prueba-geoclustering-energ-a.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6335ad2d",
   "metadata": {
    "papermill": {
     "duration": 0.00728,
     "end_time": "2023-05-03T22:44:43.574830",
     "exception": false,
     "start_time": "2023-05-03T22:44:43.567550",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Importar librerías: \n",
    "* os: para la unión de varios archivos en un solo dataframe\n",
    "* pandas: para la creación y manipulación de dataframes\n",
    "* GeoPandas: para la generación y visualización de los mapas interactivos\n",
    "* libpysAL:para análisis geoespacial mediante clustering, servirá de apoyo en la generación de una matriz de pesos y adyacencias\n",
    "* spopt (parte de PySAL): para métodos de ML no supervisado de regionalización (clustering con restricción de proximidad espacial)\n",
    "* seaborn: para la representación mediante gráficas de los dataframes, en este caso en concreto se usará para la representación de un heatmap (mapa de calor)\n",
    "* sklearn: es una librería muy usada para modelos de machine learning, en este caso la usaremos para normalizar los datos mediante StandardScaler."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ec6643e0",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2023-05-03T22:44:43.591205Z",
     "iopub.status.busy": "2023-05-03T22:44:43.590621Z",
     "iopub.status.idle": "2023-05-03T22:44:49.347725Z",
     "shell.execute_reply": "2023-05-03T22:44:49.346600Z"
    },
    "papermill": {
     "duration": 5.768826,
     "end_time": "2023-05-03T22:44:49.350718",
     "exception": false,
     "start_time": "2023-05-03T22:44:43.581892",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)\n",
    "warnings.simplefilter(action='ignore', category=UserWarning)\n",
    "\n",
    "import os\n",
    "import pandas as pd\n",
    "import geopandas as gpd\n",
    "import libpysal\n",
    "import spopt\n",
    "\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5a384712",
   "metadata": {
    "papermill": {
     "duration": 0.006972,
     "end_time": "2023-05-03T22:44:49.365279",
     "exception": false,
     "start_time": "2023-05-03T22:44:49.358307",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Concatenar ficheros de datos de potencia (W) de energía eléctrica de cada provincia en un solo dataframe, generada a partir de fuentes de energía renovables, cogeneración y residuos. Los tipos de fuentes de energía están etiquetados según: https://www.boe.es/buscar/act.php?id=BOE-A-2014-6123. En el siguiente paso trataremos de limpiar el dataset con la información relevante"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4f250ae6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:44:49.382886Z",
     "iopub.status.busy": "2023-05-03T22:44:49.380962Z",
     "iopub.status.idle": "2023-05-03T22:45:02.643978Z",
     "shell.execute_reply": "2023-05-03T22:45:02.642575Z"
    },
    "papermill": {
     "duration": 13.274625,
     "end_time": "2023-05-03T22:45:02.647137",
     "exception": false,
     "start_time": "2023-05-03T22:44:49.372512",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# https://energia.serviciosmin.gob.es/Pretor/Vista/Informes/InformesInstalaciones.aspx\n",
    "basepath = 'Data/energia'\n",
    "d_energia_raw = pd.DataFrame()\n",
    "for entry in os.listdir(basepath):\n",
    "    file = os.path.join(basepath, entry)\n",
    "    if os.path.isfile(file):\n",
    "        file_pd = pd.read_html(file)[0]\n",
    "        d_energia_raw = pd.concat([d_energia_raw, file_pd], axis=0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8ce0a377",
   "metadata": {
    "papermill": {
     "duration": 0.006829,
     "end_time": "2023-05-03T22:45:02.660810",
     "exception": false,
     "start_time": "2023-05-03T22:45:02.653981",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Limpieza de datos y agregación para obtener totales de potencia (W) por municipio, cada código viene representado por 1 tipo de generación de energía renovable, aqui se deja una breve descripción de lo que significa cada código, para más información, puede visitarse https://www.boe.es/buscar/act.php?id=BOE-A-2014-6123:\n",
    "\n",
    "* a.1-Cogeneración mediante gas natural, petróleo, etc\n",
    "* b.1-Energía solar\n",
    "* b.2-Energía eólica\n",
    "* b.3-Geotérmica, hidrotérmica, aerotérmica, la de las olas, la de las mareas, la de las rocas calientes y secas, la oceanotérmica y la energía de las corrientes marinas\n",
    "* b.4-Hidroeléctricas <10MW\n",
    "* b.5-Hidroeléctricas >10MW\n",
