXdMy0AMAINYyjadfx8Zqx1VrC+UWlAEA0nOKayk625GSVQQA2HHkNm4m5WHZ/xQfcO46fgdSmRwXbmbqu9ymMbFEREQNUtWd4N6L7GXyvRp5OiMmKgI92/mbGxYRERFRnSKTKYoqaZtxM3mk+m5vUpl6dcvQpl4AgHbNvGsnOBviaC/Ren5TrGLZ3N20AkuGU6OYWCIiogZJOQVbyU/lEzTltqxEREREpJ+yWLchK7mG9ghWO542tiOeGxaKAB8XHVfUH707KHZya+TpBADo1MrXmuHUKCaWiIioQdL4VEjlzZCoBncmISIiIqrP2jb1BAD079xEa7tEJePUrrn6zCRPVwcM7hZUe8HZELFIhJioCCx5+QEA0Fj6tvT++bqIFUaJiKjBe+e5cNjbidG+hTce7tms+guIiIiICEBlSQBdVJe/1eH61LUiMjpWeOx7fyZTXcQZS0RE1CCpFpNs1cQDYpEIb0zshs6t68+0ZCIiIiJbIpEwBVEf8atKREQNkvJToXGDWlk5EiIiIqKGQdvOcQ3Rf0Z3sHYINYqJJSIiapD+OHQLADCiT3PrBkJERERUj7k6sQJPVd3aNrJ2CDWKiSUiImrQWKibiIiIqPZMGNLG2iHYHAd7ibVDqFFMLBERERERERFRrWjVxAMA0KSRq5UjsS0xURF6i57XJZyTRkREDY7qDhxEREREVHuC/dwwY1wntGvmbe1QqJYwsURERA2KTGXLWyIiIiKqfd3a+lk7BJvW2MfF2iGYhYklIiJqMPafS8L3f12xdhhERERERACAb+YNrvM1P5lYIiKiBqNqUumFR9pZKRIiIiIiIkAirvulr+v+KyAiIjJReKi/tUMgIiIiIqrTmFgiIqIGaXDXJnB2rF9bvRIRERERWRqXwhERUYPw3vcnhMdfvTYIjg5MKhERERERmYuJJSIiqtdkMjmkMhlupeQL55hUIiIiIiKqGWYnlkpLS/HRRx/hyJEjcHR0RNeuXfH+++8jISEBUVFRyMnJgZeXFxYvXowWLVoAgMltRERExnppyR61489n9rdSJERERERE9Y/ZNZaWLl0KR0dH7Ny5E9u2bcOsWbMAAAsXLsTTTz+NnTt34umnn8aCBQuEa0xtIyIiMkZxaYXGOXcXBytEQkRERERUP5mVWCosLMTvv/+OWbNmQSQSAQAaNWqEzMxMXLp0CSNHjgQAjBw5EpcuXUJWVpbJbURERNUpKatAcmahcCyTy9Xau7VtZOmQiIiIiIjqNbOWwt29exdeXl5YsWIFjh07BldXV8yaNQtOTk4ICAiARKKoYSGRSODv74/k5GTI5XKT2nx8fMx8qUREVJ+9vvIQsvNLAQAfTe2Dxj4uWLn5glqfVx7raI3QiIiIiIjqLbMSS1KpFHfv3kX79u0xf/58nDt3DtOmTcPnn39eU/GZxNfXzarP31D5+blbO4R6ieNqPRx7yzJ3vJVJJQB4a81RtbbObRrhw5f7mXX/+orf55bF8bYejr31cOwth2NtPRx7y+J42xazEkuBgYGws7MTlq516dIF3t7ecHJyQmpqKqRSKSQSCaRSKdLS0hAYGAi5XG5SmzEyMwsgk8mr70g1xs/PHenp+dV3JKNwXK2HY29Z+sa7tEyKT38+ixmPdzK5PlJIsCe/nlrw+9yyON7Ww7G3Ho695XCsrYdjb1kcb+sQi0U6J/GYVWPJx8cHvXv3xqFDhwAodnTLzMxEixYtEBYWhu3btwMAtm/fjrCwMPj4+MDX19ekNiIianhuJuVh/T/xuJ6Yi1lfHMTa7ZcQGR1r1IcHfToE4OGeTWsxSiIiIiKihkskl8vNmtpz9+5dvPXWW8jJyYGdnR1mz56NQYMG4caNG4iKikJeXh48PDywePFitGrVCgBMbjMUZyxZHrPGtYPjaj0ce8vSNt5H4lLwzfZLOq/x9XBEZl4pYqIi8NnP53DhZqZGnzVzB8NOYvYGqPUWv88ti+NtPRx76+HYWw7H2no49pbF8bYOfTOWzFoKBwBNmzbFjz/+qHG+devW+OWXX7ReY2obERE1HPqSSgCQmaeoqfTbvhtCUsnLzQGPD2yNmD8v47GBrZhUIiIiIiKqZWYnloiIiKxpx5HbwuOxA1rhgU6NIZXJ0K+TcfX5iIiIiIjIePwol4iIbFJIsKfw+LlhoQZdM6BzIMQiEQZ1DeJsJSIiIiIiC+C7biIisknxibkAgJioCAzuFqTWFhMVgY4tfTTOiUQii8VHRERERERMLBERkR7pOcU4ez0DV+9k42ZSHu6lF1jkeSukMp1tn7zyAABgxrjOwrn3X+pd6zEREREREZEm1lgiIiKdFsYcR0mZVO3cqAda4LGBxu3WaYzSMineX3dS4/ycJ7ogJasIPh5OAAB7OzFioiJQXiGDvR0/JyEiIiIisga+EyciIp2qJpUAYNvhW8JjuVwOmVxeo8/58rJ9SMooBAD0CPETzndq5YuHwptq9GdSiYiIiIjIejhjiYiIjDKgcyBkMjlEImDy4j0AgCA/V7w/2fzlaKXl6omsKaPam31PIiIiIiKqPUwsERHZAJlcDqlUBns7ibVD0WrOE13w2c/nAAAHzifjwPlktfZ76YU18jwvf7pPeNwrzB8O9rY5HkREREREpMD1A0RENuDzX87jP5/sQ3mF7qLVlrb1YILwuFMrX7z+ZNcafw6ZXI7UrCIs/ekkIqNjhfPOjhJMG9Oxxp+PiIiIiIhqFmcsERHZgAs3MwEAa3dcskhCRVkX6filVPQKC4BIBIhEIqFdLperJZYAoENLH733jIyOxdj+LRHk5wpXJ3ss2XgGnVr5Cq8NAKKn9YW/lzMA4Ex8Or7cfEHrvVbOGWTS6yIiIiIiIstiYomIyAI2/XsNIhHwZERbvf2OX07DtDG1E8MfBxNQVFqB7iF+iF5/Wji/Ztsl4fHnM/vD3cUB73x7TDgnUrnH7AldsPyXc+jS2hcP92qGxPQCbNx9TWj/vUoySjWpBABRq49g7fwhEIlEWpNKox5ogbEDWpr6EomIiIiIyMKYWCIiMtLk6FjIAcRERRh8za4TdwEA/ToG4v11J/HymI7o2raR1r5FJRVwcaqZH89SmQxTluzVGos2s744qHFurcrr7NzaV+11O5iwI5uy4LeqPz4ZjYyMAqPvRURERERE1sXEEhGRkeTK/8vlasvHVKXlFMPTxQGODhJIZZV1kxbEHAcAfPHbeXw7fwiu3c3B4g1n1K49HZ+O/p0DTY5vy/6b2Hb4lsnXq3rxkXZ6228m5else3tSD/h4OMHJQYKFMcfh5GCHxHT15NGzD4cgonuwznEkIiIiIiLbxsQSEZGJZHI5JDoSIlGrjwAAPFzskVdUrrXPS1pm7gBAzJ+XDUosxd/NQfT603huWCgGdwsCAKzacgEnr6YbEj4Gd22C54a3Q3mFDAXF5Xh95SG1duWSNX0kEs12bTO5lrz8ADJyijHv/rgoRXQPNihWIiIiIiKyTUwsEREZQSaTqzwGJFpWgikLYwPQmVSq6uWxHfHV73FGxaKsk7Ru51Ws23kVMVERWpNKY/q3xJj+irpF0z7Zi/+M7oBuIX5Cu72dGN7ujnjmoRCk5xRjUNcmCPR1NSgGv/uFuMUiEb6dP0Rv30ZezvjklQeQXVCKnPxSdGvrp7c/ERERERHZPiaWiIiMcPlOtvBYNcmkqqJCpvW8Lr3C/NGznT/KR4bh2+2XAQB5hWVwc7HH1dvZ+GHnVaRlFwv9X3ikHUrKpBr3iYyOFR53bOWDFx8Jg7OjBE4OlT/qV78xWGccQ3sYP3uoY0sfTB3VHj1C/Q3q7+PhBB8PJ6Ofh4iIiIiIbBMTS0REBjoSl4KkzELh+JvtlzBtTAfYScSokMpw+XY2mvi6IqewVO99Fr3YE4u+O4GQpl6Ieqa7cP6BjoH4aVc8SsqkmP2lZhFtpe//ulJtrI/0agZvd0cDXpV5RCIR+nRoXOvPQ0REREREtomJJSIiA32z/ZLa8en4dExduhezJ3TG8l/O67zupZFhaN7YA2IRcOxSKpoFuOvcUU7bTCR9Vr02EK8s269xPqyFj1H3ISIiIiIiMoXx+0QTETVAqnWTqtKXVHrz2e54oGMgghq5ItDXFWMHtDL6uT1c7LHwhZ4AFLOdAKBnO3/EREXAycFOOAcA857qhtee6GL0cxAREREREZmCM5aIiAwQdzPLpOsa+7gY1f/hnk2x68RdtXPLZw4AULnbWtXZTk393dCnQwAGdw1CSFMvk+IkIiIiIiIyBRNLREQGuH4vt9o+K+cMhLOjeT9Wxw9ujdCmXlj9x0WUV8iwtpqd1gBFnaOpozqY9bxERERERESmYGKJiEiPU1fTkJJVhOYB7hptAzoH4sD5ZADAhCGtzU4qAYCdRIxuIX5Y9GJPXE/MhUgkMvueREREREREtYWJJSIiPVZuidM4161tI5y5loHQZl5CYumR3s1r9HkDfRU1mYiIiIiIiGwZi3cTERnpwR7BAICQpl54uGdTtA32tHJERERERERE1sEZS0RERoh6pjtCmnoJBbQnDm1r5YiIiIiIiIishzOWiIiMwF3XiIiIiIiIKjGxRERkIEN2aCMiIiIiImpImFgiIjLAmrmDuUMbERERERFRFayxRESkR+fWvsjMK4GdhHl4IiIiIiKiqviXEhFRNZhUIiIiIiIi0o5/LRER6SGXA2KugCMiIiIiItKKiSUiIj1kcjlrKxEREREREenAxBIRkR5yuRxiJpaIiIiIiIi0YmKJiEgPuRxgXomIiIiIiEg7JpaIiPSQy+VgXomIiIiIiEg7JpaIiPSQycEaS0RERERERDrYWTsAovoqv6gMs744CACIiYqwcjRkLJlMjrJyKeLv5lg7FCIiIiIiIpvFxBJRLSgoLheSSgAQGR3L5JINSc4sRHmFDIu+O4Gv3xgEezsJSsulEAGQA3j5033WDpGIiIiIiKhOqLHE0ooVK/Dll19i27ZtCAkJwdmzZ7FgwQKUlpYiKCgIS5cuha+vLwCY3EZUV8z8/IC1QyAdTsenY8XmC8Lxfz6pPon04ZTetRkSERERERFRnVUjNZYuXryIs2fPIigoCAAgk8kwd+5cLFiwADt37kR4eDg++eQTs9qsqaC4HDF/XkZ5hdTaoVAdEejrIjx2tJegXTMv6wVjYX8fu4Nvtl3SOF9QXI4KqUztXHmFFGXlUsjkcmTnlyIyOlb4r6ikwuxYDscl49z1DCz7+Sxup+Tjj4MJakklfR7u3RzfzBuMmKgIBPq6mh0LERERERFRfWT2jKWysjK89957+PTTT/Hcc88BAOLi4uDo6Ijw8HAAwMSJEzF06FB8/PHHJrdZQnFpBV79bL9w3KdDAEY90AJvf3MMAHDtbg5Ss4sx54kusJOIcTMpF4/2bWGR2Mi6UrKK4OvhCHs7iXBu2+Fb6NLaF80C3AEApeVS2EvEEItFGNojGD/tisdnM/rji1/Pw07ScOrk/7znOgDgyMUUAMDkR8OwdsdloT3qme6IXn+62vtMX74fQ7sH49/TiQY975q5g4VxLiqpwPTl+9Xa425mqR0vfKEnvt1+CZl5JSgpk2LVawNhb6e4/tKtbET0boH09HyDnpuIiIiIiKihMjux9Pnnn2P06NEIDg4WziUnJ6NJkybCsY+PD2QyGXJyckxu8/LyMjgmX183k17Lufh0teOjF1Nx9GKqcJyaXQwA+Oznc8K58Q+Gws3FwaTnq2/8/NytHUKN+2P/DZSWS7HuT0ViZHD3YOxVSXRs2X8Tn84aiNc/36/1+uBATzg6SGBnLzF5fGp7XD+IOYZmjd3xzPAwSMQ1v/uZalIJgEFJJSVDk0oAMHXpXjz7SDs80rclIqNj9fad81R3hHdqgvBOTbS2Nw7wBFA/v6dtGcfbOjjulsXxth6OvfVw7C2HY209HHvL4njbFrMSS2fOnEFcXBzeeOONmoqnRmRmFkAmkxt9XUVZudHX3LyTxWUyUPzDrm+zO/advYcf/r6qdm6vlkSHrqQSAOTmFqG8QgqxCCaNT02Mq0wux7nrGejaphHkANb8cRHHL6ep9Tl2MQW//HtN7Vy7Zl6Y93R3AMDtlHwE+blqzLzKKShFflE5mvq7VZvM0SWsuTduJOXi3Rd7IcDHBXEJmVj2v3Ma/ewkYo2ldFX99NcV/PTXFeH4qQfb4qHwplr7Vjeu9fF72pZxvK2D425ZHG/r4dhbD8fecjjW1sOxtyyOt3WIxSKdk3jMSiydOHECN27cwNChQwEAKSkpmDx5MiZNmoSkpCShX1ZWFsRiMby8vBAYGGhSmylkMjk2/XsNJeVSRI4Iq76/XJGMmvF4J2Tll6JVEw+8/8NJAMDa+UNw4WYm2rfwweXb2SgurcDqrRdNSmBR3VA1qWSIF0e0w5G4FFy5k4OJEW0gFokggvosoC37b2JwtyB4uztqvUd6TjFW/R6HN59RJHXkcjnkcsU/ZEPM++ownB3t0DLQA/vPJVV/gQ5X7uSYnCx6ZWxHhLfzh0wmh0gEiESGz4TydlOMS+fWvpg9oYtB1+iKU1dSiYiIiIiIiGqGWYmlqVOnYurUqcJxREQEVq9ejTZt2uDnn3/GyZMnER4ejk2bNmH48OEAgI4dO6KkpMToNlO8tGSP8Pjg+WR8+mo/4Y/5e+kF+O/a42js44Ipo9qjSSNXXE/MBQDY24kxtIdiaZ/qFvGdWzcCAHRq5YuTVxQzPuTMK9VLJWWahaPdnO1RUFyOFx5ph4FdmkAulyOnoEwjQTSgs/ryqsy8EhSVVKCkrAJvf3MM2fml2Hb4ltAe4O0sLLNUNc2ALe/7dmiMgV0CEdrMGwBw/V4uMnJLAAB30wqqvR4AQpp64aHwYKzcEmdQf30CfFzw/uRewswmQ5NhqoL83PD6xK5oE+Rp8DVfzh6AGcsP4LMZ/fHrnusY078lGnk5G/3cREREREREZByzayxpIxaLsWTJEixcuBClpaUICgrC0qVLzWozxr+nEvHjTs3ZJq+vPKRxLiWrSJiVpJRbWFbtcyhnYJSWc6e4+uhmUp7wOPo/feDv7aLRRyQS6Zx1pCo7vxQA8Moy7UvmtCWVDHXkYgqOXEwREqA7VBJW2ix6sSfcnO0BAG+sOgxAUUwbUE+iyuVyTF5cmZgd3rsZxvRvCTuJCFOW7NW477fzh0BsxKyk6nRo4WNUf1cneyH+ySPb11gcREREREREpF+NJpZiYyuXo3Tv3h3btm3T2s/UNkPtPH5H7djJQYKSMsMTQD1C/art88/JuwCAD388hbXzhxi11IcMJ5fLUVYhg6O9pPrONWTuqkPIzFMkg9o189KaVLKUGU90RYCHI9759pjefkcvpaBtkBfO3cgEADzSpxn6dwrUW//ruWGh6B6i/XtdJBKpJZpUqZ6XyeSQymQ1mlQiIiIiIiKiuqNWZizZkm/nDYFYLMIH606qzUJReue5cOw5k4hDFxRbo38xawCcHKofljH9W2LpxjMAgG+2X8LUUR1qNvAqMnNLMPcrxQyTD6f0hlQmR7Cfabvf1SUHLyTjuz+v4KOpfdDYp3YTPDkFpSgurRCSSgDw+KDWNfocjw9sheG9m0EiFkEOqCVklHWCVBM3ysJ0MVERyMgphq+nk5DElMnkwnLPNX9cUnue0Q+0hKOD/mTc4G5BZr8esVgEsdhyST8iIiIiIiKyLfU6saT6B/qbz3ZHalYxGvu44KutcTh1NR1fzh4AVyd7tGrSHpMfNW75TFhzb+Fx0/sJHrlcjuJSKVycdA/riStpuJ6YizH9W8DFyd7g59uosmPX298oZq9MGdUefTs0NiruquRyuU3PtjpwLhkA8Naaozr7TBzaFievpGHMgJbo0MIHP/x9BQnJefByc8TMcZ016vzI5XIkpheiqb/i65aUUYgAH2e8tkJzqaQxdX50mTW+M/46dkdYcqakbdQlemoSVa0ZJBaLMGNcJ3z52wW182P6V59UIiIiIiIiIqoJIrm8/pWfnvzBLkwY3MagJW3myMgtxryvjiCkqRcKisuRlFEIAIgcEYYubXxx6mo6BnQJBABIxIpixtp2r3o3shcWxhzHsw+HoF/HQGw7fAt/Hr0ttPfr2BiH4lI0rps0LBRDtMw6KSoph5OjHSAH5JALz11VQnIe3v/hJGY83gnddCyJMlRtbflo6q5kNUHXUrDaUlouhVgE2NtVJoUMGdfDcckoKK7A0B5BOr/WZDxuY2pZHG/r4LhbFsfbejj21sOxtxyOtfVw7C2L420dYrEIvr7aV03VyxlLC17oCRfH2n9pyj/i4+/mqJ2P+fOy8Hjd/SLi/TsHInJEmNb7LIw5DgD4aVc8ftoVr9GumlT6cvYAlJXL8PrKQ1pnvHz28zlcuJmpdm7muM64dDsLu08man3+Lzdf0DhnK3WjnhseinV/K8bQy80BOQXVF1avCf99Ptwiz6PK1DpSD3QMrOFIiIiIiIiIiAxTLxNLbs72kMlsayLWwfPJkKvE9EifZvjr6B09V2iKHBEGVyd7yOXlABRJq3Vadr+r6ovfzhsXLICXP92H1W8MNvo6VcqizuYkqOzuJ++WTOurd/v48goZvvj1HBKS8zH98U5o19wbyZmFwrJBXWaO74xOrXyEJGFmbgncXezhYMFi4URERERERER1Vb1MLFmKl5sDACC0qReaBrihb4fGeP+Hk7C3E+PFEe3Qrpk3nB3s8PKyfQDUZx5NGNwGXds0wsc/nQag2Aa+tFyKlKwiHDyfjHlPdxOSHcWlFXBWmYGlrw6P0odTelebVAGAJS/3xQfrTiGvUH0mUFmFTOc1t1LycDslH4fiUvDK2I7wcnNEhVSGPWfu4cedV/Hf58Nx9loGth2+BUCREOsV5o+07GKUS2UoKq1AY28XXL+Xiw4tfVBSVoFGntqTRofjFDWWikor9L4OezsxXp/YTe1coK+r0cvZfD2djOpPRERERERE1JAxsWSGqluyF5dWwMlBglce64iOLX11Xte3QwAAoG2wF1bMHqhW7LttsBcGdG6i1t+5yrI+RwcJAn1dkJxZhD7tA+DsZIfyChkOnk8W+iiTKnmFZRCLRXBz1l0ofPmM/gCAsnIpjl9OE5byKesbfTilN/IKy+BgL8H7P5xUu1ZbweuqfWL+vKy2PFCbr98YBKlMDicHO6GgeGmZFFfu5Nx/PbW7IxwRERERERERGY+JpRrk7GiHVa8N0tk+fnBrjOjTXO2cvh3kdBGLRPhwSh+N8y8+0g7LfzmPoT0qC3p7uDoYfF8Hewn6dw7USAIZMvPJXP/5ZJ/edtWC1kRERERERERkG5hYsgBL7S4mEokw54kuZt/n46l98Oaao3r7xERFYPP+G9h+uHL3OjuJCGvmDoFMLofofjxKH/10CpEjwmAnFsHX00loW/7LOZy/kVn19mpmTzD/NRERERERERFRzWNiiTQE+LggJioCt1PyYWcnxn+/VcxYip7WFx4u9nByUHzbPD6wNR4f2BqA+paPYi3Fut96tofW55o5vjNKSitQIZXjYkIWencIgAjA8ctpOH45FZOGhcLLzbEWXiURERERERERmYuJJdKpeWN3AMC384doTRbVBLFIBBcnRf2nvh0bC+d7tw9A7/YBtfKcRERERERERFQzxNYOgGxfbSWViIiIiIiIiKhuY2KJiIiIiIiIiIhMwsQSERERERERERGZhIklIiIiIiIiIiIyCRNLRERERERERERkEiaWiIiIiIiIiIjIJEwsERERERERERGRSZhYIiIiIiIiIiIikzCxREREREREREREJmFiiYiIiIiIiIiITGJn7QBqg1gssnYIDRLHvXZwXK2HY29ZHG/r4LhbFsfbejj21sOxtxyOtfVw7C2L4215+sZcJJfL5RaMhYiIiIiIiIiI6gkuhSMiIiIiIiIiIpMwsURERERERERERCZhYomIiIiIiIiIiEzCxBIREREREREREZmEiSUiIiIiIiIiIjIJE0tERERERERERGQSJpaIiIiIiIiIiMgkTCwREREREREREZFJ6mRiKSIiAvHx8dYOo0GJiIjA8OHDMWbMGIwZMwYfffSRzr6bN2/GzJkzLRhd3RQREYH+/ftDKpUK5zZv3ozQ0FD89NNPVoysYcjNzUXnzp3xwQcfWDuUeovf47aBvzOtw5Bx59emZvDnufX89ddfGDt2LMaMGYPhw4fj9ddft3ZI9VpZWRmio6Px4IMPYvjw4Rg7dix2796t95rExET873//s1CE9VNERARGjhwJmUymdo4/v2ue8m/O0aNH46GHHsLLL7+M06dPWzssMoCdtQOguuOLL75ASEiItcOoV/z9/XHw4EEMGjQIALBlyxZ06NDBqHtUVFTAzo7/lI21fft2dOnSBTt27MC8efPg4OBg8LUymQwikQgikagWI6wfauJ7nIhIH3N+npPp0tLS8O6772LLli0IDAyEXC7H5cuXrR1WvbZo0SIUFRVhx44dcHR0RHx8PF566SV4enqiZ8+eWq+5d+8e/ve//+HJJ5+0cLT1S1FREbZu3YrHHnvM2qHUe6p/c+7atQtTp07F2rVr0aVLFytHRvrUyRlLSjExMRg3bhzGjh2LJ598Uu2XWWhoKFavXo1x48Zh6NCh2LlzpxUjrZ+2bNmCCRMm4PHHH8dzzz2HmzdvCm35+fmYNm0aRowYgeeeew6pqalWjNR2PfbYY9i8eTMA4O7duygqKhJ+kB45cgRPPvkkxo4di1GjRmHHjh3CdZMmTcKHH36IJ554Ai+//LJVYq/rfvvtN7zyyisIDQ3Fv//+CwD48ssvMWvWLDz33HMYPnw4ZsyYgfz8fKFt5syZiIyMxIgRI5CXl2fN8OsMU77Hz58/j5EjR6rdZ/To0fzEykxVP11VPY6IiMDnn3+OJ598EhEREZxRVoP0jTvVDG0/z6OiotS+j1WPU1NT8fzzz+PRRx/FtGnTMG3aNH7PmyAjIwN2dnbw8vICAIhEIrRv3x4AcO7cOUyaNAmPP/44Hn/8cezduxeAYvZM7969ER0djVGjRmHUqFE4efKklV5B3XLv3j389ddfWLRoERwdHQEAISEhmDZtGlasWAEA+PrrrzFq1CiMHj0aEydOhEwmw3vvvYcbN25gzJgxXFFghunTp2PFihUoKytTO3/79m08//zzGDVqFB577DHs378fALBq1Sq1FR7Z2dno3bs3ioqKLBp3Xffwww9j4sSJWLt2LcrKyrB48WKMHz8eo0ePxty5c1FYWAhA8bfnm2++KXz/v/fee1aOvOGp09Mcxo4di8jISADA4cOHsXDhQvz8889Cu5ubG3777TecOnUKs2fPxrBhw6wVar0wc+ZM4RfZQw89hPPnz2P9+vVwcHDAvn378NZbb2HTpk0AgFOnTuH3339Hq1atsGLFCnz44Yf44osvrBm+TerVqxc2bNiA3NxcbNmyBWPHjsXFixcBAO3bt8eGDRsgkUiQkZGBxx9/HP3794enpycAxR/pGzZs4GwlE1y5cgU5OTno06cP0tPT8dtvv+GRRx4BUPm926hRI7z55ptYtWoV5s+fD0CR8Ni8eTN8fHysGX6dYsr3eOfOneHi4oLjx4+jV69eOHnyJMRiMbp3727lV1O/lZSU4H//+x8SExOFN8iurq7WDotIL30/z3X54IMP0Lt3b7zyyiu4d+8eRo0ahf79+1so4vqjXbt26Ny5MwYPHozevXuje/fuGDNmDCQSCRYuXIg1a9bA398faWlpGD9+PLZv3w4AyMnJQbt27RAVFYVjx47htddew+7duznTrBrx8fFo1qyZkMhT6tq1Kz7//HNs2bIFsbGx2LhxI9zc3JCdnQ2xWIwFCxZg8eLFwoc8ZJqOHTuiQ4cO2LhxI55//nnh/BtvvIEnnngCEyZMwPXr1/HMM88IS0SfeOIJzJs3D3Z2dti+fTsiIiLg4uJixVdRN3Xp0gWxsbH49ttv4e7ujl9//RUAsHTpUqxZswZz5szBRx99BBcXF2zduhVisRhZWVlWjrrhqdN/kcbFxeHrr79Gbm4uRCIRbt26pdY+YsQIAIofuGlpaSgtLRUSI2Q81WmJS5YswZUrVzBhwgQAgFwuV5vB0aNHD7Rq1QoAMGHCBIwaNcryAdcBIpEIjzzyCHbs2IEdO3Zg06ZNwh/dWVlZeOutt3D79m1IJBLk5uYiISEBXbt2BQCMGjWKSSUT/frrrxgzZgxEIhEefvhhfPDBB8KsusGDB6NRo0YAgPHjx6vV7Bg4cCCTSkYy9Xt80qRJ2LBhA3r16oX169fjmWeesfIrqf+UvzODg4Ph4eGBlJQUtG7d2spREemn7+e5LseOHcM777wDAAgKCkLfvn0tEWq9IxaLsWrVKsTHx+PEiRPYvXs31q5di3nz5iExMRFTpkwR+opEIty+fRve3t6wt7fH6NGjAQC9e/eGk5MTbt68iXbt2lnrpdQJcrlcb/uePXvw1FNPwc3NDQDg7e1tibAalNmzZ+O5557D+PHjAUBY/jlu3DgAQJs2bRAWFoazZ88iIiICbdq0wb59+zB06FBs2bIFb775pjXDr7OU3/uxsbEoKCgQViKVlZUJPzf27NmDzZs3QyxWLMji+3XLq7N/lcpkMsyaNQs//fQTOnTogNTUVAwcOFCtjzKJJJFIAChq0TCxVDPkcjnGjRuHWbNmWTuUOu+xxx7DhAkT0LNnT7U3AYsWLUJERARWrFgBkUiEYcOGobS0VGjnJx6mKSsrw/bt2+Hg4ICtW7cCAMrLyw36JI+zN0xjyvf48OHDsWzZMly6dAnHjh3Tu2EAGUYikagVHlX9eQJA7fejRCJRK7pOpqtu3Ml0+n6ec9wtJyQkBCEhIXjmmWcwYsQIyOVyhIaGYv369Rp9ExMTrRBh/RASEoI7d+4gJydHbdbS2bNnERoaar3AGpBWrVph0KBB+O677wzq/9hjj+H3339HcHAw8vPzER4eXssR1k8XLlxA27ZtkZiYiIULF/LDABtVp2ssVVRUIDAwEACwYcMGK0fTsERERGDr1q1ISUkBAEilUsTFxQntp0+fFmaQ/fbbb+jTp481wqwTmjZtijlz5uCVV15RO5+fn4+goCCIRCIcOnQIt2/ftlKE9cu///6Lli1bYv/+/YiNjUVsbCxiYmKwZcsWAMDevXuF6bObN2/m924NMOV73N7eHuPGjcPLL7+MUaNGwdnZ2dJh1zvNmjXDhQsXACjqW2VkZFg5ooaB41579P08b968uTDuaWlpOHbsmHBdr169hJ/5ycnJOHr0qFXir+tSU1Nx5swZ4TglJQVZWVlo06YNbt++rTau58+fF2YdlJeXY9u2bQCAkydPoqSkRJjlTroFBwdj+PDhWLRokZAojY+Px+rVqzF9+nQMGTIEGzduREFBAQBFTR9AURpEeY7MN2PGDGzYsAGFhYUQiUQICwsTfp7cuHEDV65cEVYXPPzwwzhx4gS+++47PPbYY9z0xQS7d+/Gxo0bERkZiYiICHz//fcoKSkBABQUFODGjRsAgCFDhmDt2rXCzxkuhbO8OjljqaKiAs7Ozpg5cybGjx8PLy8v1k+ysJ49e2L27Nl4+eWXIZVKUV5ejuHDh6Njx44AgO7du2Px4sW4ffs2GjVqhKVLl1o5YtumbaeO119/He+++y6+/PJLdOrUiZ9G1ZDffvtNY2lmt27dIJPJcPz4cYSHh2POnDlITU1FmzZtEBUVZaVI6xdTvscnTJiAFStW4KmnnrJUmPWScrburFmzhALGffr0QZMmTawdWr3Gca99+n6ed+3aFQcOHMCIESPQokULdO7cWejz9ttvY968edi2bRuCg4PRuXNnYfkQGa6iogJffvkl7t27BycnJ8hkMsyePRvt27fHqlWrsHTpUnz00UcoLy9H06ZNsXr1agCAl5cXrly5gm+//RYAsGzZMtZXMtDChQuxbNkyjBgxAvb29nB0dMTbb7+NXr16QS6XIzU1FU8++STs7Ozg4uKC9evXIzQ0FC1btsTIkSPRqlUr1jw1U+PGjTFmzBjExMQAAD755BMsWLAA33//Pezs7LBkyRJhGZazszOGDh2KzZs3CxsLUPVmzpwJBwcHFBcXo3Xr1lizZg26dOmC9u3bY8WKFRg/frywO/P06dPRunVrvPnmm/joo48wcuRISCQS9OrVS1jyTJYhkle3YNfGpKWl4ZFHHsGhQ4fg5ORk7XCIqB758ssvUVRUJBTrJuvaunUrduzYgTVr1lg7lDqLvzOtg+Nu20pKSmBnZwc7OzuhsPT333/PWTMWkJiYiHHjxqnNICMiorqvTs1YWrduHTZs2ID58+fzjRoRUT02efJk3LlzB1999ZW1Q6mz+DvTOjjutu/WrVuYP38+5HI5KioqMH36dCaViIiIzFDnZiwREREREREREZFtqNPFu4mIiIiIiIiIyHpsOrGUnZ2NKVOmYNiwYRg1ahSmT58uVHg/e/YsRo8ejWHDhiEyMhKZmZnCda+//jr69++P0NBQFBYWqt1TWejx0UcfxbRp05CTk2PJl0REREREREREVG/YdGJJJBLhpZdews6dO7Ft2zY0bdoUn3zyCWQyGebOnYsFCxZg586dCA8PxyeffCJcN378eGzdulXjfjdu3MDy5cvxww8/YMeOHejcuTOWLVtmyZdERERERERERFRv2HRiycvLC7179xaOu3btiqSkJMTFxcHR0RHh4eEAgIkTJ+Lvv/8W+vXt2xe+vr4a94uPj0dYWJiwBeSgQYOwbdu2Wn4VRERERERERET1k00nllTJZDJs3LgRERERSE5ORpMmTYQ2Hx8fyGSyape1tWvXDhcuXMDdu3chl8uxfft2FBUVcTkcEREREREREZEJ6kxi6f3334eLiwueffZZk+/RsmVLvPPOO5gzZw6eeOIJeHp6AgDs7OxqKkwiIiIiIiIiogajTmRUFi9ejNu3b2P16tUQi8UIDAxEUlKS0J6VlQWxWAwvL69q7/Xoo4/i0UcfBQCcP38eGzZsgJubW22FTkRERERERERUb9n8jKVly5YhLi4OK1euhIODAwCgY8eOKCkpwcmTJwEAmzZtwvDhww26X3p6OgCgtLQUX3zxBSIjI2sncCIiIiIiIiKiek4kl8vl1g5Cl2vXrmHkyJFo0aIFnJycAADBwcFYuXIlTp8+jYULF6K0tBRBQUFYunQpGjVqBACYPn06zp8/j9TUVPj7+yMkJARr164FALz00ktISkpCeXk5RowYgVmzZkEstvn8GhERERERERGRzbHpxBIREREREREREdkuTtUhIiIiIiIiIiKTMLFEREREREREREQmYWKJiIiIiIiIiIhMwsQSERERERERERGZhIklIiIiIiIiIiIyCRNLRERERERERERkEiaWiIiIiIiIiIjIJEwsERERERERERGRSf4PCsgc0eXoN0AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# TODO(Grisha): replace with a function that does the plotting.\n",
