{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "032578d6-7f00-4bed-ba92-615a598e2963",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Florian Ellsäßer: Using a LSTM network and SHAP to determine the impact of drought and season on winter wheat\n",
    "\n",
    "[Presentation](http://spatial-ecology.net/docs/source/STUDENTSPROJECTS/Proj_2022_Matera/Using_a_LSTM_network_and_SHAP_to_determine_the_impact_of_drought_Florian_Ellsäßer.pdf)  \n",
    "[Video Recording](https://youtu.be/EqxW-PfZsKY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b76be5c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import xarray as xr\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy \n",
    "import datetime as dt\n",
    "#import rioxarray as rio\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a7cac6f2",
   "metadata": {},
   "source": [
    "## load the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "63230794",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load the data\n",
    "spei_data = xr.open_dataset(r'C:\\Users\\Florian\\Desktop\\Jupyter_Skripts\\05_ML_crop_vulnerability\\data/spei01_germany.nc')\n",
    "yield_data = xr.open_dataset(r'C:\\Users\\Florian\\Desktop\\Jupyter_Skripts\\05_ML_crop_vulnerability\\data/all_crops_productivity_gapfilled_detrended.nc4')\n",
    "yield_index_data = xr.open_dataset(r'C:\\Users\\Florian\\JLUbox\\Erntedaten_CROP_Projekt\\09_Final_output_files/all_crops_SHI.nc4')\n",
    "natural_areas_raster = xr.open_dataset('C:/Users/Florian/Desktop/E-OBS_data_25.0e/natural_areas_germany.nc')\n",
    "phen_data_winterwheat = pd.read_csv('C:/Users/Florian/Desktop/CROP_indices_data/phen_data_gapfilled.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1ee3f7f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x1172ac23630>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize the data for a quick check\n",
    "spei_data.spei.sel(time='2016-05-16T00:00:00.000000000').plot(cmap='seismic_r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dbba4cf3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x1172b40bef0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yield_data.winter_wheat.sel(year=2018).plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5d05b8b4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x1172b4dd518>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yield_index_data.winter_wheat.sel(year=2016).plot(cmap='seismic_r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b005704b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x1172b4b2a58>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "natural_areas_raster = natural_areas_raster.__xarray_dataarray_variable__\n",
    "natural_areas_raster = natural_areas_raster.where(natural_areas_raster != -9999.)  \n",
    "natural_areas_raster.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "def0b186",
   "metadata": {},
   "source": [
    "## align the data properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "41379aa7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# get the same order of dimensions \n",
    "# rename year to time in yield data \n",
    "yield_data = yield_data.rename({'year':'time'})\n",
    "yield_index = yield_index_data.winter_wheat.rename({'year':'time'})\n",
    "\n",
    "# transpose dimensions of \n",
    "spei_data['spei'] = spei_data.spei.transpose('longitude','latitude','time')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "88e63741",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a function to round coordinates (this is taken from the crops package)\n",
    "def round_coordinates(in_array):\n",
    "    \"\"\"\n",
    "        This function rounds the coordinates two two decimals. This prevents errors\n",
    "        due to weird roundings that sometimes appear\n",
    "        \n",
    "        Parameters:\n",
    "            in_array (xarray.core.dataarray.DataArray) = input Data array e.g. Tmax\n",
    "            \n",
    "        Returns:\n",
    "            in_array (xarray.core.dataarray.DataArray) = array with rounded coordinates\n",
    "    \"\"\"\n",
    "    \n",
    "    # round longitude\n",
    "    in_array.coords['longitude'] = np.round(in_array.coords['longitude'],2)\n",
    "    # round latitude    \n",
    "    in_array.coords['latitude'] = np.round(in_array.coords['latitude'],2)\n",
    "    \n",
    "    return in_array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "eb178269",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Frozen(SortedKeysDict({'latitude': 79, 'longitude': 93, 'time': 32}))\n",
      "('latitude', 'longitude', 'time')\n",
      "Frozen(SortedKeysDict({'longitude': 90, 'latitude': 77, 'time': 852}))\n"
     ]
    }
   ],
   "source": [
    "print(yield_data.dims)\n",
    "print(yield_index.dims)\n",
    "print(spei_data.dims)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "85802944",
   "metadata": {},
   "outputs": [],
   "source": [
    "spei_data = round_coordinates(spei_data)\n",
    "yield_data = round_coordinates(yield_data)\n",
    "yield_index = round_coordinates(yield_index)\n",
    "natural_areas_raster = round_coordinates(natural_areas_raster)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "df54e4f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# now dims are the same, adjust the extend of the lon and lat dims\n",
    "# get the dims of the smaller xarray\n",
    "min_lon = spei_data.longitude.min()\n",
    "min_lat = spei_data.latitude.min()\n",
    "max_lon = spei_data.longitude.max()\n",
    "max_lat = spei_data.latitude.max()\n",
    "#create a mask for later\n",
    "mask_lon_spei_data = (spei_data.longitude >= min_lon) & (spei_data.longitude <= max_lon)\n",
    "mask_lat_spei_data = (spei_data.latitude >= min_lat) & (spei_data.latitude <= max_lat)\n",
    "# now drop the coords that are too big\n",
    "yield_data = yield_data.where(mask_lon_spei_data.longitude, drop=True)\n",
    "yield_data = yield_data.where(mask_lat_spei_data.latitude, drop=True)\n",
    "\n",
    "yield_index = yield_index.where(mask_lon_spei_data.longitude, drop=True)\n",
    "yield_index = yield_index.where(mask_lat_spei_data.latitude, drop=True)\n",
    "\n",
    "natural_areas_raster = natural_areas_raster.where(mask_lon_spei_data.longitude, drop=True)\n",
    "natural_areas_raster = natural_areas_raster.where(mask_lat_spei_data.latitude, drop=True)\n",
    "\n",
    "# transpose again\n",
    "spei_data['spei'] = spei_data.spei.transpose('latitude','longitude','time')\n",
    "#yield_data = yield_data.transpose('longitude','latitude','time')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "ff6b301e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Frozen(SortedKeysDict({'latitude': 77, 'longitude': 90, 'time': 32}))\n",
      "('latitude', 'longitude', 'time')\n",
      "Frozen(SortedKeysDict({'latitude': 77, 'longitude': 90, 'time': 852}))\n"
     ]
    }
   ],
   "source": [
    "print(yield_data.dims)\n",
    "print(yield_index.dims)\n",
    "print(spei_data.dims)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0a6e15d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# concatenate the data into one set\n",
    "in_data = xr.concat([natural_areas_raster,yield_data.winter_wheat,yield_index,spei_data.spei], dim='time')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71d2c30b",
