Update Preprocessing

ThisIsSimple · May 11, 2024 · 18335f3 · 18335f3
1 parent fc6129b
commit 18335f3
Show file tree

Hide file tree

Showing 31 changed files with 36,708 additions and 0 deletions.
diff --git a/...db/proprecessing/disability_facilities/.ipynb_checkpoints/extract_latlon-checkpoint.ipynb b/...db/proprecessing/disability_facilities/.ipynb_checkpoints/extract_latlon-checkpoint.ipynb
@@ -0,0 +1,118 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "8349c3c3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import os\n",
+    "import json\n",
+    "from time import sleep\n",
+    "import requests\n",
+    "import googlemaps"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "dd65a263",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from urllib.parse import quote\n",
+    "from urllib.request import Request, urlopen\n",
+    "import ssl\n",
+    "import json"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "c67b8c21",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "API_KEY = \"********\"\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "624a943f",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|████████████████████████████████████████████████████████████████████████████████| 149/149 [00:17<00:00,  8.42it/s]\n",
+      "100%|██████████████████████████████████████████████████████████████████████████████████| 26/26 [00:04<00:00,  6.50it/s]\n",
+      "100%|██████████████████████████████████████████████████████████████████████████████████| 52/52 [00:08<00:00,  6.18it/s]\n",
+      "100%|████████████████████████████████████████████████████████████████████████████████████| 6/6 [00:00<00:00,  6.64it/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "from tqdm import tqdm\n",
+    "\n",
+    "maps = googlemaps.Client(key=API_KEY)\n",
+    "\n",
+    "lst = ['공동생활가정_경기도', '단기거주시설_경기도', '중증장애인거주시설_경기도', '지체장애인시설_경기도']\n",
+    "for fname in lst:\n",
+    "    try:\n",
+    "        df = pd.read_csv(f'{fname}.csv')\n",
+    "    except:\n",
+    "        df = pd.read_csv(f'{fname}.csv', encoding='cp949')\n",
+    "    \n",
+    "    lat, lon = [], []\n",
+    "\n",
+    "    for address in tqdm(df['시설주소']):\n",
+    "        try:\n",
+    "            geo_location = maps.geocode(address)[0].get('geometry')\n",
+    "            lat.append(geo_location['location']['lat'])\n",
+    "            lon.append(geo_location['location']['lng'])    \n",
+    "        except:\n",
+    "            lat.append('')\n",
+    "            lon.append('')\n",
+    "            print(f\"{address} 에러\")\n",
+    "\n",
+    "    df['lat'] = lat\n",
+    "    df['lon'] = lon\n",
+    "\n",
+    "    df.to_csv(f'{fname}_위경도.csv', index=False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b4c1fdd5",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.13"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/src/db/proprecessing/disability_facilities/2023년_장애인_복지시설_일람표(2022.12월말_기준) (1).docx b/src/db/proprecessing/disability_facilities/2023년_장애인_복지시설_일람표(2022.12월말_기준) (1).docx
diff --git a/src/db/proprecessing/disability_facilities/2023년_장애인_복지시설_일람표(2022.12월말_기준) (1).pdf b/src/db/proprecessing/disability_facilities/2023년_장애인_복지시설_일람표(2022.12월말_기준) (1).pdf
diff --git a/src/db/proprecessing/disability_facilities/disability_facility.pdf b/src/db/proprecessing/disability_facilities/disability_facility.pdf
diff --git a/src/db/proprecessing/disability_facilities/extract_latlon.ipynb b/src/db/proprecessing/disability_facilities/extract_latlon.ipynb
@@ -0,0 +1,118 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "8349c3c3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import os\n",
+    "import json\n",
+    "from time import sleep\n",
+    "import requests\n",
+    "import googlemaps"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "dd65a263",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from urllib.parse import quote\n",
+    "from urllib.request import Request, urlopen\n",
+    "import ssl\n",
+    "import json"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "c67b8c21",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "API_KEY = \"********\"\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "624a943f",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|████████████████████████████████████████████████████████████████████████████████| 149/149 [00:17<00:00,  8.42it/s]\n",
+      "100%|██████████████████████████████████████████████████████████████████████████████████| 26/26 [00:04<00:00,  6.50it/s]\n",
+      "100%|██████████████████████████████████████████████████████████████████████████████████| 52/52 [00:08<00:00,  6.18it/s]\n",
+      "100%|████████████████████████████████████████████████████████████████████████████████████| 6/6 [00:00<00:00,  6.64it/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "from tqdm import tqdm\n",
+    "\n",
+    "maps = googlemaps.Client(key=API_KEY)\n",
+    "\n",
+    "lst = ['공동생활가정_경기도', '단기거주시설_경기도', '중증장애인거주시설_경기도', '지체장애인시설_경기도']\n",
+    "for fname in lst:\n",
+    "    try:\n",
+    "        df = pd.read_csv(f'{fname}.csv')\n",
+    "    except:\n",
+    "        df = pd.read_csv(f'{fname}.csv', encoding='cp949')\n",
+    "    \n",
+    "    lat, lon = [], []\n",
+    "\n",
+    "    for address in tqdm(df['시설주소']):\n",
+    "        try:\n",
+    "            geo_location = maps.geocode(address)[0].get('geometry')\n",
+    "            lat.append(geo_location['location']['lat'])\n",
+    "            lon.append(geo_location['location']['lng'])    \n",
+    "        except:\n",
+    "            lat.append('')\n",
+    "            lon.append('')\n",
+    "            print(f\"{address} 에러\")\n",
+    "\n",
+    "    df['lat'] = lat\n",
+    "    df['lon'] = lon\n",
+    "\n",
+    "    df.to_csv(f'{fname}_위경도.csv', index=False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b4c1fdd5",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.13"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}