    "ccxt_data_filtered[config[\"data\"][\"close_price_col_name\"]].plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e70c561a",
   "metadata": {},
   "source": [
    "## Plot timeseries distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b395189f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:27.895706Z",
     "start_time": "2022-01-14T09:58:27.402910Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/venv/lib/python3.8/site-packages/seaborn/_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# TODO(Grisha): fix the function behavior in #204.\n",
    "coplotti.plot_timeseries_distribution(\n",
    "    ccxt_data_filtered[config[\"data\"][\"close_price_col_name\"]],\n",
    "    datetime_types=[\"hour\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0efaed41",
   "metadata": {},
   "source": [
    "## NaN statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "427be7c5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:29.259651Z",
     "start_time": "2022-01-14T09:58:28.720495Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index in [2019-01-01 00:00:00-05:00, 2019-12-31 23:59:00-05:00]\n",
      "num_rows=525,600\n",
      "data=\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-01-01 00:00:00-05:00</th>\n",
       "      <td>3697.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-01 00:01:00-05:00</th>\n",
       "      <td>3697.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-31 23:58:00-05:00</th>\n",
       "      <td>7216.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-31 23:59:00-05:00</th>\n",
       "      <td>7217.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_days=365\n",
      "num_weekdays=261\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_rows</th>\n",
       "      <th>zeros [%]</th>\n",
       "      <th>nans [%]</th>\n",
       "      <th>infs [%]</th>\n",
       "      <th>valid [%]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>close</th>\n",
       "      <td>525600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.0</td>\n",
       "      <td>99.65</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>num_rows</th>\n",
       "      <th>zeros [%]</th>\n",
       "      <th>nans [%]</th>\n",
       "      <th>infs [%]</th>\n",
       "      <th>valid [%]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>close</th>\n",
       "      <td>525600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.0</td>\n",
       "      <td>99.65</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       num_rows zeros [%] nans [%] infs [%] valid [%]\n",