   "metadata": {},
   "source": [
    "## apply ufunc to go through the raster cells "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6196e76",
   "metadata": {},
   "outputs": [],
   "source": [
    "## use apply ufunc to go through the xarrays along the time coordinate\n",
    "# the purpose of this is to create a list of lists that can later be converted into a pandas data frame\n",
    "# for each harvest date, we want the previous 24 month of SPEI data\n",
    "test_list = []\n",
    "test_list.append(['harvest_year','harvest_month','natural_area','yield','index']+\n",
    "                ['SPEI'+str(i) for i in range(1,25)])\n",
    "# first we create a yield_calculation class and a function to return the three best years\n",
    "class calcualte_impact:\n",
    "    def __init__(self,in_time):\n",
    "        self.in_time = in_time\n",
    "        pass\n",
    "        \n",
    "    # now define a function that takes the yield data and returns the three best years            \n",
    "    def run_model(self, in_data):\n",
    "        # first get the times\n",
    "        in_time_index = self.in_time[1:33]\n",
    "        in_time_yield = self.in_time[33:65]\n",
    "        in_time_spei = self.in_time[65:]\n",
    "        # the get the data\n",
    "        natural_area = in_data[0]\n",
    "        index_data = in_data[1:33]\n",
    "        yield_data = in_data[33:65]\n",
    "        spei_data = in_data[65:]\n",
    "        \n",
    "        if np.isnan(spei_data).all():\n",
    "            # create a list with the length of  total_years_in_data full of Nones\n",
    "            result = [None] * len(index_data)\n",
    "            return np.array(result)\n",
    "        \n",
    "        else:\n",
    "            # create a pandas dataframe from in_time_yield and yield_data\n",
    "            yield_df = pd.DataFrame({'time': in_time_yield, 'yield': yield_data, 'index':index_data}, \n",
    "                                    columns=['time', 'yield','index'])\n",
    "            # add a column for the harvest dates\n",
    "            yield_df['natural_area'] = [natural_area]*len(index_data)\n",
    "            harvest_date_list = []\n",
    "            for year in in_time_yield:\n",
    "                try: \n",
    "                    harvest_date = phen_data_winterwheat[(phen_data_winterwheat.natural_area_group_code == natural_area) &\n",
    "                             (phen_data_winterwheat.reference_year == int(year))]['start_date'].values[0]\n",
    "                    harvest_date_list.append(harvest_date)\n",
    "                except: \n",
    "                    harvest_date_list.append(None)\n",
    "                      \n",
    "            # add the harvest date to the df\n",
    "            yield_df['harvest_date'] = harvest_date_list\n",
    "            # convert the time strings to datetime objects\n",
    "            yield_df['harvest_date'] = yield_df['harvest_date'].astype('datetime64[ns]')\n",
    "            # convert the time strings to datetime objects\n",
    "            yield_df['harvest_date'] = yield_df['harvest_date'].astype('datetime64[ns]')\n",
    "            yield_df['harvest_date'] = pd.to_datetime(yield_df['harvest_date']).apply(lambda x: x.date())\n",
    "            # create a column for years and month\n",
    "            yield_df['harvest_year'] = pd.DatetimeIndex(yield_df['harvest_date']).year\n",
    "            yield_df['harvest_month'] = pd.DatetimeIndex(yield_df['harvest_date']).month\n",
    "            \n",
    "            # create the same for in_time spei and spei data\n",
    "            spei_df = pd.DataFrame({'time': in_time_spei, 'spei': spei_data}, columns=['time', 'spei'])\n",
    "\n",
    "            # convert the time strings to datetime objects\n",
    "            spei_df['time'] = spei_df['time'].astype('datetime64[ns]')\n",
    "            spei_df['time'] = pd.to_datetime(spei_df['time']).apply(lambda x: x.date())\n",
    "            # create a column for years and month\n",
    "            spei_df['year'] = pd.DatetimeIndex(spei_df['time']).year\n",
    "            spei_df['month'] = pd.DatetimeIndex(spei_df['time']).month\n",
    "            \n",
    "            # now form an output for every year: year, harvest_month, natural_area, index, yield, 24 SPEI\n",
    "            for index, row in yield_df.iterrows():    \n",
    "                \n",
    "                # get the SPEI for each harvest month\n",
    "                index_df = spei_df[(spei_df.year==yield_df['harvest_year'][index])&\n",
    "                       (spei_df.month==yield_df['harvest_month'][index])]\n",
    "                \n",
    "                \n",
    "                try: \n",
    "                    new_index = index_df.index.values.astype('int')[0]\n",
    "                    # create a range of indices +23\n",
    "                    index_df_range = list(range(new_index,new_index+24)) \n",
    "                    spei_range = spei_df.iloc[index_df_range]['spei'].values\n",
    "                except:\n",
    "                    spei_range = [None]*24\n",
    "                 \n",
    "                \n",
    "                # append all to a list\n",
    "                test_list.append([yield_df['harvest_year'][index],\n",
    "                                  yield_df['harvest_month'][index],\n",
    "                                  yield_df['natural_area'][index],\n",
    "                                  yield_df['index'][index],\n",
    "                                  yield_df['yield'][index]]+\n",
    "                                  # list(spei_range).reverse()\n",
    "                                 list(spei_range)\n",
    "                                 )\n",
    "            return np.array(yield_df['harvest_month'])        \n",
    "        \n",
    "# now we create an object and give it the in_years of the yield data\n",
    "yield_object = calcualte_impact(in_data.time)\n",
    "\n",
    "# now we can apply ufunc and get the years with the highest productivity\n",
    "test_out = xr.apply_ufunc(yield_object.run_model, \n",
    "                     in_data,\n",
    "                     input_core_dims=[['time']],\n",
    "                     output_core_dims=[['year']], \n",
    "                     dask = 'parallelized', \n",
    "                     vectorize = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42d880e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(test_list)\n",
    "df.columns = df.iloc[0] \n",
    "df = df[1:]\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6564f67",
   "metadata": {},
   "outputs": [],
   "source": [
    "# revove all the rows with a Nan\n",
    "df.dropna(inplace=True)\n",
    "df = df.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "77e93a6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('C:/Users/Florian/Desktop/Jupyter_Skripts/05_ML_crop_vulnerability/data/yield_spei_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ae589cc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b5f4d29",
   "metadata": {},
   "source": [
    "## ML part"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "bf6ec909",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import tensorflow as tf\n",