    "* b.6-Biomasa procedente de cultivos energéticos, de actividades agrícolas, ganaderas o de jardinerías\n",
    "* b.7-Biogas o bioliquidos\n",
    "* b.8-biomasa sector agrícola\n",
    "* c.1-Residuos domésticos y similares\n",
    "* c.2-Centrales que utilicen como combustible principal otros residuos no contemplados anteriormente\n",
    "* c.3-Centrales que utilicen como combustible residuos, siempre que éstos no supongan menos del 50 por ciento de la energía primaria utilizada\\\n",
    "* c.4-Centrales que usan combustible productos de explotaciones mineras de calidades no comerciales\n",
    "* d.1-Instalaciones de tratamiento y reducción de purines de explotaciones de porcino de zonas excedentarias\n",
    "* d.2-Instalaciones de tratamiento y reducción de lodos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "a41fde73",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:02.677673Z",
     "iopub.status.busy": "2023-05-03T22:45:02.676545Z",
     "iopub.status.idle": "2023-05-03T22:45:02.999100Z",
     "shell.execute_reply": "2023-05-03T22:45:02.997835Z"
    },
    "papermill": {
     "duration": 0.333651,
     "end_time": "2023-05-03T22:45:03.001775",
     "exception": false,
     "start_time": "2023-05-03T22:45:02.668124",
     "status": "completed"
    },
    "tags": []
   },
   "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>Tipo</th>\n",
       "      <th>a.1</th>\n",
       "      <th>a.2</th>\n",
       "      <th>b.1</th>\n",
       "      <th>b.2</th>\n",
       "      <th>b.3</th>\n",
       "      <th>b.4</th>\n",
       "      <th>b.5</th>\n",
       "      <th>b.6</th>\n",
       "      <th>b.7</th>\n",
       "      <th>b.8</th>\n",
       "      <th>c.1</th>\n",
       "      <th>c.2</th>\n",
       "      <th>c.3</th>\n",
       "      <th>c.4</th>\n",
       "      <th>d.1</th>\n",
       "      <th>d.2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Municipio</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Abadiño</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>95000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Abadía</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1155000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Abadín</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>42600.0</td>\n",
       "      <td>263510000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Abaltzisketa</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Abanilla</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12612000.0</td>\n",
       "      <td>91440000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1064000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Tipo          a.1  a.2         b.1          b.2  b.3  b.4  b.5  b.6   \n",
       "Municipio                                                             \n",
       "Abadiño       0.0  0.0     95000.0          0.0  0.0  0.0  0.0  0.0  \\\n",
       "Abadía        0.0  0.0   1155000.0          0.0  0.0  0.0  0.0  0.0   \n",
       "Abadín        0.0  0.0     42600.0  263510000.0  0.0  0.0  0.0  0.0   \n",
       "Abaltzisketa  0.0  0.0      7500.0          0.0  0.0  0.0  0.0  0.0   \n",
       "Abanilla      0.0  0.0  12612000.0   91440000.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "Tipo                b.7  b.8  c.1  c.2  c.3  c.4  d.1  d.2  \n",
       "Municipio                                                   \n",
       "Abadiño             0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "Abadía              0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "Abadín              0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "Abaltzisketa        0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "Abanilla      1064000.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_energia = d_energia_raw.iloc[1:, [2,6,7]] \n",
    "d_energia.columns = [\"Municipio\", \"Potencia\", \"Tipo\"]\n",
    "d_energia = d_energia[d_energia.Tipo != \"Almacenamiento\"]\n",
    "d_energia[\"Tipo\"] = d_energia[\"Tipo\"].str[0:3]\n",
    "d_energia[\"Potencia\"] = pd.to_numeric(\n",
    "                            d_energia[\"Potencia\"].str.replace('.', '').str.replace(',', '.'), \n",
    "                        errors=\"coerce\")\n",
    "d_energia = d_energia \\\n",
    "                .dropna() \\\n",
    "                .sort_values(by='Municipio') \\\n",
    "                .groupby([\"Municipio\", \"Tipo\"], as_index=False) \\\n",
    "                .agg({\"Potencia\": \"sum\"}) \\\n",
    "                .pivot(index=\"Municipio\", columns='Tipo', values='Potencia') \\\n",