       "close    525600       0.0     0.35      0.0     99.65"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nan_stats_df = coexplor.report_zero_nan_inf_stats(ccxt_data_filtered)\n",
    "nan_stats_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "f471b2f9",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:31.817101Z",
     "start_time": "2022-01-14T09:58:31.541091Z"
    }
   },
   "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>nan_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-05-15 00:00:00-04:00</th>\n",
       "      <td>540.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-08-15 00:00:00-04:00</th>\n",
       "      <td>360.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-12 00:00:00-04:00</th>\n",
       "      <td>240.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-12 00:00:00-05:00</th>\n",
       "      <td>142.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-06-07 00:00:00-04:00</th>\n",
       "      <td>133.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-11 00:00:00-04:00</th>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-08-14 00:00:00-04:00</th>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-24 00:00:00-05:00</th>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-05-14 00:00:00-04:00</th>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-13 00:00:00-05:00</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           nan_count\n",
       "2019-05-15 00:00:00-04:00      540.0\n",
       "2019-08-15 00:00:00-04:00      360.0\n",
       "2019-03-12 00:00:00-04:00      240.0\n",
       "2019-11-12 00:00:00-05:00      142.0\n",
       "2019-06-07 00:00:00-04:00      133.0\n",
       "2019-03-11 00:00:00-04:00      120.0\n",
       "2019-08-14 00:00:00-04:00      120.0\n",
       "2019-11-24 00:00:00-05:00      120.0\n",
       "2019-05-14 00:00:00-04:00       60.0\n",
       "2019-11-13 00:00:00-05:00        3.0"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# TODO(Grisha): pretify the function: add assertions, logging.\n",
    "# TODO(Grisha): add support for zeros, infinities.\n",
    "# TODO(Grisha): also count NaNs by exchange, currency, asset class.\n",
    "def count_nans_by_period(\n",
    "    df: pd.DataFrame,\n",
    "    config: cconconf.Config,\n",
    "    period: str,\n",
    "    top_n: int = 10,\n",
    ") -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Count NaNs by period.\n",
    "\n",
    "    :param df: data\n",
    "    :param period: time period, e.g. \"D\" - to group by day\n",
    "    :param top_n: display top N counts\n",
    "    :return: table with NaN counts by period\n",
    "    \"\"\"\n",
    "    # Select only NaNs.\n",
    "    nan_data = df[df[config[\"data\"][\"close_price_col_name\"]].isna()]\n",
    "    # Group by specified period.\n",
    "    nan_grouped = nan_data.groupby(pd.Grouper(freq=period))\n",
    "    # Count NaNs.\n",
    "    nan_grouped_counts = nan_grouped.apply(lambda x: x.isnull().sum())\n",
    "    nan_grouped_counts.columns = [\"nan_count\"]\n",