    "import time\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "import keras\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.models import load_model\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from tensorflow.keras import optimizers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "791b5442",
   "metadata": {},
   "outputs": [],
   "source": [
    "# on LSTMs: https://machinelearningmastery.com/gentle-introduction-long-short-term-memory-networks-experts/\n",
    "# create data similar to this:\n",
    "# based on this: https://machinelearningmastery.com/how-to-develop-lstm-models-for-time-series-forecasting/\n",
    "\n",
    "# X -> array([[10, 20, 30],\n",
    "#       [20, 30, 40],\n",
    "#       [30, 40, 50],\n",
    "#       [40, 50, 60],\n",
    "#       [50, 60, 70],\n",
    "#       [60, 70, 80]])\n",
    "# y -> array([40, 50, 60, 70, 80, 90])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "fbbe03de",
   "metadata": {},
   "outputs": [],
   "source": [
    "# read the data again\n",
    "df = pd.read_csv('C:/Users/Florian/Desktop/Jupyter_Skripts/05_ML_crop_vulnerability/data/yield_spei_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4cb9271",
   "metadata": {},
   "source": [
    "### do with all data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "119de43b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create X and y data\n",
    "X = np.array(df[['SPEI1', 'SPEI2', 'SPEI3', 'SPEI4', 'SPEI5', 'SPEI6', 'SPEI7',\n",
    "       'SPEI8', 'SPEI9', 'SPEI10', 'SPEI11', 'SPEI12', 'SPEI13', 'SPEI14',\n",
    "       'SPEI15', 'SPEI16', 'SPEI17', 'SPEI18', 'SPEI19', 'SPEI20',\n",
    "       'SPEI21', 'SPEI22', 'SPEI23', 'SPEI24']].values.tolist())\n",
    "y = np.array(df['index'].values.tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "fee1e9bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create train and test data by splitting\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "8f629322",
   "metadata": {},
   "outputs": [],
   "source": [
    "# define model\n",
    "n_steps = 24 # --> month of SPEI\n",
    "n_features = 1\n",
    "model = Sequential()\n",
    "model.add(LSTM(50, activation='relu', input_shape=(n_steps, n_features)))\n",
    "model.add(Dense(1))\n",
    "model.compile(optimizer='adam', loss='mse')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ee48c129",
   "metadata": {},
   "outputs": [],
   "source": [
    "# reshape from [samples, timesteps] into [samples, timesteps, features]\n",
    "X_train = X_train.reshape((X_train.shape[0], X_train.shape[1], n_features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2593616a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# fit model -> outcommented because it takes like 4h\n",
    "#model.fit(X_train, y_train, epochs=200, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "94c4c8a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# save the model\n",
    "#model.save('C:/Users/Florian/Desktop/Jupyter_Skripts/05_ML_crop_vulnerability/models/model_v001')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a865b8ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# load the model again\n",
    "model = keras.models.load_model('C:/Users/Florian/Desktop/Jupyter_Skripts/05_ML_crop_vulnerability/models/model_v001')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "b01afa0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# test the model\n",
    "X_test = X_test.reshape((X_test.shape[0], X_test.shape[1], n_features))\n",
    "y_pred = model.predict(X_test, verbose=0)\n",
    "# reformat the output\n",
    "y_pred_list = []\n",
    "for item in y_pred:\n",
    "    y_pred_list.append(item[0])\n",
    "y_pred = y_pred_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "9b433d8a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n predictions: 45036\n",
      "slope: 1.0054949082304028\n",
      "intercept: 0.01712772477690782\n",
      "r: 0.9177724506004394\n",
      "r²: 0.8423062710811361\n",
      "p-value: 0.0\n",
      "standard error: 0.002050131923845098\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEGCAYAAABsLkJ6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABon0lEQVR4nO2dd3yV5fn/39c5GRAIEGZYARFEDI6ycUJFKxYXanG2ahW1ttVfl1ZbtHSoX+2wlSqIo1pAi6goihvEwQyCJCICgYSwRwhhZZxz//54Rp4z8yQ5J+ckud+vV145z3OecZ91j2t8LlFKodFoNJqWhyfRDdBoNBpNYtADgEaj0bRQ9ACg0Wg0LRQ9AGg0Gk0LRQ8AGo1G00JJSXQD6kLnzp1V3759E90MjUajaVLk5eXtU0p1Cd7fpAaAvn37smrVqkQ3Q6PRaJoUIlIUbr82AWk0Gk0LRQ8AGo1G00LRA4BGo9G0UPQAoNFoNC0UPQBoNBpNC0UPABqNRtNC0QOARqNpEeQVlTJt0SbyikoT3ZSkoUnlAWg0Gk19yCsq5fqZy6is9pOW4mHWraMY2icr0c1KOHoFoNFomj3LCvdTWe3Hr6Cq2s+ywv2JblJSoAcAjUbT7BnVrxNpKR68AqkpHkb165ToJiUF2gSk0WiaPUP7ZDHr1lEsK9zPqH6dtPnHRA8AGo2mRTC0T1ajd/x5RaVJPejoAUCj0WjiQF5RKT+YvhSfX+H1CP+7fXT9BoHnn4d58+CVV6BNm5i2UfsANBqNJg48unA9Pr8CwOdXPLpwfd0usGoViMAtt8Dbb8OhQzFvo14BaDSaFkFjmWOs+6zfFdhhb9pz2N0F9u2D7Gzw+Wr2bdkC3bvHsJUGCRsARKQVsARIN9vxqlLqwUS1R6PRNF+i5QEEDwwNGSic91Eq8Ln+XdtGP7m6GsaPhw8/rNn3/vtwwQV1akNdqHUAEJGuwFlAD+AYkA+sUkr5G3jvCuC7SqnDIpIKfCYiC5VSyxp4XY1GowkgXB6A1dk7B4YpE3KZuqCg3gljzvt4BFCgAK8H7h0/KPKJf/4z/O53NdsPPwz33Vfv1+uWiAOAiIwF7gM6Al8Ce4BWwOXAiSLyKvBXpVS9DFNKKQVYa6JU809FPkOj0Wjqh5UHUFXtD8gDcHbYldV+FubvDDtQ1Pc+UybkUnq0MvJq4r334KKLarbHj4e33gKvt6Ev2RXRVgAXA7cppYqDnxCRFGACcAEwr743FxEvkAf0B6YppZaHOWYyMBkgJyenvrfSaDQtmEh5AFkZaZh+WvwKcru3Y+XWA3YHnpWRxrRFm1ybg1znG2zZAv361Wy3agUlJdCpcRPURAUbqhKAiHQAXgd+ppTKj3TcsGHDlK4JrNFo3ODGlj9t0SYef28DCiMk8pffG8iofp1YVrifrIw0HnqrgKpqPyle4ephvblySK96O5DzikpZ9XUJP7rrClpt2ex4Ig+GDKnXNd0iInlKqWHB+2sNAxWRu0WknRg8KyKrReTCWDZOKXUQWAxcFP1IjUbTknGr6GnZ9v/6/gaun7ks4vGj+nUiPdWQiEhL9diDxV1j+1Owo8xw5gJVPsXs5cVcO2NpvdRE87YeYOPE67n94tNqOv///AeUinvnHw03UUC3KKWeEJHvAV2Am4HngfcbcmMR6QJUKaUOikhrYBzwaEOuqdFomhfWLD4rI42CHWXMXbWNar+qNZInktM3mKF9srhpdF/eLdjFRbnZAceEs41U+hTzVpfUbRXw4osM/dGPGGpuzj7jIkr/+k/u+u6AOrwT8cHNACDm/4uB55VSa0VEop3gku7Af0w/gAf4n1JqQQyuq9FomgHWLL6iyh/SGUeL5Jl166iITt/ggWL28mKeXlIIwNNLCsnp1IbrRhq+xsE92odtV22dn3WPsUe3c8qEMfb+wk69uOSmf+BrncGsEzvX5y2JOW4GgDwReR84AfitiGQCDQ0BRSn1FfCdhl5Ho9E0T6xZfHDnLxAxkscaGO4a2z/EGTt7eTFT5ufj8yvSU42BYmH+zoBrL8zfaQ8ApUcr8Qi2kxggLcXDxCG9IrY5r6iUO594n0//eSPpvqqaJwoLKfV04CdJpgvkZgD4MXAGUKiUOioinTDMQBqNRhM3rFl8ZZUfP0bH7/EI3z25K3ecd2JAJI+IIErh9dYMDE7xt7yiUn5vdv4AlVXGQNGpTVrAPZ3bzlWE12M4gSdGcwL7fGRePoEVa76wd73wu39z0x/vNNpjtimZcKMFpIBTgJ+b220w8gE0Go2mTtSlLKMVUvnL7w3kjnP74fUIPr9i0Td72LCr3L7eQ28V4PMrFOD3hzdOvLa6xO78wZDYGdWvE/uPVAYc59we2ieLKRNyObN/Zx66dDB/vuLUyB34ww9DSgonmZ3/4+fcQN97F/D5gOG1vs5INEYJSzcrgH9jmHy+C0wFyjFi/+v/yjQaTYsjXNZt1CQpYMOucpYV7ud4lY9qswOv9it+/8Y6CnaUGRE61TWdvs9PWIfvnvKKgG1rdTB+cHc+3bjP3j9+cHdmLy9mYf5Ocru344WlW6ms9rNy6wEGZmeGtvODD+DCmqDIjaeP5qIL78PnMRK59hw6Tl5RaYMkJeJZwtLNADBSKTVERL4EUEqVikhabSdpNBoN1DhFdxw8VpN1W+Vnyvx8/EqR4vVw1dBeITH2s5cXc//r68Je06eM51NTPHi9QrXPGBxEDJNQMOEct3lFpZQereTyM3qwrHA/vTtmULz/iO0U/nTjPkw1ByqDI4mKiqBv35qLpabCjh0cOuLF+8wy/Kbv4quSMq6fuaxBkhL1yUh2i5sBoMqM1FFgh2822Ams0WiSl1gpZzpnsikeweMRlE8hAn6lbAmGOcuLeW11id1R5hWV8qe3v456bQX4fH6G9slixVbDTOJX8Pv5+eTvKGNwj/b2CqNLZnrAuSu3lvKDp7/A53Dw7jpUwaqtgeYW62m/MgeWY8dg6FBY75B2XrkShhk5VkM7w5zbRvGPD7/ls437UEBFlb/OoaORophijZsB4J8YWbpdReTPwFXA76KfotFomiqxND84Z7LVPgVizSQhxWN0cApsU45VrP36mcs4XhV+ntmnYwY7Dx3H5/Pj8Yjd+Vv4/EbSFhiCbJa5yRnRoyCg87eIqIugFB1+dTd89kbNvueeg5tD42GG9sninnEnsbxwP5U+wzfxal5JnbKIG6uEZa1OYKXULOA3wMPATuBypdTcuLRGo9EknHDmh/oyql8nUjximGAcHbDPDzkdM7jglG6keSWgWLt1/0gUlx7lljP7Mrhn+wDHbjis11B6tJLzB3Wrtb2CodwJhjkJ4PKCRWz9v0u42Oz89026Afz+sJ2/xdA+WVw9rLdtevL56v4+WhnJ8YwciqYG2k4pdUhEOmIogc5xPNdRKXUgbq3SaDQJI+bmB7Es6YFs2nOYktKj3HLWCRTsPMT4wd3tzi4txWMPQsEoBTM+LUQpd/LBXo8wql8nRvXrxCcb9lAZbupv0i0znT2HDYfxoN2FvPP8z+3ntnbozvib/8U5Z/Th9MWbw87MnaaziUN6MW91SdzNOA0hohiciCxQSk0QkS0Y77M4/yul+oU9MY5oMTiNpnGIlQ9g2qJN/PX9DfhVjSPW2eM4nbOpXuGhSwdTerSSrIw0XllZzNqSsnrf2yLFK0x1XHfRhj188PXusMdmt0tHHTjAx//8EW2qjtv7z538DMVZRkUur2nGCidHEWw6A5KiKHwkMbiIKwCl1ATz/wnxbJhGo0k+nElUteHU6wkO6wxO5nISvC6o9BnhnVbnelrP8FIMbnBeu9qn7GiiVqmRr+vx+3j21T+Q++Vn9r6brnqIxSfW9Jte04zl9FlYrzVSRnKyJX85iWYCiipRp5RaHfvmaDSapoRz1mvN8i2ZBWsQmXXrKO6d91VITdysjFQOHK0K2GdZZyqr/a4SoMIblyKbho5X+Vm7PXRVMXn5PO5f/HzNjoce4oHTJrJ4eU05lNN7tWfS8BymLigIa9ZprMidWBItCuivUZ5TGIlhGo2mheE0DzlnvWDGzFeFxq0X7g0tiB7c+YPpgFUgIrU6eK371ZUKR3TR6KK1zHn5AXt719DRZC9bAikpTCwqZW5ejQ1/yiW5DO2TxcDszLBmncaK3Ikl0UxAYxuzIRqNJvkJl82b4vUERO34CUzGWla4P6wzNxzZ7Vpx3cg+lB+rYrqZkBUPehzawxdP3WJv+xGG//Qlbpk4klHby+1OfM5tdevQ62I6Swbc5AEgIoMx9IBsDSCl1IvxapRGo0lOgu3cBTvKjLCcIBZv2GOrapYfC53pR2L7weOUH6viuc+3xKVAeHp1JW+8+AsG7d1q77vsxr+ytsdA0rxCVkZaiCP3rrH97WMbS6Khsah1ABCRB4ExGAPAO8B44DNADwAaTYKJVbSOW2ynbrUfEWFPeQVVYcIqC/cdYdqiTZQfq7KlFdzy3+VFUUM1g8lul05ltT+sSclGKR78aAY3571l77r3op/xyunfA4xIVSsCKZoEQ2NJNDQWblYAVwGnA18qpW4WkW7AzPg2S6PR1EY8Z6ORBhargtaMTwvx+RUff7M77Ey9cO9hHntvQ73ufbjCV6fj26anMHJQJ2Y5HLb9u7ShcN8R/Aou/foT/vnWY/ZzcweP49cX312T6YWxiCnYUcbEIb0CBrhgXaGm6OiNhpsB4JhSyi8i1SLSDiMprNFzADQaTSDxmo06BxaPCFMvG2ybc/KKSpn52ZaAjN5wuLX5x4JUr4fcHu1tuWiPQMc2aaStL+CdZ39qH1fSrisX/ngaR9Nah73O3FXbmDikF1Mm5NqFY6YuKAhQAW2Kjt5ouBkAVolIB+AZIA84DKxo6I1FpDeGGSkbw280Qyn1REOvq9G0FOI1G3UOLH5lxOYPzM4E4B8ffmvLMkPkMMzGZN/hCh54fZ3djrbHDjPznkm0qzhiHzPmtuls7dgz5FyRGheGz69suQa/UmFj/aHpOXqjUesAoJT6ifnwaRF5F2hnlnNsKNXAL5VSq80yk3ki8oFSKroEoEajAeI3Gx3Vr1NAhq5PwfRPNrNk496AEEpIfOcPsPewUcRFlJ/pr/+FCzcus5+75copfNx/RMRzbz+nHy8s3WoPolkZaeTvKLNVSy0ZCTc0tj8mFriNAjoN6GsdLyL9lVKvNeTGSqmdGOJyKKXKRWQ90BPQA4CmUUjUDzaW943HbHRonyzOH9SN9x1yCYV7D4ctzp4s/HjF6/x+0bP29hNnXsPfz7mh1vMuyM0mp1MbuwDMQ2/mBzqgRSKfbJJXVMq81SW8mldCta9pRQe5iQJ6DjgNKKCmDoACGjQABN2jL0aB+OVhnpsMTAbIycmJ1S01LZxEhfM19n3rOthYx48Z2JWPvtljJ2Nt3nskKTv/UcVf8fKc++3t5b0Hc/2kP1HtDe3awpmrHlm4ni+LD+LzKz7fuC9ErsJS8Yz03lmfp3NwbErRQW5WAKOUUqfEqwEi0hajxOQ9SqlDwc8rpWYAM8AQg4tXOzQti0SF80WSWo7HSsTtYOPU8pm6oICKKr9ZsKXmmGT74WUf2seyp24K2Dfspy+xr03d3r9VRaW2DyD4NQrU6luxPk9Vh3OSCTcDwFIROSUetnkRScXo/Gc11KSkiU5TtE/Gk0SF8wXfN1ziUV0+n2ifq3Owqajy81qYqlSzlxczZX5+gGMXwuZ2JQVp1VW89t9fMXj3ZnvfFTc8zpc9T671XIGQgS3SyOYRuGZETq1FXJyfp9cjXD2sNxPrUPgl0bgZAP6DMQjsAiqokYM+rSE3FhEBngXWK6X+1pBraaLT3LIXY0GiwvmC79uQlUhtn+uofp1smQYFvGKGOTrli8N1/snK7z56hltXzbe37//eXcw+Y7zr81UYG5Bz5u58avI5/bjv4kG1XrOph4W6GQCeA24E1hHbWsBnWdcVkTXmvvuVUu/E8B4aml/2YqxIVDhf8H3ruxKp7XMd2ieLMSd1sZ251T4VsApYVri/SXT+F3/zGf+e/4i9PS93LL/8/i9cOWgDiFJAxuMR/H4j9NMjkNk61fVlm3JYqJsBoFgp9Wasb6yU+ozAehCaONHcshebE84ZZFZGmu0TcNOhuPlcOwcVQ3d2gMFZrsnGgL1FfPDcXfb2zradGHfrUxxJz6jztdpnpHCswhdRYkL5lT0IpHhbzm/EzQDwjYjMBt7CMAEBoG32TYemvkxtitTF52I9X1cznZvP9cohvXh11TaqfIpUr3DlkF72cwU7Gl5tKx60O36YJdNvpcPxGgnp7976NIWdekU5Kzoj+naia2Y6s5cXh3X2pqR48PtNA0eyOkDigJsBoDVGx3+hY19Mw0A18acpL1OTgbp06PXxudTXTBfuc529vJhXVhbTrV0rbj/vROZMHs2ywv1s3F3OL/+3hotys7nv4kF8u7u89hfeiIjy89QbD3PRt0vtfbdN/B0fDBjVoOt6Be4470QA5uaVhBSc79mhFZ3aprNuexmKmozglvB7cZMJfHNjNESjSVbcdOiRiqS47czra6abvbyYhfk7GT+4O9eNzOGRd9Y71DfL+ODr3Vx2Rg9Wbj3A9oNGjdunlxSy69BxVxW3GoubV83nwY+esbefHP0DHj/3hw26Zs8OrThvYFc7kievqDTs7L7k4HG2Hzxu2/+t978lRM65SQTrBfwLw2mrMKSg71ZKlcS5bRpNUlBbhx6uSEpdO/No5pxIHdHs5cV2rdtPN+6jeP8RZnwaKL2sgDfW7Ai53+Jv9zaqYFskhm/LZ+7s++ztVT0Hcc21D4dN5KorYwZ25c9XnGpvR3N6K8ADnNW/M/eMOwmou0muKeLmXX4emA1cbW7fYO67IF6N0miSidpm58EDROnRynr5XJzmHGdy1kNv1dSgnXNbTUe0MH9nwPnvFuxybb7u2zGDNUcT5wPoVr6P5f++KWDf8LteYm/b2HSyaSkeJg4J9BkExOx7PaAU1T6FH2Pmn5bi4Z5xJzG0TxbTFm1Kqsi5eK1G3AwAXZRSjmrJvCAi98SsBRpNkmPNzuetLgkbthZugAhnm3f7I3auKITAQulT3yqwa9N2ahMYxZPTMYMdZcdDbNzhWFuSmM4/rbqKV2f9mtN2bbL3Tbz+MVb3qj3m3i09s1pz15j+IRFVwasswB5kS49WBnwu8Yicq28nHs88HjcDwD4RuQGYY25fC+yPyd01mibEa6sNB+K81SUBP0I30Th1+REHF1p3srakjEnTlzL1ssFs2Xck4LmyY1VcNbQXsx2FUSKRCOvPbxc9x+0ramJHfnfBnfx3yPdjfp/WKR6mLigI+14HD8yRPoNYR841pBOPZx6PmwHgFuBJ4O8Y35svzH0aTZPFzWysLo7d2qKs3PyInWafFK8x+wyRLgCq/YZG/0ndMgP2bz1whP2HK0g2LtrwOU+/8bC9PX/Qedx9ya/qnsjlkqw2aXY1sIZ0mLGMnGtIJx7PPB43UUDFwKUxu6NGk2DcRvU01LHrvNaOg8fwmMVHRGDHwWPkFZUG2Pyt+6V4Pfh8ftsx6QkzCPgU7C4P7OzLjlZTdrS6Pm9JXDhx3zY+evZOe3t3246cf+vTHK5HIlddWFtSRopZHczrkZD3OhE0pBOPZx6PmyigLsBtOOoBACil9CpA0yRxMxurq2M3eEURTmHT6sOr/TBnRXGAKSn4ftaxPmWEM1ohnE4OHKmMw7vTcNpWHGXxjNvo7HAyn3/rU2zu1LvB1+6Z1ZpUj7B1/9GIx/h8fs4f1I3CfUco3Hs45L1OBA3txOOVx+PGBDQf+BT4EKhbtWaNJglxMxtz69iF8KsFywbtEWMmGmxzDx58nPfDPMdi/+Hk7OiDEeXnX/P/jwkbPrP33X7F/bx30pkxu0fnNmkcqQzshlI8woTTuvNuwS47wufjDXuodsg+VCZBJE8yJmO6GQAylFL3xr0lGk0j4Tbm3u2MLXj2/srKYseM3zBD+P1GuKGlOulMOApuU1ZGGr+fv84uuH7cRVRPovlh3ltM/XC6vf30yCt5ZEzsc0jTUjzsOHgsYF/fThn845rv2J/d9oPHmBPGEV5+rIppizY168SuuuJmAFggIhdrlU5NcyJSmGawb+Cusf1rvZYlu1xV7cfjEfJNSQGAFK+HW87sS8HOQ3Rqk8b+I5Xkdm9HZuvUkI5ow65y3i/YRWW13+78k52hJV8zb9Zv7O0vuw/kB9c/QpXXvZpmXVi5NTR7+Zaz+wUM3ACvBkk++BV2hnSrVOOzhfgU4mlKuBkA7gbuF5EKoIqaegDt4toyjaaRaVC4nZmB5TMTiyy6t2vFjE8LbSeuACu3HgiwR+cVlfLowvWsCNO5JStdDh9g5bRAqYYRP/kPezIbV0VzeN8sCnaU8dCbRl0Da+Cec9soXltdQv72Mr4qKQswwVVWGaG8Vlhvc870rQ03UUCZtR2j0TRlnA7btBSjgIqIuJZLtiQGFKHx9UUHAp2VisDBJa+olGtmLKUqgkxxspHqq+J/s+7jOzs32Puuuv5RVvXKbfS2eMSI+Fm1tTSkHm9WRhrFB44yul8nCnYcCpCA8HgEgaTK9E0UrgQ3RCQLGAC0svYppZbEq1EaTWPhLOrtNZ2JC77aic+vmLqggIHZmbVm7e44eIwUj1DlC3X2hkME21Tx2uqSJtP537v4Be5c/qq9PWXc7bw49JKEtadf5zYU7jsSUNUrNcVD+bEqHnvPGKA+3biPy8/owYKvdlLtV3gEbj37BC7IzWbe6hLbabxm20EeeH1dkyrnGAs8tR0gIrcCS4D3gD+Y/x+Kb7M0msZhWeF+22Fb7VfMX7sDv1IBM3UwOvr7X1/HA6+vs1U0rcFjzopiEKFbu8DiKz2zWoe9Z7UfPijYxQOvr2Pxhj3xfHkx4cJvl7L10Ql257/g5HM44TdvJrTz94ph+0/x1nRhHoEpE3Ip2Hko4Ng12w5y69knkOIxEs9eWLoVgFm3juKaETn4fH4++Ho3s5YXc+0zy5JKJTXeuPUBDAeWKaXGisjJGAOBRtPkGdWvEx6p0dtRyjAReFEBssDXzlhqV5Oam1fCnNsMbSBr8LCKgjs5cLiCLm3T2BsmjHP6ksKEyDHUhX77S/h45h329r6M9oydPIPy9DYJaY/XAz6/MWs9f1A3BmZnctXQXsxxFHkpPVrJ+MHd+XTjPvu8rfuPMvOzLfiVCjD53DW2P/NWl+BcgCVDuGhj4mYAOK6UOi4iiEi6UuobERkYi5uLyHPABGCPUmpwLK6p0bjFsv2f1C2T9btqiqNkt0tnzMCutjlg2qJNAWaaymo/d83KY+/hCrvjUWAna6V5hUqf4liVn2NV4WP4k7nzb1NxlI9n3kG3wwfsfeN+/G82dc5JYKuwI6P8wIfrd7Nk416mTMglPTU0XwNgxpLNdsKYlRUsjoEdwtekjVepzGSsL+BmACgRkQ7AG8AHIlIKhAqM148XMHSGXozR9TQaVwQobgZp0mw/eJx5q0tsOeFR/Trh9RimG4tdhyJr7kSqO5v0KMUTbz3OZes/sXfdedl9LDz57AQ2KjzWTH7xhj0M7JZJ13atuOO8E+2O9bqRxmBl1UtQGLb/4PDbiUN68fLKbXbinWCsImJNPBU9G0KtPgCl1BVKqYNKqYeA3wPPApfH4uamI/lArQdqNHUkr6iUaYs2RbTnOkM+lVIh5hun/X9onywmDU/s7Dfe3PDlO2z9v0vszn/G8Cvoe++CpOz8wbD3ezzC+1/vZm2JUflsw67AEpelRyuxPlaPQGbrVO4a2z9ExO+Plw0mxSN4gPTU+BSEDxdinAzUqeyOUuqT2o+KLSIyGZgMkJPTvH+Emtpxq+JphVameoWXJ48OOTZY6mHKhFzyd5Txal4JPp+xLysjzc4cnTikV9h6sk2dIdvX89p/f21vf5Xdn6uuf4zKlPgkcjUErwfOP7kbAJ0z0ynYXhZQ12Bh/k6uG5kTEtZbmwDbdSNzGJidGVfzTDwVPRtCw+uuxRml1AxgBsCwYcOa6NpaEwvcLqOnf7LZttlX+RTTP9nMjB8OCznu3AFdKNx7mH5d2jIwO5PrRuYwuEd7FubvJLd7uxBN+WCHYzB9OmZQfOBoUtv3LbocLmXltBsD9o38yQvszuycoBZFR4AhOVl8/M0e/MpI+LppdN+AAWD84O5hdZmCi72Ew41OT0Ns+PFU9GwIST8AaDQWkZbRwT+q3YcClTODt/OKSrn2mWX2bH7T3iMs2rCHsQO7snjDHqr9iqWb99sibhVVfv7x4bfkdm+H1xNe3A1Ck76SkRRfNXPm3M/w7V/b+35w3SOs6J3YGAzLAGe9rx7B1lBSGIJvq4sP2rb6ymo/ma1T+csVp9oDdunRSju716ni6kbOozZiYcNvkmJwIvJosBhcuH0aTbwJXkZnZaSF/VFOGp7D2pJ19nnB9vtlhfsN1U0HVT7F+1/vtreVUgERPp9u3GeHFoql6NbE+OWSl/jZ0lfs7anfvY3nhl+WsPZYb6MA3TLT2WXWN/AAp/Zsz/qdh/Cb9RPatkoNkL8WEXvQH5idWVNLwSN2PYVwppb6zuLjWZUrkbhZAVwABHf248PsqzMiMgcYA3QWkRLgQaXUsw29rqZ5EryMXla4n+NVRkdeUVXzo7xuZA7F+4/wbsEuLsrNtiNCLAzxNomYgSsQdabvtvB6sjBu43JmvvZHe3vhSWfyk8vvQ0mtMSBxRcwCOQrszh/A4xW6tmtla/goFab2geNDcHbOPr9i0oje9OzQOqzSa6RZfG0DQ7La8BtKxAFARO4EfgL0E5GvHE9lAp/H4uZKqWtjcR1Ny8G5jP6gYJe9X2HI/YLxY35h6VYqq/28sHQrOZ3aBNiBh/bJ4uphvUNq53rEUO+8amgvBvdoz9QFBfYA0xQ54cB2Fj1zu719sFVbzr19JodatU1gq2oIV/MYoG+nNizesCfqIsunYN7qkpBaCqkpHq6MIOcQaRbvxryTrDb8hhJtBTAbWAg8DNzn2F+ulNKhm5pGxRnZYXXmS4NC6ZY6fAJWhm5FlZ8p8/Ntx6H1475ySC9eWVFsZ4EKcFb/ztwz7iT7x128/0iTyNgNJqPyGB/OvJMe5TXZsBfe8iTfdumbuEbVgY4ZqWzaU/u7vs9cNbjtnCPN4t2ad5LRht9QIg4ASqkyoExEfgfsUkpViMgY4DQReVEpdbBxmqhpTJIxW9E5Q/Mro7NOT/VwWs/2Acd1a2doFWZlpAXY7y1TTvCP2+MRfOYI4BEjisRpEpjxaRPr/JXir2//jSsLFtm77rr0Xt4edE4CG1V3gt/zEX2zqKj2k57iYVVRqb1yWLxhT51q/UYaKJqreccNbnwA84BhItIfIwnsTYzVwcXxbJgm/oSrY+ssTH7V0F4Rl9PxbosT5wwNaiSVOwSl7I8Z2BUgVGBNwAsB2j7/+PDbgJKBPgUPvplP/o4yBvdoz78XbYxookhGrl3zLg+/96S9/dzQS5k6bnICW1R/Kqv9pJk+mlSvcPl3evHQm/khGdbVfmVHgrmN0Ak3i2+u5h03uBkA/EqpahGZCPxDKfUvEfky3g3TxJdwdk9nR1tZ7WfO8mJea4Ri2rXZYK0ZmnMQ8Ho9dM5MxyPYq4LFG/ZQerSSwr2HA66vFAzu1d6OBrrWTBIL7t+rfCrEL5DsnLFjA2+89Et7u6BrP6648a9JmcgVCUvkzWJ0v05MuSTX7pDnrS4JK6/hMSOBYhGh0xzNO25wMwBUici1wA8BS/+16Xy7NGEJ96OxOlrLfh7ObBLvtoRTY7RmaK+tLuGVlcVGZ6EUg3u0J8VrDAwKeP/r3Xy4fneItg8YhUPW78znxC5tm65Wj4POR0pZ9WRgItfoO59nZ7suCWpR/bnktB5km5XTlDLkmi/IzbY7933lobpLKR5h6mWD7e9JSzXhNBQ3A8DNwB3An5VSW0TkBOC/8W2WJt6Es3s6O9q5qwyBrMb4QWVlpNkze7+KrMZYfOCood2DMVtfvGEPOVmt2bT3iH2MX4FEiNOs9KkA1c+miNfvY/bLDzByW76979pr/sLSPqclsFWhpHiEk7q15eudtb/fywr3c+PovoDx2VZWB5ZsTPF6SPEa/hqvV/juwK50zkxnYLZRrLAlm3AaipuSkF+LyL1Ajrm9BXgk3g3TxJdIPxprKTxxSK9G+0GVHq20k4I8QMGOMluDx+mbsFYmmMc6E7daAv/v01nc/cUce/tPY29h5oiJCWxReMSsuvXc51tcHZ/TMSNkErCvvKImtt/n55oRhvluT3kFi7/dS7XPH2CerI8JJxkDHhobN5nAlwCPA2nACSJyBjBVKXVpnNumiTPRfjSNaRN1JmaJR5i7ahtVPkOhc+plgyk9WmmbedwQ6bgmmsDL2M0ref7VmhpM7w8Yxe1X3J/wRK5weMQo1lJeUR1Qh9dJrw6t2HnouFHcRWBAt0zyd9Ro+giG2JtzhZpr5mU4JwENMU8mqzxzY+PGBPQQMAJYDKCUWmOagTQaV8xeXszC/J2MH9w9JCsXYMOucjsr1+dX+Mz91X7FlPn53Hr2CQ2OyAlX+CPZ6VO6g09m1ETyHEpvw9l3PJs0iVzBCEaH/tH63bYkQzj11JKDx7n8jB5kpKfwal6JUVLTgQIG92jPlY5VqLP6mnWvhpgnm6u0Q11xMwBUK6XKghxrTXEipXFJLJfGs5cX20U5LC2d4EFgYf7OgG3nTN2vFAU7DwXsEwwzQ22DgiXnAE3rC9u68jgfPPsTeh2qCWe96OZ/8U3X5J53GXWVjcfVPsU1I3sjwKwwkVVrth3k6mG9qfb5Qz5HjxhmQWsVmldUyqt5JfZnmOIx9J0aUsC9Jcf+O3EzAOSLyHWAV0QGAD8HvohvszSJIpZL47yiUh5/75uAfdMWbwrQ6bnv4kEhNVwvO6MHC77aaWfvjh/cnZVbD1BV7UcEcnu0p2u7VnxQiw/A15QC+QGU4rF3nuDq/A/tXT+/5Ne8ecp5CWyUe5zhnH5g0+5yOmSk2aG6Ts7o3cGsxyz4HU57wYjocXbIweJ9k4bn8OcrTm1QW7Xj2MDNAPAz4AGgAiMB7D3gT/FslCZxxGppbA0kwVo620uP8fSSQgCeXlLI6uJS7h0/yJb1tcxEN47uGyD9YBVsmbtqG2tLyvBIWbjbNll+sPZ9/u/df9rbzw+9hD+cP9mUHk1++ndtywmd2/Dh17vtmfqKreGrsQG8W7CLG0f3Zeplgw2pDr8iJSV88mH5saqAFVxuj/ahF6wHLTX234mbAWCgUuoBjEFA08wxHLLG0tjrbbiNtTZWbDW0+a8a2itAhwdg+8FjPPHRRqp9xmrk3AFdbF9BU5vcR+K0nd/y5ou/sLfXd+nL5T/8GxUp8SlMHitG9M0iz6HPv3nPYbbsO+La1GZNLu4a2z9qNa68olJmflYTTRSvmr0tFTcDwN9EpDswF3hZKVUQ5zZpEo21JA8TTx/NP+AUbNtx8FhYx6tXIDgPq7Laz+zlxcxdtY2XJ48GCAn7rKz281XJwYa9riSi49EyVv3rBjyOLvPMO59jR7uuCWyVOwTokJEWYGKzNJfcnu+0u0ebiS8r3B9wXa9H7PPiFcbZksJD3eQBjBWRbOAHwAwRaQe8opTSZqBmyLLC/VRb4mk+xWum5C7Urqfu7LStik74FUqBx2Ms3ScNN7T631iznb2HKwN+3FU+xaML15Oe6g0xHfkV7DoUmhHa1PD6fbz4v99zVlGNwvp1k/7EF33PSFyjouDBsOcH7PNIrf6XYLweQ/zfW0eNqVH9OpGe6qGyyo/Hkf0brzDOlhYe6qokpFJqF/BPEVkE/AaYgvYDNEssE5AVdz931TY72iKaf8B6riZ6B8SvuGZEDgr436ptfGXKMcyZPJr7Lh7E7OXF/HFBAcccnX00u3FT52efz+GXn82ytx8ecxPTR16VwBbVTkjnL/Vzrg/J6cCYgV3rPKuO5KyNVxhnvK6brKsKN4lgg4BJwFXAfuBl4JdRT9I0KYK/nM7i5z5TcTFc4Q2nf8AWbKvy48cK1RRye7Qnf0eZrbxZ6VNMfauAScNzmnzBFbeM2byKF159yN7+6MTh3Dbxd/g93sQ1qh64Cb2NxNqSMu4bP6henV84E1G8wjjjcd1kXlWIqqW+nYgsA+YAc5VSO2J6c5GLgCcw1HpnKqWiSkwMGzZMrVq1KpZNaPHkFZXa6pipXmGOwwZv/Qjcls6znis/VsUznxbiU5DmFc7o3SFkZp8SpeRibTSVjN7eB3fx6fRb7e0jqa04887nKWudmcBW1Y0UrzCgS1u+2V0etRRm/65tKdxzOGDFMKJvFiu3ltoSH2cN6Bzi6G8ITcUHMG3RJv76/gb8yvCB/eLCgTEpVF8XRCRPKTUseH/UFYCIeIHNSqkn4tAgLzANo+ZwCbBSRN5USn0d63slM3X9ssX6y+mU2q30KeatLuEvV5zKrFtHMW91SYgj1zkbC26L9Xfbi6tsR2+lT1G47wjBNCRGP9k7/1ZVx3n3uZ/R92BNgtvFN/2Tr7v1S2Cr6k7PrNZccmp3W6UzGDFHYq9HGHdyV57Zd8ReIgiGxMNX28vsVeHnm/axcuuBmM2A4xXGGevrJnPSWdQBQCnlE5FOIpKmlIp17NUIYJNSqhBARF4GLgNazABQ16VhPJaSwR28c9tSY5wXpiZAuLaAMaB8tD7QQbjvcOBXx5rBN40I9zqgFI+8+y+u+ep9e9c9E37JG7ljE9io+rO99FjUkphK1fgEZn62JSRax4rXz99exrrtZS1WdiGZk87cOIGLgM9F5E3Ansoppf7WwHv3BLY5tkuAkcEHichkYDJATk6ojkxTpq4OJzfH13WFMHFIL+bmldizk4lDegFG529F9IS7V3BbXltdEqLXEo5eHVrRqW06a0vK7EEg2Wf0brhh9dv86YOn7O2XvnMxv7/gziaTyBWJ4M8m+POqUfBUdoiv11QDnbqgoEbO2SNU+xUiYst9J6tjNB4ka9KZmwFgh/nnAWJpvAz3ywjpC5RSM4AZYPgAYnj/hFPXpWFtx9dnhTC0TxYPXZJrZ+FaIXZzV22zP4xwCWHBbbF03Gv7gDJbpQYoP6ameBh1QkeWOKQgmhLBSp0bO/Vmwo/+QUVqegJbFR9O6Z5J2fFqtpces/d5xPghez2CAvw+I9SzvKI6QM75/EHd+PibPfj8iqkLjFQia4BINsdoS8JNHsAfajumnpQAvR3bvTAGmhZDXZeGtR1fnxC2vKJS+4e4cusBOyuz2mHLvWpoaMx2cFvAWDVUVftRKjR80KL8eFVA+b8TO7cJ0AFqKnQ/tJelT90csO+CW6axsUufBLUo/oQr7jL5nH5ktk5lx8FjzFlhRo75jImAc4LQOTMdv1L2inJh/k6txpkEuAkD7YIR+58LtLL2K6W+28B7rwQGmNLS24FrgOsaeM0mR12XhtGOd4ZiOpfa0YhWGtL68V5pmoXCtcW6xqh+newB4ZMNe8LG86d4hQNBafwb95Y3KRNQiq+aTY9fHrDvgQt/wqzvXJyYBsUJj2nrqS1I91BFNfddPIi8olLmrS4J+M445ZyhZoKQGiTwl2yO0ZaEGxPQLOAVYAJGacgfAXsbemOz0PxPMcTlvMBzWmbCwE2oZbjnhvbJYsqEXENcSxlL7YHZmVEHmGilIWtbmYQzOd01tj87Dh4LGQCG983iy20HOVoZ2KVU+2gyvPC/BxmzJc/e/rjfMG65+qHENSiOnD+oG8erfLWuziw7brQKcxbBz0fTANI0Dm4GgE5KqWdF5G6l1CfAJyLySSxurpR6B3gnFtdqLriRW4hmNy09WolfKddL69pKQwa3zTmj+8eH34Zdxuf2aB/gLPSI4R/wNdFi7Dd8+Q5/ev/fAfv6/+oNqr2uEumTGjH/CEryGjuwKwOzM1m6eX9AZS/BkPXw+7GDBpzfi3Dx7c78kIKdh8jKSKt3GUdNbHHzDa4y/+8Uke9j2OnD2wQ0DcaN3EK0zj14Rp+VkRZQXzcc0X6IToE3Z1QHSlHlU7buj7V6yCsq5cE380MiRdaWND355lN2F/LOCz8P2Df6zufZ2a5LglpUfyJFWylCM3wtxc2hfbKYetlgHnhjXY0+IHD+yd043dTzB6JOSsLVc45UGEjT+LgZAP4kIu0x5B/+BbQD/l9cW9WCcSO3EO455yzMmtFnZaTx0FsF9vFzboscaRHOtORccXikJnO3yhHt4wHO6l+T4fnA6+tsyeZw9OzQip1lx5Nazjm9qoJ3XribEw+U2PtuuupBFp84PIGtqjtd2qax/0glflOMD1WjxOocEII/CxFs/9F1I3NYtGFPgPhb58x0e6Y/bdGmqJOSYI0oi1dWFusBIAlwEwW0wHxYBjTNjJYmRDT7e6TnItniH3h9na3JX2nG6ocbACKZlpwrDjCKtFvmJYsUrwSk99fWr6d6PUw+px/TI2SXJhSl+PP707h+zbv2rpnDLuNP59+WwEbVn72OBDxfkDd3WN8s1paUBQzmFkoR4D8aO7BrwAAw2FGQpbbQZKe4oJNu7VqhSTxuo4BuA/o6j1dK3RK/ZrVsoplkwj0XyTQU8sN2PLaiNiwnXrjzg3/cUyYY+QKfbdxnJ3FdPay3nTswb3UJ+8orSPHU1IYNZuv+o3ZFsGTiivyP+fvbNbmNs864iAcuvKvJJ3JFolWqlzm3jeK11SW8vLIYn7/GoRuc/OfM2/AQWJCltoABS1xwtqMusEfg9vNOjNq+lpQklkjcmIDmA58CHwJNKGaj5RBpFnblkF68umqbLfRmhXNaAnCWBlCKV0jxevD5/CGFOsJFbiwv3E+VT+H1GPefvbyYh97Mt6+X6hX6dGxN0YGjjfxO1J1gO//mjj35/k1PcDy1ec9Qxw/uDsCe8gp7daAwCq4rFejTeTWvxhSW4pWQWX5tztwrh/QyZEWCNP0jkczqmc0NNwNAhlLq3ri3pIUSi5lOtEieOZNHh9VSd9rpfT7FD0b2pmeH1mHNTtYMf9qiTYZtWASFotoPc1YU4xEJiBSp8imKS5O7829/rJxl/76J1tU1RWbOvn0mJR2yE9iq+CEYnfcp3dsxabhhe580fWnA5wZG0Z4Lc7Pt78G0RZsCCrJbK766UNeEx3hp8mtCcTMALBCRi82QTU0MCBdZ09CZTm0qnU5G9etEqlfsGbvXXB24ife3nMEWfgVKqZAok2D7vumDTHjSl8fv49l5UxlbWBPP/8Or/8CSfkMT2KrYIcDJ2Zkh8s1nO6SY84pKw3b+AJOG5wQ4Z2NVkL0uIZ/JrJ7Z3Ig4AIhIOTWijfeLSAVGSKgASinVrnGa2LyIFlkTi5nO7OXFTJmfj8+vSE+tUekMHhAeunQwvzePAyNLE7A7COfxzhmZUqEa/m469WQo+3LHsle575MX7O3Hz7mBJ8+8JnENigO3n2tIM3zz3oaA/ZbOE4TW2fUA/bq0oV+XtgzMrpH7yisq5RlHQXaAAoc/IF4ks3pmcyPiAKCUajpVK5oQ4SJrlFKuZzq1ZQlPmZ9vz+wqq2pUOoNXGaVHK7GKAVX7FLOWFzM3r4SHLskNWZU4JSYSPYOvD2duXcPsV35nby/p+x1uvvohfE2sIpcbDlVUc0FuNl6PBOg5WY7bvKJSdhw8RqpXqPYpPB7h1rNP4IWlWyncd4QlG/cGRIEF121oip+/JjLRVgB9lVJbozwvQE+lVEmkYzShhIusKT1a6WqmU5tzbFnhfvyOdb/HVGi0knAqq2pWGVkZaSHx35XVfl5ZWRxgf523uoSeHVpz0+i+PPNpYVLH7wfTs2wPnz9dE6xW5fEy4q4XKc2onxkj2QhXsN2K5Bl7clc+Wr8bvzKc8pZD1/r+pHg9XDOyt