    "                .fillna(0)\n",
    "d_energia.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "16fd75c5",
   "metadata": {
    "papermill": {
     "duration": 0.006802,
     "end_time": "2023-05-03T22:45:03.015852",
     "exception": false,
     "start_time": "2023-05-03T22:45:03.009050",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Para generar los mapas interactivos de península, canarias y baleares se ha planteado representar cada uno de los municipios mediante polígonos, en el código situado inmediatamente abajo se ha importado dicha información de cada municipio, conservando nombre (NAMEUNIT) y una lista de los vértices mediante coordendas de cada polígono (geometry)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5030e491",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:03.032916Z",
     "iopub.status.busy": "2023-05-03T22:45:03.032406Z",
     "iopub.status.idle": "2023-05-03T22:45:07.627535Z",
     "shell.execute_reply": "2023-05-03T22:45:07.625963Z"
    },
    "papermill": {
     "duration": 4.607482,
     "end_time": "2023-05-03T22:45:07.630574",
     "exception": false,
     "start_time": "2023-05-03T22:45:03.023092",
     "status": "completed"
    },
    "tags": []
   },
   "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>geometry</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Municipio</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A Arnoia</th>\n",
       "      <td>POLYGON ((-8.15983 42.24513, -8.15928 42.24542...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Baña</th>\n",
       "      <td>POLYGON ((-8.83259 42.96346, -8.83205 42.96402...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Bola</th>\n",
       "      <td>POLYGON ((-7.93304 42.11927, -7.93296 42.11967...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Capela</th>\n",
       "      <td>POLYGON ((-8.10065 43.46198, -8.10019 43.46213...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Cañiza</th>\n",
       "      <td>POLYGON ((-8.39289 42.16872, -8.39335 42.16931...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    geometry\n",
       "Municipio                                                   \n",
       "A Arnoia   POLYGON ((-8.15983 42.24513, -8.15928 42.24542...\n",
       "A Baña     POLYGON ((-8.83259 42.96346, -8.83205 42.96402...\n",
       "A Bola     POLYGON ((-7.93304 42.11927, -7.93296 42.11967...\n",
       "A Capela   POLYGON ((-8.10065 43.46198, -8.10019 43.46213...\n",
       "A Cañiza   POLYGON ((-8.39289 42.16872, -8.39335 42.16931..."
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# http://centrodedescargas.cnig.es/CentroDescargas/index.jsp\n",
    "gd = pd.concat([\n",
    "    gpd.read_file(\"Data/poly_peninbal/poly_peninbal.shp\").to_crs(\"epsg:4326\"),\n",
    "    gpd.read_file(\"Data/poly_canarias/poly_canarias.shp\").to_crs(\"epsg:4326\")\n",
    "])\n",
    "gd = gd[[\"NAMEUNIT\", \"geometry\"]] \\\n",
    "        .rename(columns={\"NAMEUNIT\": \"Municipio\"}) \\\n",
    "        .sort_values(by='Municipio') \\\n",
    "        .drop_duplicates(subset=\"Municipio\", keep='first') \\\n",
    "        .set_index(\"Municipio\")\n",
    "gd.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7cbc2523",
   "metadata": {
    "papermill": {
     "duration": 0.007136,
     "end_time": "2023-05-03T22:45:07.645398",
     "exception": false,
     "start_time": "2023-05-03T22:45:07.638262",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Se genera un dataset con la unión de datos de potencia generada por tipo y municipio junto con los datos geográficos de dichos municipios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "783523ee",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:07.662495Z",
     "iopub.status.busy": "2023-05-03T22:45:07.662042Z",
     "iopub.status.idle": "2023-05-03T22:45:07.694431Z",
     "shell.execute_reply": "2023-05-03T22:45:07.693286Z"
    },
    "papermill": {
     "duration": 0.044405,
     "end_time": "2023-05-03T22:45:07.697322",
     "exception": false,
     "start_time": "2023-05-03T22:45:07.652917",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "d = gd.join(d_energia, how=\"left\").fillna(0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "71b6196e",
   "metadata": {
    "papermill": {
     "duration": 0.00715,
     "end_time": "2023-05-03T22:45:07.712410",
     "exception": false,
     "start_time": "2023-05-03T22:45:07.705260",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "A modo de ejemplo vamos a visualizar el mapa coroplético (mapa coloreado por regiones) de la distribución de la potencia generada por fuentes de energía de tipo \"b.1\", correspondiente a: \n",
