{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Create canonical labels attached to tessellation\n",
    "\n",
    "This notebook combines results of the first and second level clustering, generating a canonical singature type ID reflecting both levels."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "source": [
    "import pandas as pd\n",
    "import geopandas as gpd\n",
    "import dask.dataframe\n",
    "import dask_geopandas"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "source": [
    "level1 = pd.read_parquet(\"../../urbangrammar_samba/spatial_signatures/clustering_data/KMeans10GB.pq\")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "source": [
    "level1"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "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>kmeans10gb</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hindex</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>c000e094707t0000</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0000</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0001</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0002</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094764t0000</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0111</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0112</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0113</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0114</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0115</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14539578 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  kmeans10gb\n",
       "hindex                      \n",
       "c000e094707t0000           4\n",
       "c000e094763t0000           0\n",
       "c000e094763t0001           0\n",
       "c000e094763t0002           0\n",
       "c000e094764t0000           0\n",
       "...                      ...\n",
       "c102e644989t0111           0\n",
       "c102e644989t0112           0\n",
       "c102e644989t0113           0\n",
       "c102e644989t0114           0\n",
       "c102e644989t0115           0\n",
       "\n",
       "[14539578 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "execution_count": 3
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "source": [
    "level2_9 = pd.read_parquet(\"../../urbangrammar_samba/spatial_signatures/clustering_data/clustergram_cl9_labels.pq\", columns=['9'])"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "level2_9"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "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>9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113639</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113640</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113641</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113642</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113643</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>113644 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        9\n",
       "0       0\n",
       "1       0\n",
       "2       0\n",
       "3       0\n",
       "4       0\n",
       "...    ..\n",
       "113639  0\n",
       "113640  0\n",
       "113641  0\n",
       "113642  0\n",
       "113643  0\n",
       "\n",
       "[113644 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "execution_count": 5
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "level2_2 = pd.read_parquet(\"../../urbangrammar_samba/spatial_signatures/clustering_data/subclustering_cluster2_k3.pq\")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "source": [
    "level2_2"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "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>subclustering_cluster2_k3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hindex</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>c000e097919t0003</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e097919t0005</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e097919t0008</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e097919t0009</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e097919t0015</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e639766t0007</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e639766t0010</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e639766t0011</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e639766t0012</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e639766t0013</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1115564 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  subclustering_cluster2_k3\n",
       "hindex                                     \n",
       "c000e097919t0003                          1\n",
       "c000e097919t0005                          1\n",
       "c000e097919t0008                          1\n",
       "c000e097919t0009                          1\n",
       "c000e097919t0015                          1\n",
       "...                                     ...\n",
       "c102e639766t0007                          0\n",
       "c102e639766t0010                          0\n",
       "c102e639766t0011                          0\n",
       "c102e639766t0012                          0\n",
       "c102e639766t0013                          0\n",
       "\n",
       "[1115564 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "execution_count": 7
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "source": [
    "labels = level1.copy()\n",
    "labels[\"signature_type\"] = labels[\"kmeans10gb\"]\n",
    "labels.loc[labels.kmeans10gb == 9, \"signature_type\"] = (labels.loc[labels.kmeans10gb == 9, \"signature_type\"] * 10) + level2_9[\"9\"].values\n",
    "labels.loc[labels.kmeans10gb == 2, \"signature_type\"] = (labels.loc[labels.kmeans10gb == 2, \"signature_type\"] * 10) +level2_2[\"subclustering_cluster2_k3\"].values"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "source": [
    "labels = labels.drop(columns=[\"kmeans10gb\"])\n",
    "labels"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "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>signature_type</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hindex</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>c000e094707t0000</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0000</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0001</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094763t0002</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c000e094764t0000</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0111</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0112</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0113</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0114</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c102e644989t0115</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14539578 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  signature_type\n",
       "hindex                          \n",
       "c000e094707t0000               4\n",
       "c000e094763t0000               0\n",