63XY33mFVU14cL2wG/6AFI8RNSFiiXaCdx4RPMBPCYiHowooDwM/Z9WQH+MfIDzgQcxVD01LqmLzo5TdmFZ4X62HzwW1Tnm/MF6xIi2AEfCD7B220HyikoDQvmcpKcY2u2Wk/jlFcV24Y+m0vmnV1fy5n/uYeC+mtDDCT/6B/nZjVuGL96EM6tt3F1ul/i0Py4zlNXZ0ft8fnp2aG1/f7wewW+eM3fVNiaaPiErVFRBVD9RLNFO4MYjmgnoahE5BbgeuAXoDhwF1mPo9/xZKXW8UVrZzKjNIRYwU/MIiFDtqyms4fOHNxmFG1ymLdoU0Hl/8PVulmzcy0W54aNdFEbHoqjJGsV8nPQR8Urxhw+f5ker37Z3/Xr83cw97YIENqpxCafCWu3z848Pv2X84O4R1WL9QVFc1iogEXo92gnceNRWEvJr4IFGaovGJGAG5DNiZyyd9WtG5NAjTLhmJKwfk2UGskxBi78NL+iaV1QacabfPiOVg0erwj+ZYC79ejH/fOtxe/vl0y7kvot+1mwTudwiGIP/ZxuNerw3je7LzM+2BKjFGqbDmnOCVwFuiGXilnYCNx5NX86wGeKcAXnNFYCVpBXtRxnJdjplQi6vrCymYEcZPr8xw4/UkUeTZ0jGzn/QnkIWPl+TyLW1Q3fG3/wvjqU170SuSAhGEtdVQ3sxuEf7gMztqmo/BTsPhajFjurXifRUD8eraoxKPr9ybXqJh81eK4U2DnoASEKCZ0CAq9nQssL99ky/osrP9E820zkznVfzSqg20z2D+/dWKR6OOxJ9RAiJ9U9G2h0/zOdP3UxmZU15wnMnP0NxVvcEtip+DMrO5Jtd4YvnCMbndunpPRjQLdP+nuQVlZK/o4zUlJos7/GDuzsyuSWg/sNrq0uYu2pbRBNjJLTNvumiB4AkJXgG5CZCaM22g3YHoYD3HQJekXB2/mCYCzxJp9JWgyg/M177ExdsWmHvu+mqh1h84rAEtqr+jOibxaY9hzlQy+rq2z2Hw3b+HuAsR5KXRbAf6ZoROUw0I3gUoUl51vdtoqOKl07cav64EYM7C1ijlDoiIjcAQ4AnlFJFcW+dJirOQhszP9sSNrOzPiRD0lY4bl3xGr9b9Jy9/fezruOJs5t2FdFWqV4OHqvdtOb3KzsAAIwQX8z8keDOH4IifvyKHh1aAzD1rQI7wsvp7LWoj+lF2+ybLm5WAE8Bp4vI6Ri1gZ8FXgTOi2fDWjq1OdXCFdpwgwh0y0xn16GK2g9OEkYXfcWcl++3t7/IOY0bJ/2xWSRyfb5pn6vwWq+ZsDXzsy34/AqvwJiTu9ElM50Nu8pDvivhCgNNmv5FiEqrdeuGOnG1zb5p4mYAqFZKKRG5DGPm/6yI/KghNxWRq4GHgEHACKXUqoZcr7lRm1Mtr6iUf3z4bZ07f4DLTu/B/iOVTWIA6H5oL0ufutne9iMM/+lL7G/TIXGNijHhaucMys7kO32ybN+NiDD25K4cqqjGb0pxVPsUH3+zx3boCtjSH1Zn7JyVP/3J5pDOP83UgNKJVy0XNwNAuYj8FrgROEdEvEBqA++bD0wEpjfwOs2ScE41a78lIFfnmb/5/538XZzYuU3M2xxL0qsreePFXzBo71Z732U3/pW1PQYmrlGNyJA+Wfz5ilNtGeW5q7bx0frdAXkgIoHFecLpSTln5XsOBabsZLdLZ9r1Q+1ckYY6cbV+f9PEzQAwCbgOuEUptUtEcoDHGnJTpdR6AGnhMdqRCLd8dwrIWbNAZwEPixQPZuJYeA2Xymo/63eV2/sHZWcGbCeaBz+czs15b9nb9170M145/XsJbFHjYyluWno81X5lZ+9aeSBOJVm/MpzB0Rywk4bnsLZknb398/NPimguqqsTN1jgcOplg3W5xyaCm5KQu0RkHjDA3LUPeD2urXIgIpOByQA5OS3jSxW8fA9W4xQxfvBpqR7OHdCFD77ebXfwQ3KyGNAtk1mOCkzR2LC7PClkHiasX8KTb/6fvT138Dh+ffHdLS6RK7jiVnDn7MwDsZK4sjLSatWTsjrkhfk7GT+4e0AH3VAnrvP76VeKKfPz7XKSmuTGTRTQbRgdcEfgRKAn8DSGFlC08z4EwukNPKCUmu+2gUqpGcAMgGHDhiWkm0rE8tYp3ZuVkWan8PsBlOEUnDIhl4HZmSzasMeO7FixtdSO+HCDX0H/rm3ZtOdwwP7MdC/lFfEvADdw71bee+6n9va29t343i1PcjTN/WtIFpxDVX2+qIIxqFvCbdZ3zuqcszLSbHNgfWQarhuZE3Fm7vy+ObfdMKpfJ3tlCkbEks4FaBq4MQHdBYwAlgMopTaKSNfaTlJKjWtg25KCRDnIgpfVt559AgU7D9lRIz6/omBHGdeNzKFPxww27T1in/tuwa463etYVWhHH+/Ov93xw3z29I9pV1HT7jG3TWdrx55xvW+88XqEDhmp7DscXmzPom26lyMVvoCB4rRe7ZlySS4bdpUzZX4+fqUCJLnj+T1syPd8aJ8spl422GizX9mDmCb5cTMAVCilKi17vYik0EJkwYOjbSx55HCrgVivEoKX1TM/28LUywazfMsBKqv9tl5Lbo/2bA2qvetM6QdCqnUFs730WJRnY4soP9Nf/wsXblxm77vlyil83H9Eo7UhXigME12/zm1qHQAGdW9HXlGpLb3hFZhySS5AYE0HRxBAPLNtG5rNe93IHNskpR3BTQc3A8AnInI/0FpELgB+ArxVyzlREZErgH8BXYC3RWSNUiqpPH3OGZHCsM16PWKH5jlnSVaRdav4+pzJoxv8Awi3rC49WslVQ3sxZ3mxIQ7nVyzM3xmg5OikY5s0MlI9lBxMDtHWW1bOZ8rHz9jbT5x5DX8/54YEtii2WDo8/btlhqhyDsrOZGB2Jmu2HeSi3GwuyM0OcZxaETkB1bqkpgh7PLNtY5HNq3MBmh5uBoD7gB8D64DbMaSgZzbkpkqp12lER3J9cM6IPAJn9e9MTscM5qwoDpklzVtdYtfXrfQppn+ymdN7d2hwofdwy+oPCnYZflGFre2ycusBqqr9iBAQ633oWCUHjkS8RVjapHk5Uhlb88/I4nW8Mue39vby3oO5ftKfqPY2HyWSFK8waVhvJg7pxYYwUVXrd5Wzee/hgMlBOMerJcxWWeXH46kZGCIdD7FZfeps3paJmyggP/CM+ddiCJ4R3TPuJADmrS4JmSUFx6l8tH43H67f3WBbbfCyesOucp5eUmg/f9PovrZT75WVxZQdq2Lr/hpzkK8emg6x7PyzD+1j2VM3Bewb9tOX2Nem+XQuHmDcKd24/bwT7dXgwvydYY+t9CnufXUtj151ekQnbrSOONzxsfRR6Rl8y8NNFNAWwpiQlVL94tKiJCHSDzF4X15RKQoCaqz6Y1jo3Rmd8X6Qc7dg5yHyikp56M18ewUSQG3G/ziRVl3Fa//9FYN3b7b3XXHD43zZ8+TGb0ycSUnxBHT+VmcciU17jzBp+he8cvuZEb8XdemItRKnpiG4WYM7ZRZbAVdjhIQ2eyLN0Kx9wYqL143MIbdHe6YuKIhqS3WzZA8WevP5FV5P4DHjB3dn+iebw3f+wMndMtmwu7xRY/wf+Hgmt618w96+/3t3MfuM8Y3XgEamqtqotnXPuJMCOmOLVG9NeU2Lar+xkoxFR62VODUNwY0JaH/Qrn+IyGfAlPg0qekQTnExOOGmPkt25zHOzsTnr6nNm+oVivcfiSr53LZVSqN1/uO/+Yyn5j9ib8/LHcsvv/+LZpfIFZwzoaiptjVlQm5A9TUwvhfhFmINeVeCJxDadq+pL25MQEMcmx6MFUFm3FrUhAg3+8orKrVT9FduPRCSEelmyR5uJhmM369qjfdfGaY+bKwZsLeID567y97e2bYT4259iiPpGXG/d2PhNc16qSkebjnrBOPzrfKDORhb5r7So5UhhVVEhOqg+gpej9ja/HUl0gRCd/ya+uDGBPRXx+NqYCvwg7i0pokRqQh7tA7ezZLdOiY4nv/k7Ew27ztilIr0ejijd4cAp29jkllxhE+m30bHY4fsfd+99WkKO9WvY0tWhvfN4orv9AqQWnBKMASb+4ILq2RlpAXE9QNMGt673h22tvlrYokbE9DYxmhIc6G2Dt7Nkt065t55XwWYG9q2Sqkp2qsUR2McrukGUX6enP8o39/wub3vtom/44MBoxq9LbEknLCeYMy4120vCzDVBc+4F+bvJLd7u7AyDXlFpXz35K589M0elJnZe2U9Z/9QN5u/VujU1EbEAUBEfhHtRKXU32LfnKZFpOW4mw7e6hymLdoUVsxraJ8s+nVuEzAAlB6ppNqMMPL5FYX76hjk30B+lPcWf/iwRsH7ydE/4PFzf9iobYgHp3TP5Nvdh0PqIBuZvZFn2rOXFzNlfj4+v+LTjftCNPmDgwSuHp4TIOZWH9za/LXGv8YN0VYA2s4fRPCMKtJy3I1NNtjRG67z2B2k4d6vS1uKDxylyqcQj7B13+HwF48xw0oKeHXWvfb2qp6DuObah5tNIleq12NLbAdjZfcGz7TzikpDTDsKQ64hXFSQFSQQi064tu+XJWGiTUWa2oj4C1ZK/aExG5LshJtRNSQEL9jR68wbAOxyj1DTCY0Z2JUP1xvSz74IoZ+xpGv5flb8O7D42/C7XmRv2+YTBZzmFUb368TakjJ7n+X0TfEKVw/rzZVhZu3LCveHrBgEwykcHBXUmCGawaVCxXw9OjxUEw43UUCtMKQgcjHyAABQSt0Sx3Y1Km5speFm+3eN7V/vEDxr8HAW9PB6PWw/eMyQlnBoEJ01oDP3jDuJ37+xLmDAiBepvirmzvoNZ+zcaO+beP1jrO41KI53jT8ZaV6OmX4Tr1fI7d6OScNzKD1aaYdqCth+FoGwnT+ESjZYaq2fbdxnD+b5O8q4ckgvVJTrRKK+9nvre2p9PxQ0u1BcTexws4Z/CfgG+B4wFbgeWB/PRjUmbm2lkWb79Q3Bc9pyszLSKNhRxtxV23h5RbFd+q/ar/CIMH5wdwC+3hmoMROPRN/7Fj3HHStes7d/d8Gd/HfI92N8l8RQUW10/h6zP1y3vYwNuwuYMiGX9FSPqadUU3HNV4uu/cQhvRDzv2W2s3SZgoUD6+L4bYj93vqeOnMRqn3aBKQJj5sBoL9S6moRuUwp9R8RmQ28F++GNRZuw+pimXDjnN3dNbY/ANMWbQoo/Xf+oG58/M0efH7F1AUFnDugS8h1Ytn5f2/DF0x/4y/29vxB53H3Jb9qVrNHv994z/wK8KmQ+P1IoZ1O8opK7Tj/ar8R1WPF9Du/IzsOHgsrHOiGhoR6Wm2Y/slmO0nQryArI83t26RpQbgZAKrM/wdFZDCwC+gbtxY1MnWx48ci4cYpHe31GLVaJw7pFdKOzpnp9ky0ospP4d74OHxP3LeNj569097e3bYj59/6NIebUSKXYGROI4LPZ8zOrcfB8ftARF37YPs6RC7EnldUGlY40A0NlXcY2ieL03t3sEuFBpeZ1Ggs3AwAM0QkC/g98CbQ1nzcLGiMVHrnjN8pHV3th1nLi5m3uoRZt44KaAfAq6u2UWnOVDfvjW3IZ9uKoyx6ZjJdjhy0951/61Ns7tQ7pvdJBD2zWrPz4DGUKZl91dBetgnG+f5G+swjDfTB9vVIEULWNaZMyA2bI1AbsfhOjurXiVRzEEnRGkGaCLgZAJ5XSvmAT4BmqQAai5n97OXFIQW3w5kLTuvZPuRcp1PZ2Y4xA7vay/iYmXuU4l9v/h+XfPOpvev2K+7nvZPOjNUdEs4lp3bngtzsiJLK4R67wTkz93qNgWVwj/ZhO3enJEi4HIHaiIm8gyNpUKMJh5sBYIuIvAu8AnyslP42BTN7eTH3v74OgE837gMMM0KwuaCyyk9e8cGQ8yOF6XXOTI9pO29cvYA/fvC0vf30yCt5ZMzNMb1HvMls5aX8eOQMaAFeWLqVC3Kzbf9KrAiemQMRnbXRwnwbwxm7rHB/QNKgdgJrwuFmABgIXIJRHP45EXkLeFkp9VlcW9aECC4AMmPJZs7s3znEXODx1JR4xLH/6mHhtWGuHNKLV1YW16uwi5OhJV8zb9Zv7O0vuw/kB9c/QpU3tWEXTgCpXi8QeQCI1NHGShbBOTOPpvsULsy3MeWatUy0xg1utICOAf8D/mf6Ap7AMAd563tTEXkMY1CpBDYDNyulDtb3eonC6lRyu7ezZ/4ARfuPsqOshBSPGDr+Xg9jTuqCAj75di/V1X78GOGIziiScHRv16reNX27HD7AymmBUg0jfvIf9mQ23c7gwJFAZ6YH8GMUVfd6PQGOXYt4ySJE62SDw3yDpT7ijZaJ1rhB3Fh0ROQ8YBIwHlgJvKKUmlfvm4pciGFOqhaRRwGUUvfWchrDhg1Tq1atqu9tY0pwp3LT6L68W7CLov1HURgd0jUjcujRobUdWmhrwgzrTW6P9hTsKGNveQWdM9MZ3KN9QCcxe3kxv3MkftWFFF81r8y+j6E7vrH3XXX9o6zqlRu7NyAJ8HpARPD5DKnmhy7JDeloLVkEK0HLI/DLCwfGzDykBdc0TQERyVNKDQve77Yk5BqMVcCvlVINDkdRSr3v2FwGXNXQazY2ThtvZbWfzNap/PUHZ3DtM8tsJ6GVIOQ0FViaMAOzM5nyZj7VDkkHa0UwZUIuv3t9HfWx/Pz6k/9w17K59vaUcbfz4tBLYvCKE4/XUW4TjLh+MO3cPiOe39mxW4O0U1Y71jHxWotf05Rx4wM4XSl1qPbD6s0tGA7msIjIZGAyQE5OThybUTeyMtLs2XlApxIm8iIrIw2PCChlmwrmrS4J6Pyt6xyv8vP4e9/UufO/YOMynnntT/b2goFn87PLfoMST5SzkpdgeWaPGDr6AnaxFa/XA0rh86uwdm5rkHZSn5h4PcvXNFfc+ADq1fmLyIdAdpinHlBKzTePeQCjyMysKPefAcwAwwRUn7bEA6d+jNWphIu8AJi6oAC/MgrGT5mQy9A+Wby2uiTitQ8crYr4XDD99pfw8cw77O19Ge0ZO3kG5elt6vfCEojXIyilSPF6OO+kLizesMeup5viEVtPxyq2Uls8v+2IrQr0udTFIapllTXNmbjp+SqlxkV7XkR+BEwAzm+KoaVZGWmGScIs8mF1KsFOQaepyK8Ur6wsZmB2JpnpgW99ZrqX8gr3BV7aVBzlo5l3kH34gL1v3I//zabOybNKqgsj+mZx7/hBdme+rHA/H62vqXc8qHs7+7HT7JJXFLnsZSwcsboCl6Y5kxBBdxG5CLgXOE8plZiahg3ASvLx+RVex6weCIm82LCrPMCRu7akjGtnLA3o0AD3nb9S/H3BX7ni68X2rjsvu4+FJ5/d0JeVUPKKDzJvdUmAaqZT1OyrkjKun7ksYAYeaXbuNNlYBNdmdosOp9Q0ZxJVEexJIB34QAyxsWVKqTuin5I4IhWCMSpGqQCbcrBTMJy9ucqn6NquFVAW8lw0rluzkL+8N83efmb45fz5u7fW+fUkIz6/Ys7yYl7NK7GlG6ZMyOUBM8HOKrbinIGHm51DTXJWiukjsLKw62O+0eGUmuaMm4pgA4HhGDpAYMTvL2nITZVSsU3RjCN5RaV2ZE9qioc5t0UvBBM8+9x+8BgpXglx+LZJ89I2zcthF3V9v7P9G17/76/s7a+y+3PV9Y9RmdL0ErmiYXXyc5YX89rqEs4Z0CWwRq8EZkyH+xyCBwXrug0x3+hIH01zpdaKYCLyPjBEKVVubj8EzI10XnPjNbM4Cxid06ML19O/WybnDuhCl8z0gBqvATVgHbNPj0D/Lm2o9iu2mnkCb6zZUeu9Ox8pZdWTNwbsG/mTF9id2TnmrzPepHiEAV3bsn5XecRjLKe61WHvCSqJef7JXUO0fMLNzp16PdGihILR0T6aloYbH0AORsauRSXNSA66NoK90yu2lrJiq+F4DM7ijTT79CvYVAc1zxRfNbNefoCRJQX2vh9c9wgreg+u9+tobFI8cFK3TG4Y1ZfSo5VkZaTx4Jv5EY8f0TeLAd0y7RDP1BQPk4bnsH6Xoc3v8QhjBnYNOS94dh5Or8dNp97Son30YKcB9xXBVojI6xj92RXAi3FtVRJx5ZBevGwW9ggm2CbtNEl4PEaGal3Dm36x5CV+vrQmLeIP59/G88Mua8ArSAzVfti057DtfH3g9XV2SCdAl7Zp7D9Sab+va0rKuHf8oIAQT+t9nTI/3y6M48aZG25QqI2WFO3T0gY7TWRqzRJSSv0ZuBkoBQ5i6Pb8JepJLYi12w7aoYjW7POaETlGgfA6XOf8TcvZ+ugEu/N/96TRnPCbN5tM528VH3dS6VPMM/MdggfCvYcrcR7tc5QtdMpilx6txGfmVlRW1Th6Y401eHulcUXbEkEk57mm5eE2DDQDOKSUel5EuojICUqpLfFsWLIwb3VJVD2e97/ezZKNe+1ZlBUlVO1SxKfvge0sfuZ2e/tgq7ace/tMDrVq29CmNypeD0y9bDDPfb6FTXtqqpdZnfzgHqF1EJQyfANKRbbRZ2Wk1Ug/EL/Shi0p2keHtmos3GgBPQgMw4gGeh5IBf4LnBXfpiUOp33UTUXciio/81aXhEoBmxmoTtqmezlc4SOj8hjvP3sXvQ7tsZ+78JYn+bZL35i9jlhiFTS5yCy04vV62F56zH5+0nAjAW2Lo3RlqldsH0np0Uo8gj2YCpCWaugeRUvQcp7nkfiWNmwp0T4tabDTRMfNCuAK4DvAagCl1A4RyYx+StMlnMqns+NyKlBaM1MFvJpXk8Rk/cDmrS4J8R+MO7krZ//lN1yV/5G9765L7+XtQec02musjTvO7Ud5RTWzlhcDRmd99oDOjB/cvUbV1OshxVujxJnbo71hq3d08M46B+GqaTmTviKhZ6vxoaUMdprouBkAKpVSSkQUgIg0PZGZOhCs8vnMZ1sCZq1/vOxUu2j42m0H7cLblg3buobVUTk7/2u+eo9HHv2Xvf3c0EuZOm5yI70ydwiQ2TqVnE5tSHFIXdwz7qSA96ba569ZHSlFwY6ygGI3XlO7x6K+s86GzlZ1tItGExk3A8D/RGQ60EFEbsNQ75wZ32YlDueMU0QCbPkKKNhRxnUjc2zJgSUb99qz06yMNK6dsZQqn7JXCgCn79jA/Jd+aV9nfdd+XH7j41SkxMee3RC8HrHrF4STunC+N5Zz1vpvVcDyiDD1ssGui63XRn3P09EuGk103KiBPi4iFwCHMPwAU5RSH8S9ZQkiWEDs92YIooWKcKwl8Vxp2kCq/dDpSCl5T94QcP2vvviKqh69OO+TzXy0frdt20ZocOnHuuI0bWE0gamXDab0aGVYqYvg92bqggJ78LtyiGHSSabZdksK7dRo6oMbJ/CjZrWuD8Lsa5YEzzitylwpHgLMGsHHWhLPXr+PWS8/wKhtNYlPM/8wk+/8aKJ97Om9O9jmI7+CEzu3oXDvkXoVgQlmUHYmm/cetgejcM8P6ZNFbo/2PPRmPlU+I1v5j5efynUjc8grKrWd2CISEHnjfL2WKczZ4SdTB6v9BxpNdNyYgC7AUO50Mj7MviZDXezCA7MzEVOjoLbAzolDepH990f42aez7X1/GvtjXhw9kTk/Gh0gGbF4w54AJ7IzU9gDnNqrPUcqquuUQWwxpE8WRyt9FB0IFVoVYMLpPezKWZE68SkTcpkyPx+/ipyAleyORB3totFEJ5oa6J3AT4ATReQrx1OZwBfxbli8qKtd+OlPNtumGZ/f2H7mhyGlNeGddxj6/e8z1NxcM+Q8rrzgl/jw4DWLw1h+g0nTv6A6ylQ/JcVD13atKDtaiew7QrhqCQJ4PMLQnA6sLi61r+c1U/vKK6rDXjtc8fJI4Zd+pZq8+STZBymNJpFEWwHMBhYCDwP3OfaXK6UOhD8l+amrXXhNcWnUbTZvhv414qaH0ttw9h3PciSjLR4R/D6jdKHV6c5bXRK18xeMdn3w9e6Ix3g9cP7J3eicmW6bpOatLmFfeQWLv93LnBXFSIQEhquG9grRzA/3+rX5RKNp/kRTAy0DykTkCeCAQw00U0RGKqWWN1YjY0ldO7aKoN76iCXffOQIFYNySd9WZD/3hz/O5vnDZqEXP/hNI4/f77f9A7UllkUyM/Xv0gZE6JiRSvuMNBZ/u5dqn3HdWbeO4i9XnMq0RZv40HQsi4I+HTMCzEBpXiM0080qSJtPNJrmjxsfwFPAEMf2kTD7mgx17dhOzs601T8BjldWs+8H19N57mzSzX0/v+TXvHvqGHIy2sDhmkxYqzOv9sOs5cXMzSvhljP72rLHdeGWs/sxMDuT62cus6tkQeAqZlS/TqR4PXYET0npUS48pRuAvVoY2ieLaYs2uVoFafOJRtO8cTMAiLNmr1LKLyIJKSUZK+rSsV3+nV72AHD1V+/z2MJ/2s+9MPQSHjp/MoggPkW/zm0CdHBSPIbfwHrzKqv9PPNpYcTOX4Sw9n6A5z4rpE16CserAlckTvPS0D5ZnHdSF9t85FOGVlGr1MBZfkBWrkfYcfAYeUWlurPXaFoYtaqBAoUi8nMRSTX/7gYKG3JTEfmjiHwlImtE5H0R6dGQ68WDvKJSpi3aRMGOMk7duZGtj06wO/+jJ5/C6m928Ofv3YFlbFfAmIFd+csVp3LOgM785YpTeeX2MzmtV6AIWoTITC48pRvD+mRhCWoGm4o27T3C2pLQEpI+n58NjiIrXTPTQ46pqKoxQUHNKmjSiBwQYc6KYq6fuSxqgXWNRtP8cDMA3AGcCWwHSoCRQEP1Cx5TSp2mlDoDWABMaeD1YoplI3/+jRX84cozeOvF/2c/d/adz7H+3c8YMrA7Vw/rbe/3YETODMzOZFS/TnbYpCWS5sTZuXs9wh3n9uPjb3azcmupnZjl1kTkU4ZevtV5TxzSi7SUwI9VAXNXbQvo4If2yaJnh9ZU+2pkL2qTBbYGRes6wdsajaZp4SYTeA9wTSxvqpQ65NhsQ91N4nFl+cbdPPvSbzmraK2977pJf+KLvmcA2Br3Qo2ZJ8VrJExZ9YNTvBIwQDixhNUsFcxHF66PGBlkhHtGzxL2q5ow06F9sphzm+HjWLPtIB9aWkWOUFSLrIw0e8Dxq+hSy8GO4ykTcm1hOC2zoNE0TaLlAfxGKfV/IvIvwnTQSqmfN+TGIvJn4IdAGTA2ynGTMVccOTmhs+mY88c/8pMpNQuSh8fcxPSRVwUcsml3Odc+s8yuFWw2lMUb9tj7qnyKOcuLSU3xhEgujB/cnetGGq9l9vLiACdzMIIR8rn70HHSUzysLi7F5zf9BQDK0OAJF9ufV1TKpw6touCIp9KjlbZD2lrBRCI4fHZh/k4ts6DRNHGirQDWm/9X1efCIvIhkB3mqQeUUvOVUg8AD4jIb4GfAg+Gu45SagYwA2DYsGHxWym89x5cdJG9efC7F/Lzax7ks8LSkOHvwNGqwM4fwxa/O6iIucJQzXSeLgR2tK+sLA5pyuVn9GDBVzvx+xUpKR4+/mY3Pr+hrz/1slPJ31HGq3klVPuM0pNOsTYnVkbvwvydjB/cPeSYUf06kZ7qLiQ2OHx2/ODuLC/cbwrfic4T0GiaINHyAN4y//+nPhdWSo1zeehs4G0iDABxZ8sW6NevZrtNGygupkPHjtxdVMqKmcvswi6CkUl7QlC0D+b+0f06hXXUOiN7gjvLbu1aYSyCDEb0zeIf13yHG0f3tc04VlRPpU+Rv6MswHYvqIgz97yiUttMs3LrgRA5h7qExIYrtm5rZETKOtNoNElNNBPQW0SxzSulLq3vTUVkgFJqo7l5KfBNfa9Vb44ehTPOgI0ba/Z9+aWxz8TZ6ZUfq6Jg5yHGD+7OwOxMPtmwhyqfQoBxp3Tj9vNODIi0AWPACOj8hRCZ5NvPO5FF5rW8HqF/t0w7JNMqph58TbfJbG6ynusSEus8dtqiTfbqxlnPV6PRNB2iRQE9DvwV2AIcA54x/w4D+VHOc8MjIpJvagxdCNzdwOu5RymYPNmY6Vud/0svGfsdnf/s5cXc+OxyNuwqZ1S/TrywdCufb9rH1AUFADx06WCjCLrAko17AdhbXhF4KwJHUK/Xw8DswGJqQ/tk8fLk0UbCljL8Bs6QTCuqRzBs/RPNZK5Zt47iFxcOjOp8jWeh85ZURF2jaa5EMwF9AkbMvlLqXMdTb4nIkobcVCl1ZUPOrzf/+Q/cdFPN9p13wrRpISaM2cuLud+ceX+6cR8XntLNzr6trKoJlwwWS+scJgbfSXXQTNnS4yk/VmVLOOC4hzOq57XVJSG1CGqbccdTzkFLRWg0TR83Gb1dRKSfUqoQQEROALrEt1kxZts2cEYQDRoEeXnQunXYwxfm7wzYXlVUane+fmDttoOMGdg1xAzzQcGugPO8YnSUVpSPM9Qyr6iUa2csDavZ7wnjVJ23uoTK6hrtn/qYbZzEolSilorQaJo2bgaA/wcsFhEr+7cvcHvcWhQP5s6tebxlC/Tta2/OXl5sR8lYoZnjB3fn04377GMOHAl0sn7w9W6WbNzLiL4d+Wp7GWNOMsbDmZ9tCTjumhE59OjQmpVbS0NCLZ3Vw4KZcFpgxE6sK1vpUokajQbcJYK9KyIDgJPNXd8opSqinZNsrL7kevIGjGHIaScEdHTBph6A60bmMDA7k1SvUBWhg1bA8So/S8xz3lizgx0HjwUURU/xCBNNqeYU81peb83MPlrczIKvdnLj6L5htXtiYW/XpRI1Gg24kIIQkQzg18BPlVJrgRwRmRD3lsWIvKJSrnkhj798vpNrZywNkC0INvVY28sK9wfUARaMTrxL28iZsl+VlNnOWo/ArWefAMD0TzbbA0mVT9m6PU7nrlFAvuZafjNr18Kt09ct2oGr0WjAnQnoeSAPGG1ulwBzMTR8kh6nqaXSp5i3usTuQINNPeMHdweC1DK9Hs47qQuLN+xh7+HImbLHq/12ApfPr3ju8y0898XWkISxhfk7uW5kToBkw6h+ndiwq9wowehXpKWGdsqxtLdrB65GowF3A8CJSqlJInItgFLqmEjTyfwJbqhz27L5B/sAgjvI11aXRDQHOVny7V58foUC8/jQc6xBxrqPsw5vuPq88UI7cDUajZsBoFJEWmP2ZiJyItBkfAATh/Ribl6JbT+37PIW143MsTt+J84Ocl5Qgld2u3RG9evEG2t2BOw/eKzK7vK9HvB4PPhMuYZTurdj0vDw9wp3T41Go4k3bgaAB4F3gd4iMgs4C7gpno2KJcGmlvp0sIN7BGr6//z8kxiYncnb63ZGXBkMycni3vGDtJlFo9EkLVEHABHxAFnARGAUhgXlbqXUvmjnJRu1zawfeWc97xbs4qLcbO67eFDI86VHK21FT48Y28GOYo+AOEp6rdhayoZd5dw1tn/I9TQajSYZiBoFpJTyY0T/7FdKva2UWtDUOn+okXWYvTxUefORd9bz9JJCtu4/ytNLCnnknfUhxzijZizp5VH9OuFxukIUIVFCwVFGGo1Gk0y4qQj2gYj8SkR6i0hH6y/uLYsRVqz/pxv3cf/r60IGgXeDsneDtyF8GObQPllMvWwwKR7BA2Ejd3K7t4v569FoNJpY4cYHcIv5/y7HPgX0C3Ns0hEu1t/piD2jdwe27j8asB2OcGYkK2nMsvMvK9xvF1gRILN1aoxehUaj0cQeN5nAJzRGQ+JFpFh/izbpKVG3ayN4YHBbYEWj0WgSTa29nYi0An4CnI0xuf0UeFopdTzqiUlCpFh/i+AYnoaUHNMJVhqNpikhSkXv8kTkf0A58F9z17VAllLq6ji3LYRhw4apVavqVaEyIpYqZ5VPkeoV5kwerTtujUbTrBCRPKXUsOD9buwdA5VSpzu2F4nI2tg1Lf5Ekz4e2ieLOZNH61m7RqNpcbgZAL4UkVFKqWUAIjIS+Dy+zYodbqSPdQauRqNpibgJAx0JfCEiW0VkK7AUOE9E1pklHZOacNLHGo1Go3G3ArgoXjcXkV8BjwFd4pVgFmstfY1Go2kuuAkDLYrHjUWkN3ABEJqeG0OG9sliyoRcOwpIm3o0Go3GoG5B77Hl78BvgPnxvEleUSkPvVVAVbWf5VsOMDA7Uw8CGo1GgzsfQMwRkUuB7WaFsdqOnSwiq0Rk1d69e+t8r9fMYuoK7KLqGo1Go4njCkBEPgSywzz1AHA/cKGb6yilZgAzwMgDqGs79pZXRN3WaDSalkrcBgCl1Lhw+0XkVOAEYK1ZWKwXsFpERiilQpXYGkjnzPSo2xqNRtNSaXQTkFJqnVKqq1Kqr1KqL0aN4SHx6PwBrhzSizSvIECaV7gyqCKYRqPRtFQS6QRuFHSmr0aj0YQn4QOAuQqIKzrTV6PRaEJJSBSQRqPRaBKPHgA0Go2mhaIHAI1Go2mh6AFAo9FoWih6ANBoNJoWih4ANBqNpoVSa0nIZEJE9gJu1Ek7A3GRl44Run31J5nbBsndvmRuGyR3+5K5bVB7+/oopboE72xSA4BbRGRVuPqXyYJuX/1J5rZBcrcvmdsGyd2+ZG4b1L992gSk0Wg0LRQ9AGg0Gk0LpbkOADMS3YBa0O2rP8ncNkju9iVz2yC525fMbYN6tq9Z+gA0Go1GUzvNdQWg0Wg0mlrQA4BGo9G0UJr9ACAivxIRJSKdE90WJyLyRxH5SkTWiMj7ItIj0W2yEJHHROQbs32vi0iHRLfJiYhcLSIFIuIXkaQIzRORi0Rkg4hsEpH7Et0eJyLynIjsEZH8RLclGBHpLSKLRGS9+Zneneg2ORGRViKyQkTWmu37Q6LbFIyIeEXkSxFZUNdzm/UAICK9gQuA4kS3JQyPKaVOU0qdASwApiS4PU4+AAYrpU4DvgV+m+D2BJMPTASWJLohYPwAgWnAeOAU4FoROSWxrQrgBeCiRDciAtXAL5VSg4BRwF1J9t5VAN9VSp0OnAFcJCKjEtukEO4G1tfnxGY9AAB/B34DJJ2nWyl1yLHZhiRqo1LqfaVUtbm5DKNuc9KglFqvlNqQ6HY4GAFsUkoVKqUqgZeByxLcJhul1BLgQKLbEQ6l1E6l1GrzcTlGR9Yzsa2qQRkcNjdTzb+k+a2KSC/g+8DM+pzfbAcAEbkU2K6UWpvotkRCRP4sItuA60muFYCTW4CFiW5EktMT2ObYLiGJOrGmgoj0Bb4DLE9wUwIwTSxrgD3AB0qpZGrfPzAmuf76nJzwkpANQUQ+BLLDPPUAcD9wYeO2KJBo7VNKzVdKPQA8ICK/BX4KPJgsbTOPeQBjiT6rsdpl4aZ9SYSE2Zc0s8SmgIi0BeYB9wStjhOOUsoHnGH6wl4XkcFKqYT7U0RkArBHKZUnImPqc40mPQAopcaF2y8ipwInAGtFBAwTxmoRGaGU2pXo9oVhNvA2jTgA1NY2EfkRMAE4XyUgWaQO710yUAL0dmz3AnYkqC1NDhFJxej8ZymlXkt0eyKhlDooIosx/CkJHwCAs4BLReRioBXQTkT+q5S6we0FmqUJSCm1TinVVSnV1yw6XwIMaczOvzZEZIBj81Lgm0S1JRgRuQi4F7hUKXU00e1pAqwEBojICSKSBlwDvJngNjUJxJihPQusV0r9LdHtCUZEulhRcCLSGhhHkvxWlVK/VUr1Mvu4a4CP69L5QzMdAJoIj4hIvoh8hWGqSqbwtyeBTOADM0z16UQ3yImIXCEiJcBo4G0ReS+R7TEd5j8F3sNwYv5PKVWQyDY5EZE5wFJgoIiUiMiPE90mB2cBNwLfNb9ra8wZbbLQHVhk/k5XYvgA6hxumaxoKQiNRqNpoegVgEaj0bRQ9ACg0Wg0LRQ9AGg0Gk0LRQ8AGo1G00LRA4BGo9G0UPQAoLERkQ4i8pMGnH+PiGTEqC2L46n0KSKtReQTM81/TH2UFOPQpjOSLAQyLCJy2PzfQ0Rerec1PhSRrNi2TFNX9ACgcdIBqPcAANwDxGQAaARuAV4z0/wbhIjEKqP+DCDpBwALpdQOpdRV9Tz9JRr2XdPEAD0AaJw8ApxoJuM8BiAivxaRlWZtgD+Y+9qIyNumRnq+iEwSkZ8DPTCSZhY5Lyoi40Xkf47tMSLylvn4KRFZFU1r3Zpxmo+vEpEXzMddRGSe2b6VInKWuf88R1LRlyKSGeay1wNOTaG2IvKqGHUQZpkZqojIFPPa+SIyw7F/sYj8RUQ+wdBz2ioiHvO5DBHZJiKpInKiiLwrInki8qmInGwec7V5zbUissTMIJ4KTDLbPSnoPehrnr/a/DvT8V4ujtD2883Xv06MmgDp5v6tZtuXmu/9EBF5T0Q2i8gd5jFtReQj817rRCRE3dRsU7752CtGHQnru3K7ub+7+frWmK/3HPP0N4Frw33emkZEKaX/9B9KKYC+QL5j+0KMYtOCMVlYAJwLXAk84ziuvfl/K9A5zHVTMGoytDG3nwJuMB93NP97gcXAaeb2YmCY+fiw41pXAS+Yj2cDZ5uPczDkBADeAs4yH7cFUoLakwbscmyPAcowNHw8GFmzZzvbZz5+CbjE0b5/O56bD4w1H08CZpqPPwIGmI9HYqTrA6wDepqPO5j/bwKejPDZZACtzMcDgFXR2o6hDbMNOMk87kUMoTXrc7rTfPx34CuMzO8uGOJi1mfWznzcGdhETeLo4eDvCzAZ+J35OB1YhaHH9UsMAT/rM850vKaNQKdEf+9b8p9eAWiicaH59yWwGjgZo/NZB4wTkUdF5BylVFm0iyhDKuFd4BLTXPJ9ambfPxCR1eY9cjEKqrhlHPCkGFK9b2KIYWUCnwN/M1clHVRNbQOLzsDBoH0rlFIlSik/sAajcwMYKyLLRWQd8F2zjRavBD22Zu3XAK+IoXB5JjDXbON0DGkBzDa+ICK3YXSMtZEKPGO2Yy6B71O4tg8EtiilvjWP+Q/G4G1haRWtA5YrpcqVUnuB42Jo3wjwFzEkED7EkLfuFqV9FwI/NF/ncqATxndlJXCziDwEnKoMzX+LPRirRk2CaNJqoJq4I8DDSqnpIU+IDMWwVz8sIu8rpabWcq1XgLswCpOsVEqVi8gJwK+A4UqpUtO00yrMuU69EufzHmC0UupY0PGPiMjbZvuWicg4pZRTwOtYmPtUOB77gBQRaQX8G2Mlss3sxJznHXE8fhPjvegIDAU+xij0c1AZVd8CX5BSd4jISIzBcI2IhBwTxP8DdgOnm6/7eLS2E16i2ol1jj/ofL95/vUYK4KhSqkqEdlK+M/GQoCfKaVCdJlE5FyM1/mSiDymlHrRfKoVxmehSRB6BaBxUo5hCrB4D7jFnMkiIj1FpKsY9YuPKqX+CzwODIlwvpPF5nG3UTNzbofRiZaJSDeMkorh2C0ig0wb+xWO/e9jiLBhtu8M8/+JylCEfRTDFHGy82JKqVLAa3bw0bCe32e+BxEdnsqoGrUCeAJYoJTyKUPXfouIXG22S0TkdEcblyulpgD7MOSko71/7YGd5iz/RmpfNXwD9BWR/ub2jcAntZwTfL89Zuc/FuhTy/HvAXeKIe2MiJwkhq+oj3mdZzBUP4eYzwtGvYetdWiTJsboAUBjo5TaD3xuOuseU0q9j2FnX2qaHl7F6KBOBVaYy/0HgD+Zl5gBLJQgJ7B5bR+GD2G8+R9lVGv7EigAnsMwi4TjPvOcj4Gdjv0/B4aZTsevgTvM/fdYDlaMGWa4imbvY9jKo70fB4FnMMwkb2CYM6LxCnADgaah64Efm20poKZU5GOmczUfo7bxWmARcEo4JzDGSuRHIrIMOInA1Ue4th8HbsYwP63DmNnXRdV1FsZ7u8p8DbVJIM8Evsaou5GPYe5KwfBRrBGRLzF8R0+Yxw8FloUxz2kaEa0GqmmRiMh3gF8opW5MdFtaIiLyBPCmUuqjRLelJaNXAJoWiVLqS4yQVTcOWE3sydedf+LRKwCNRqNpoegVgEaj0bRQ9ACg0Wg0LRQ9AGg0Gk0LRQ8AGo1G00LRA4BGo9G0UP4/2z65DOKbh3IAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot it \n",
    "plt.scatter(y_test,y_pred, marker='.')\n",
    "m, b = np.polyfit(y_test,y_pred, 1)\n",
    "plt.plot(y_test, m*y_test + b, color='red')\n",
    "plt.xlabel('test values (harvest anomalies)')\n",
    "plt.ylabel('predicted values (harvest anomalies)')\n",
    "# check all other things\n",
    "n_pred = len(y_pred)\n",
    "slope, intercept, r_value, p_value, std_err = scipy.stats.linregress(y_pred, y_test)\n",
    "print('n predictions: ' + str(n_pred))\n",
    "print('slope: ' + str(slope))\n",
    "print('intercept: ' + str(intercept))\n",
    "print('r: ' + str(r_value))\n",
    "print('r²: ' + str(r_value**2))\n",
    "print('p-value: ' + str(p_value))\n",
    "print('standard error: ' +str(std_err))\n",
    "plt.savefig(r'C:\\Users\\Florian\\Desktop\\Jupyter_Skripts\\05_ML_crop_vulnerability/figures/scatterplot.png', format='png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c06d16e6",
   "metadata": {},
   "source": [
    "## now implement shap for variable importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2cc8cee0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\tqdm\\auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "a49a9dc1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a smaller subset\n",
    "test_images= X_test[100:110]\n",
    "test_labels=y_test[100:110]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "850af659",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\shap\\explainers\\tf_utils.py:28: The name tf.keras.backend.get_session is deprecated. Please use tf.compat.v1.keras.backend.get_session instead.\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keras is no longer supported, please use tf.keras instead.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Florian\\anaconda3\\envs\\yield_ml_2\\lib\\site-packages\\shap\\explainers\\_deep\\deep_tf.py:631: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
     ]
    }
   ],
   "source": [
    "# fit the explainer\n",
    "explainer = shap.DeepExplainer(model, X_test[0:4999])\n",
    "shap_values = explainer.shap_values(test_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "2338fd0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# reformat output\n",
    "shap_list=[]\n",
    "for row in shap_values[0]:\n",
    "    row_list = []\n",
    "    for item in row:\n",
    "        row_list.append(item[0])\n",
    "    #print(row_list)\n",
    "    shap_list.append(row_list)\n",
    "# make it numpy array    \n",
    "shap_array = np.array(shap_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "adb81cd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a feature name list\n",
    "feature_names = ['SPEI-24','SPEI-23','SPEI-22','SPEI-21','SPEI-20','SPEI-19',\n",
    "                'SPEI-18','SPEI-17','SPEI-16','SPEI-15','SPEI-14','SPEI-13',\n",
    "                'SPEI-12','SPEI-11','SPEI-10','SPEI-9','SPEI-8','SPEI-7',\n",
    "                'SPEI-6','SPEI-5','SPEI-4','SPEI-3','SPEI-2','SPEI-1']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "0960ff28",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot and save the figure\n",
    "shap.summary_plot(shap_array, plot_type = 'bar', feature_names = np.array(feature_names), show=False)\n",
    "plt.savefig(r'C:\\Users\\Florian\\Desktop\\Jupyter_Skripts\\05_ML_crop_vulnerability/figures/barplot.png', format='png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "c4f4e3d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot and save the figure\n",
    "shap.summary_plot(shap_array,feature_names = np.array(feature_names),plot_type='violin', show=False)\n",
    "plt.savefig(r'C:\\Users\\Florian\\Desktop\\Jupyter_Skripts\\05_ML_crop_vulnerability/figures/violinplot.png', format='png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1cccef92",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1a99ca4-e128-4306-b9da-6ad60cc3a82f",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}