    "* Grupo b.1 Instalaciones que utilicen como energía primaria la energía solar.\n",
    "\n",
    "Para una representación eficiente del mapa interactivo se va a usar el agoritmo de Douglas-Peucker, mediante este algoritmo se reduce el número de vértices de cada polígono, suavizando de esta manera el polígono y por contra reduciendo el nivel de detalle del mapa en favor de una mayor rapidez de ejecución"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "72875ed7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:07.731497Z",
     "iopub.status.busy": "2023-05-03T22:45:07.730467Z",
     "iopub.status.idle": "2023-05-03T22:45:20.867414Z",
     "shell.execute_reply": "2023-05-03T22:45:20.863655Z"
    },
    "papermill": {
     "duration": 13.317661,
     "end_time": "2023-05-03T22:45:21.038355",
     "exception": false,
     "start_time": "2023-05-03T22:45:07.720694",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Douglas-Peucker Algorithm to simplify geometries\n",
    "d_plot = gpd.GeoDataFrame(d)\n",
    "d_plot.geometry = d_plot.geometry.simplify(0.005)\n",
    "d_plot.explore(column=\"b.1\", \n",
    "               cmap=\"BuPu\",\n",
    "               scheme='HeadTailBreaks',\n",
    "               tooltip=False,\n",
    "               popup=True,\n",
    "               style_kwds={\"weight\": 0.4}).save('Data/Mapas/mapa_b1.html')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1aca60e2",
   "metadata": {
    "papermill": {
     "duration": 0.14868,
     "end_time": "2023-05-03T22:45:21.333816",
     "exception": false,
     "start_time": "2023-05-03T22:45:21.185136",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Como segundo ejemplo, visualicemos el mapa coroplético de la distribución de la potencia generada por fuentes de energía de tipo \"b.2\", correspondiente a: \n",
    "* Grupo b.2 Instalaciones que únicamente utilicen como energía primaria la energía eólica\n",
    "\n",
    "De nuevo usando el algoritmo de Douglas-Peucker"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "61ed9e92",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:21.641652Z",
     "iopub.status.busy": "2023-05-03T22:45:21.641202Z",
     "iopub.status.idle": "2023-05-03T22:45:26.685276Z",
     "shell.execute_reply": "2023-05-03T22:45:26.683186Z"
    },
    "papermill": {
     "duration": 5.364471,
     "end_time": "2023-05-03T22:45:26.850911",
     "exception": false,
     "start_time": "2023-05-03T22:45:21.486440",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Douglas-Peucker Algorithm to simplify geometries\n",
    "d_plot = gpd.GeoDataFrame(d)\n",
    "d_plot.geometry = d_plot.geometry.simplify(0.005)\n",
    "d_plot.explore(column=\"b.2\", \n",
    "               cmap=\"BuPu\",\n",
    "               scheme='HeadTailBreaks',\n",
    "               tooltip=False,\n",
    "               popup=True,\n",
    "               style_kwds={\"weight\": 0.4}).save(\"Data/Mapas/mapa_b2.html\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3cb33cbf",
   "metadata": {
    "papermill": {
     "duration": 0.293919,
     "end_time": "2023-05-03T22:45:27.441811",
     "exception": false,
     "start_time": "2023-05-03T22:45:27.147892",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Para poder realizar el clustering (agrupación en conjuntos) de los municipios con características similares (en lo que a generación de energías renovables respecta) necesitaremos una restricción espacial, es decir, agruparemos por conjuntos con características similares, siempre y cuando dichos municipios compartan fronteras, en otras palabras, siempre y cuando sean contiguos entre sí.\n",
    "\n",
    "Para dicha restricción espacial, usaremos la matriz 'w', la cual indica la adyacencia entre los distintos municipios que comparten frontera (Rook neighbours)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "c7232f95",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:28.026915Z",
     "iopub.status.busy": "2023-05-03T22:45:28.026410Z",
     "iopub.status.idle": "2023-05-03T22:45:30.586210Z",
     "shell.execute_reply": "2023-05-03T22:45:30.584846Z"
    },
    "papermill": {
     "duration": 2.855912,
     "end_time": "2023-05-03T22:45:30.589932",
     "exception": false,
     "start_time": "2023-05-03T22:45:27.734020",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Rook neighbors are those states that share an edge on their respective borders\n",
    "# Queen neighbors are based on a more inclusive condition that requires only a shared vertex between two states\n",