       "c000e094763t0001               0\n",
       "c000e094763t0002               0\n",
       "c000e094764t0000               0\n",
       "...                          ...\n",
       "c102e644989t0111               0\n",
       "c102e644989t0112               0\n",
       "c102e644989t0113               0\n",
       "c102e644989t0114               0\n",
       "c102e644989t0115               0\n",
       "\n",
       "[14539578 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "execution_count": 9
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "source": [
    "labels.to_parquet(\"../../urbangrammar_samba/spatial_signatures/signatures/signatures_combined_levels_labels.pq\")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "source": [
    "cells = dask_geopandas.read_parquet(\"../../urbangrammar_samba/spatial_signatures/tessellation/\")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "source": [
    "cells = cells.merge(labels, how=\"left\", left_on=\"hindex\", right_index=True)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "source": [
    "cells = cells.drop(columns=\"buildings\")\n",
    "cells = cells.rename(columns={\"tessellation\": \"geometry\"})\n",
    "cells"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div><strong>Dask DataFrame Structure:</strong></div>\n",
       "<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>hindex</th>\n",
       "      <th>geometry</th>\n",
       "      <th>signature_type</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>npartitions=103</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>object</td>\n",
       "      <td>object</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "<div>Dask Name: rename, 413 tasks</div>"
      ],
      "text/plain": [
       "Dask DataFrame Structure:\n",
       "                 hindex geometry signature_type\n",
       "npartitions=103                                \n",
       "                 object   object        float64\n",
       "                    ...      ...            ...\n",
       "...                 ...      ...            ...\n",
       "                    ...      ...            ...\n",
       "                    ...      ...            ...\n",
       "Dask Name: rename, 413 tasks"
      ]
     },
     "metadata": {},
     "execution_count": 12
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "source": [
    "cells = dask_geopandas.from_dask_dataframe(cells)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "source": [
    "cells"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div><strong>Dask-GeoPandas GeoDataFrame Structure:</strong></div>\n",
       "<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>hindex</th>\n",
       "      <th>geometry</th>\n",
       "      <th>signature_type</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>npartitions=103</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>object</td>\n",
       "      <td>geometry</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "<div>Dask Name: GeoDataFrame, 516 tasks</div>"
      ],
      "text/plain": [
       "<dask_geopandas.GeoDataFrame | 516 tasks | 103 npartitions>"
      ]
     },
     "metadata": {},
     "execution_count": 14
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "source": [
    "%%time\n",
    "cells.to_parquet(\"../../urbangrammar_samba/spatial_signatures/signatures/signatures_combined_tessellation/\")"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stderr",
     "text": [
      "/opt/conda/lib/python3.8/site-packages/dask/utils.py:35: UserWarning: this is an initial implementation of Parquet/Feather file support and associated metadata.  This is tracking version 0.1.0 of the metadata specification at https://github.com/geopandas/geo-arrow-spec\n",
      "\n",
      "This metadata specification does not yet make stability promises.  We do not yet recommend using this in a production setting unless you are able to rewrite your Parquet/Feather files.\n",
      "\n",
      "To further ignore this warning, you can do: \n",
      "import warnings; warnings.filterwarnings('ignore', message='.*initial implementation of Parquet.*')\n",
      "  return func(*args, **kwargs)\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "CPU times: user 17min 44s, sys: 2min 26s, total: 20min 11s\n",
      "Wall time: 20min 11s\n"
     ]
    }
   ],
   "metadata": {
    "tags": []
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "We can check the result and its spatial partitioning."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "source": [
    "cells = dask_geopandas.read_parquet(\"../../urbangrammar_samba/spatial_signatures/signatures/signatures_combined_tessellation/\")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "source": [
    "cells"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div><strong>Dask-GeoPandas GeoDataFrame Structure:</strong></div>\n",
       "<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>hindex</th>\n",
       "      <th>geometry</th>\n",
       "      <th>signature_type</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>npartitions=103</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>object</td>\n",
       "      <td>geometry</td>\n",
       "      <td>int32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "<div>Dask Name: read-parquet, 103 tasks</div>"
      ],
      "text/plain": [
       "<dask_geopandas.GeoDataFrame | 103 tasks | 103 npartitions>"
      ]
     },
     "metadata": {},
     "execution_count": 17
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "source": [
    "%%time\n",
    "cells.calculate_spatial_partitions()"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "CPU times: user 27min 22s, sys: 34 s, total: 27min 57s\n",
      "Wall time: 3min 22s\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "source": [
    "cells.spatial_partitions"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "0      POLYGON Z ((339010.000 426700.000 0.000, 33596...\n",
       "1      POLYGON Z ((459529.390 339487.330 0.000, 45951...\n",
       "2      POLYGON Z ((380696.000 228900.000 0.000, 37766...\n",
       "3      POLYGON Z ((363161.000 541631.000 0.000, 33680...\n",
       "4      POLYGON Z ((431181.406 393369.256 0.000, 40940...\n",
       "                             ...                        \n",
       "98     POLYGON Z ((448173.875 209362.352 0.000, 44814...\n",
       "99     POLYGON Z ((513455.520 192183.030 0.000, 50044...\n",
       "100    POLYGON Z ((172420.000 603651.000 0.000, 16125...\n",
       "101    POLYGON Z ((419376.825 272261.845 0.000, 41514...\n",
       "102    POLYGON Z ((367870.000 68210.000 0.000, 367820...\n",
       "Length: 103, dtype: geometry"
      ]
     },
     "metadata": {},
     "execution_count": 19
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "source": [
    "cells.spatial_partitions.plot(figsize=(12, 12), cmap=\"tab20\", alpha=.2)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "metadata": {},
     "execution_count": 20
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}