    "    nan_grouped_counts_sorted = nan_grouped_counts.sort_values(\n",
    "        by=[\"nan_count\"], ascending=False\n",
    "    )\n",
    "    return nan_grouped_counts_sorted.head(top_n)\n",
    "\n",
    "\n",
    "nan_counts = count_nans_by_period(\n",
    "    ccxt_data_filtered,\n",
    "    config,\n",
    "    \"D\",\n",
    ")\n",
    "nan_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd3a24d2",
   "metadata": {},
   "source": [
    "## Detect outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "f7ade676",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-01-14T09:58:35.536230Z",
     "start_time": "2022-01-14T09:58:35.461912Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape=20925\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>close</th>\n",
       "      <th>z-score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-01-01 01:06:00-05:00</th>\n",
       "      <td>3706.95</td>\n",
       "      <td>3.119872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-01 01:07:00-05:00</th>\n",
       "      <td>3710.40</td>\n",
       "      <td>4.048159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-01 01:09:00-05:00</th>\n",
       "      <td>3709.03</td>\n",
       "      <td>3.012456</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             close   z-score\n",
       "2019-01-01 01:06:00-05:00  3706.95  3.119872\n",
       "2019-01-01 01:07:00-05:00  3710.40  4.048159\n",
       "2019-01-01 01:09:00-05:00  3709.03  3.012456"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# TODO(Grisha): add support for other approaches, e.g. IQR-based approach.\n",
    "def detect_outliers(df: pd.DataFrame, config: cconconf.Config) -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Detect outliers in a rolling fashion using z-score.\n",
    "\n",
    "    If an observation has abs(z-score) > `z_score_boundary` it is considered\n",
    "    an outlier. To compute a `z-score` rolling mean and rolling std are used.\n",
    "\n",
    "    :param df: data\n",
    "    :return: outliers\n",
    "    \"\"\"\n",
    "    df_copy = df.copy()\n",
    "    roll = df_copy[config[\"data\"][\"close_price_col_name\"]].rolling(\n",
    "        window=config[\"stats\"][\"z_score_window\"]\n",
    "    )\n",
    "    # Compute z-score for a rolling window.\n",
    "    df_copy[\"z-score\"] = (\n",
    "        df_copy[config[\"data\"][\"close_price_col_name\"]] - roll.mean()\n",
    "    ) / roll.std()\n",
    "    # Select outliers based on the z-score.\n",
    "    df_outliers = df_copy[\n",
    "        abs(df_copy[\"z-score\"]) > config[\"stats\"][\"z_score_boundary\"]\n",
    "    ]\n",
    "    return df_outliers\n",
    "\n",
    "\n",
    "outliers = detect_outliers(ccxt_data_filtered, config)\n",
    "_LOG.info(\"shape=%s\", outliers.shape[0])\n",
    "outliers.head(3)"
   ]
  }
 ],
 "metadata": {
  "jupytext": {
   "formats": "ipynb,py:percent"
  },
  "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": "3.8.10"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}