    "w = libpysal.weights.Rook.from_dataframe(d)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7b147f93",
   "metadata": {
    "papermill": {
     "duration": 0.295449,
     "end_time": "2023-05-03T22:45:31.221771",
     "exception": false,
     "start_time": "2023-05-03T22:45:30.926322",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "A continuación, procedemos a agrupar por conjuntos (clusters) los diferentes municipios. Para ello usaremos el algoritmo 'max-P regions' que nos generará la división de conjuntos teniendo en cuenta la matriz de adyacencia 'w' generada en el paso anterior, así como un threshold (umbral inferior) minimo de municipios por cluster.\n",
    "\n",
    "En resumen, generamos clusters con municipios con generación de energía eléctrica de fuentes renovables similares entre sí, que, a su vez, compartan frontera, con un mínimo de 80 municipios por cluster.\n",
    "\n",
    "El clustering es un algoritmo de aprendizaje automático (machine learning) muy usado para el aprendizaje no supervisado cuando tenemos un conjunto de datos sin etiquetar y del cual queramos sacar conclusiones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "cec8a79f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T22:45:31.811869Z",
     "iopub.status.busy": "2023-05-03T22:45:31.810570Z",
     "iopub.status.idle": "2023-05-03T23:46:16.335776Z",
     "shell.execute_reply": "2023-05-03T23:46:16.334374Z"
    },
    "papermill": {
     "duration": 3645.118285,
     "end_time": "2023-05-03T23:46:16.634784",
     "exception": false,
     "start_time": "2023-05-03T22:45:31.516499",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max_p:  91\n",
      "number of good partitions: 1\n",
      "0\n",
      "totalWithinRegionDistance after SA: \n",
      "5808014553485.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5809967335088.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5788138463466.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5801963808356.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5785676345476.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5807692399652.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5771797934413.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5807109012982.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5775221838254.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5833148416385.0\n",
      "best objective value:\n",
      "5771797934413.0\n"
     ]
    }
   ],
   "source": [
    "# Max-P Heuristic with a minimum number of towns per cluster\n",
    "d[\"Count\"] = 1\n",
    "attrs_name = list(d.columns)[1:-1]\n",
    "# https://pysal.org/spopt/generated/spopt.region.MaxPHeuristic.html#spopt.region.MaxPHeuristic\n",
    "model = spopt.region.MaxPHeuristic(d, w, attrs_name, \n",
    "                                   verbose=True,\n",
    "                                   # At least 80 towns per cluster\n",
    "                                   threshold=80, \n",
    "                                   threshold_name=\"Count\", \n",
    "                                   # Computational cost\n",
    "                                   max_iterations_sa=10,\n",
    "                                   max_iterations_construction=999,\n",
    "                                   top_n=10,\n",
    "                                   # Attach infeasible components using a single linkage between the area in the infeasible component with the smallest nearest neighbor distance to an area in a feasible component\n",
    "                                   policy=\"single\")\n",
    "model.solve()\n",
    "d['Cluster'] = model.labels_\n",
    "# Store df in an csv file\n",
    "d.to_csv(\"Data/Output/output.csv\", index=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9c0a7304",
   "metadata": {
    "papermill": {
     "duration": 0.299193,
     "end_time": "2023-05-03T23:46:17.235020",
     "exception": false,
     "start_time": "2023-05-03T23:46:16.935827",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Repetimos la representación mediante el mapa coroplético, esta vez mediante los clusters generados y no mediante las fuentes renovables para la generación de energía eléctrica como hicimos con anterioridad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "585a7285",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T23:46:17.816083Z",
     "iopub.status.busy": "2023-05-03T23:46:17.815510Z",
     "iopub.status.idle": "2023-05-03T23:46:23.104731Z",
     "shell.execute_reply": "2023-05-03T23:46:23.103245Z"
    },
    "papermill": {
     "duration": 5.761269,
     "end_time": "2023-05-03T23:46:23.286117",
     "exception": false,
     "start_time": "2023-05-03T23:46:17.524848",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Douglas-Peucker Algorithm to simplify geometries\n",
    "d_plot = gpd.GeoDataFrame(d)\n",
    "d_plot.geometry = d_plot.geometry.simplify(0.005)\n",
    "d_plot.explore(column=\"Cluster\", \n",
    "           cmap=\"Set1\",\n",
    "           tooltip=False,\n",
    "           popup=True,\n",
    "           legend=True,\n",
    "           style_kwds={\"weight\": 0.4}).save(\"Data/Mapas/mapa_clusters_1.html\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "486e06e2",
   "metadata": {
    "papermill": {
     "duration": 0.404998,
     "end_time": "2023-05-03T23:46:24.104710",
     "exception": false,
     "start_time": "2023-05-03T23:46:23.699712",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Mapa de calor que muestra los valores medios de cada variable (potencia de la fuente de producción de energía) en los municipios de cada cluster generado. \n",
    "En los mapas de calor (heatmaps), se muestran los valores medios de la producción eléctrica con cada fuente de energía en cada agrupación de municipios. Dado que los datos se centran y escalan previamente, transformándolos hasta que cada variable (la producción con cada fuente de energía) tenga media 0 y desviación típica 1, los valores superiores a 0 indican que la producción eléctrica es superior a la media, y viceversa. Con la visualización en forma de mapa de calor, trasladando la escala de números a una escala de colores, se pueden apreciar de un vistazo las características más sobresalientes de cada cluster de municipios, identificando las fuentes de energía que destacan por su producción superior (tonos verdes) o inferior (tonos amarillos) a la media. Así, por ejemplo, el conjunto de municipios que constituyen el cluster identificado en los mapas con el número 20 se caracteriza principalmente por tener una producción eléctrica procedente de la energía eólica muy superior a la media"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "4998d2d2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T23:46:24.900765Z",
     "iopub.status.busy": "2023-05-03T23:46:24.900335Z",
     "iopub.status.idle": "2023-05-03T23:46:25.897596Z",
     "shell.execute_reply": "2023-05-03T23:46:25.896372Z"
    },
    "papermill": {
     "duration": 1.398329,
     "end_time": "2023-05-03T23:46:25.900311",
     "exception": false,
     "start_time": "2023-05-03T23:46:24.501982",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: ylabel='Cluster'>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x1500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "aux = d.drop(columns=[\"geometry\",\"Count\",\"Cluster\"])\n",
    "d_scaled = pd.DataFrame(\n",
    "    StandardScaler().fit_transform(aux),\n",
    "    columns = aux.columns,\n",
    "    index = aux.index\n",
    ")\n",
    "d_scaled[\"Cluster\"] = d.Cluster\n",
    "sns.set(rc={'figure.figsize':(8,15)})\n",
    "sns.heatmap(d_scaled.groupby(\"Cluster\").mean(), cmap=\"RdYlGn\", center=0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e098144c",
   "metadata": {
    "papermill": {
     "duration": 0.392626,
     "end_time": "2023-05-03T23:46:26.693697",
     "exception": false,
     "start_time": "2023-05-03T23:46:26.301071",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Anteriormente, para establecer el tamaño de cada agrupación (cluster) poníamos como threshold (umbral) el número de municipios. \n",
    "En este caso, vamos a realizar las agrupaciones estableción un mímo en función del número total de habitantes, y no del mínimo total de municipios.\n",
    "Para ello necesitaremos el número total de habitantes por municipio, el cual importaremos y uniremos al dataframe original\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "77fc9360",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T23:46:27.495671Z",
     "iopub.status.busy": "2023-05-03T23:46:27.494648Z",
     "iopub.status.idle": "2023-05-03T23:46:33.641621Z",
     "shell.execute_reply": "2023-05-03T23:46:33.640095Z"
    },
    "papermill": {
     "duration": 6.556868,
     "end_time": "2023-05-03T23:46:33.644777",
     "exception": false,
     "start_time": "2023-05-03T23:46:27.087909",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# https://www.ine.es/jaxiT3/files/t/es/csv_bdsc/30832.csv\n",
    "# d_poblacion = pd.read_csv(\"https://www.ine.es/jaxiT3/files/t/es/csv_bdsc/30832.csv\", sep=\";\", decimal=\",\")\n",
    "# d_poblacion.to_parquet(\"poblacion.parquet\")\n",
    "d_poblacion = pd.read_parquet(\"Data/poblacion/poblacion.parquet\")\n",
    "d_poblacion[\"Municipios\"] = d_poblacion[\"Municipios\"].str[6:]\n",
    "d_poblacion[\"Total\"] = pd.to_numeric(\n",
    "    d_poblacion[\"Total\"].str.replace('.', '').str.replace(',', '.'), \n",
    "    errors=\"coerce\")\n",
    "d_poblacion = d_poblacion[(d_poblacion.Periodo == 2020) &\n",
    "                          (d_poblacion[\"Indicadores demográficos\"] == \"Población\") &\n",
    "                          (d_poblacion.Distritos.isna()) &\n",
    "                          (d_poblacion.Secciones.isna())] \\\n",
    "            .drop([\"Distritos\", \"Secciones\", \"Indicadores demográficos\", \"Periodo\"], axis=1) \\\n",
    "            .dropna() \\\n",
    "            .rename(columns={\"Total\": \"Poblacion\",\n",
    "                             \"Municipios\": \"Municipio\"}) \\\n",
    "            .set_index(\"Municipio\")\n",
    "d_poblacion.head()\n",
    "d_energia = d_energia.join(d_poblacion, how=\"inner\")\n",
    "d = gd.join(d_energia, how=\"left\").fillna(0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9fe5c13c",
   "metadata": {
    "papermill": {
     "duration": 0.41993,
     "end_time": "2023-05-03T23:46:34.503239",
     "exception": false,
     "start_time": "2023-05-03T23:46:34.083309",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Creación de la matriz de adyacencias 'w' que será usada para el algoritmo de clustering con el nuevo threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "14b15308",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T23:46:35.320154Z",
     "iopub.status.busy": "2023-05-03T23:46:35.319304Z",
     "iopub.status.idle": "2023-05-03T23:47:09.920034Z",
     "shell.execute_reply": "2023-05-03T23:47:09.917872Z"
    },
    "papermill": {
     "duration": 35.013545,
     "end_time": "2023-05-03T23:47:09.923723",
     "exception": false,
     "start_time": "2023-05-03T23:46:34.910178",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Rook neighbors are those states that share an edge on their respective borders\n",
    "# Queen neighbors are based on a more inclusive condition that requires only a shared vertex between two states\n",
    "w = libpysal.weights.Rook.from_dataframe(d.drop(columns=[\"Poblacion\"]))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e0fa3ebb",
   "metadata": {
    "papermill": {
     "duration": 0.416397,
     "end_time": "2023-05-03T23:47:10.769847",
     "exception": false,
     "start_time": "2023-05-03T23:47:10.353450",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Volvemos a ejecutar el mismo clustering mediante el algoritmo de max-P regions, pero modificando el umbral (threshold & threshold_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "29f1a49a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-03T23:47:11.573809Z",
     "iopub.status.busy": "2023-05-03T23:47:11.573109Z",
     "iopub.status.idle": "2023-05-04T07:39:50.205770Z",
     "shell.execute_reply": "2023-05-04T07:39:50.204461Z"
    },
    "papermill": {
     "duration": 28359.447022,
     "end_time": "2023-05-04T07:39:50.619524",
     "exception": false,
     "start_time": "2023-05-03T23:47:11.172502",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max_p:  123\n",
      "number of good partitions: 1\n",
      "0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ignac\\OneDrive\\Documents\\Hackaton 2023\\PythonVoilaDemo\\.venv\\lib\\site-packages\\spopt\\region\\maxp.py:672: RuntimeWarning: overflow encountered in double_scalars\n",
      "  prob = np.exp(diff / t)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "totalWithinRegionDistance after SA: \n",
      "5560046681953.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5119416293089.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5219147546940.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5193505158963.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5214892322352.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5218599466982.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5186081063303.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5617176358188.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5225923946995.0\n",
      "totalWithinRegionDistance after SA: \n",
      "5057759361286.0\n",
      "best objective value:\n",
      "5057759361286.0\n"
     ]
    }
   ],
   "source": [
    "# Max-P Heuristic with a minimum number population per cluster\n",
    "attrs_name = list(d.columns)[1:-1]\n",
    "# https://pysal.org/spopt/generated/spopt.region.MaxPHeuristic.html#spopt.region.MaxPHeuristic\n",
    "model = spopt.region.MaxPHeuristic(d, w, attrs_name, \n",
    "                                   verbose=True,\n",
    "                                   # At least 200,000 people per cluster\n",
    "                                   threshold=200000, \n",
    "                                   threshold_name=\"Poblacion\", \n",
    "                                   # Computational cost\n",
    "                                   max_iterations_sa=10,\n",
    "                                   max_iterations_construction=999,\n",
    "                                   top_n=10,\n",
    "                                   # Attach infeasible components using a single linkage between the area in the infeasible component with the smallest nearest neighbor distance to an area in a feasible component\n",
    "                                   policy=\"single\")\n",
    "model.solve()\n",
    "d['Cluster'] = model.labels_\n",
    "# Store df in an csv file\n",
    "d.to_csv(\"Data/Output/output2.csv\", index=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "de6ed356",
   "metadata": {
    "papermill": {
     "duration": 0.407518,
     "end_time": "2023-05-04T07:39:51.421116",
     "exception": false,
     "start_time": "2023-05-04T07:39:51.013598",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Representamos de nuevo el mapa coroplético con los nuevos clusters creados en función del threshold de población"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "4c4c82f3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-04T07:39:52.230878Z",
     "iopub.status.busy": "2023-05-04T07:39:52.230242Z",
     "iopub.status.idle": "2023-05-04T07:40:03.473194Z",
     "shell.execute_reply": "2023-05-04T07:40:03.472020Z"
    },
    "papermill": {
     "duration": 11.81333,
     "end_time": "2023-05-04T07:40:03.640268",
     "exception": false,
     "start_time": "2023-05-04T07:39:51.826938",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Douglas-Peucker Algorithm to simplify geometries\n",
    "d_plot = gpd.GeoDataFrame(d)\n",
    "d_plot.geometry = d_plot.geometry.simplify(0.005)\n",
    "d_plot.explore(column=\"Cluster\", \n",
    "           cmap=\"Set1\",\n",
    "           tooltip=False,\n",
    "           popup=True,\n",
    "           legend=True,\n",
    "           style_kwds={\"weight\": 0.4}).save(\"Data/Mapas/mapa_clusters_2.html\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "cfcb62a5",
   "metadata": {
    "papermill": {
     "duration": 0.572669,
     "end_time": "2023-05-04T07:40:04.783380",
     "exception": false,
     "start_time": "2023-05-04T07:40:04.210711",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Volvemos a generar el heatmap (mapa de calor), esta vez con los nuevos clusters generados en función de la población total establecida como umbral.\n",
    "\n",
    "Recordemos que mediante el heatmap podemos ver de un vistazo los clusters que superan la media (aquellos superiores a 0), en la producción de energía eléctrica según el tipo de fuente renovable utilizada."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f7e237f2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-04T07:40:06.020786Z",
     "iopub.status.busy": "2023-05-04T07:40:06.020336Z",
     "iopub.status.idle": "2023-05-04T07:40:06.954934Z",
     "shell.execute_reply": "2023-05-04T07:40:06.953779Z"
    },
    "papermill": {
     "duration": 1.555805,
     "end_time": "2023-05-04T07:40:06.957377",
     "exception": false,
     "start_time": "2023-05-04T07:40:05.401572",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: ylabel='Cluster'>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x1500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "aux = d.drop(columns=[\"geometry\",\"Poblacion\",\"Cluster\"])\n",
    "d_scaled = pd.DataFrame(\n",
    "    StandardScaler().fit_transform(aux),\n",
    "    columns = aux.columns,\n",
    "    index = aux.index\n",
    ")\n",
    "d_scaled[\"Cluster\"] = d.Cluster\n",
    "sns.set(rc={'figure.figsize':(8,15)})\n",
    "sns.heatmap(d_scaled.groupby(\"Cluster\").mean(), cmap=\"RdYlGn\", center=0)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.10.11"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 32138.851331,
   "end_time": "2023-05-04T07:40:10.476413",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2023-05-03T22:44:31.625082",
   "version": "2.4.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}