{
"cells": [
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"import gradio as gr\n",
"import tensorflow as tf\n",
"import numpy as np\n",
"import pandas as pd"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"\n",
"model_path = \"kia_apartment_keras_model.keras\"\n",
"model = tf.keras.models.load_model(model_path)\n",
"\n",
"df_bfs_data = pd.read_csv('bfs_municipality_and_tax_data.csv', sep=',', encoding='utf-8')\n",
"df_bfs_data['tax_income'] = df_bfs_data['tax_income'].str.replace(\"'\", \"\").astype(float)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"locations = {\n",
" \"Zürich\": 261,\n",
" \"Kloten\": 62,\n",
" \"Uster\": 198,\n",
" \"Illnau-Effretikon\": 296,\n",
" \"Feuerthalen\": 27,\n",
" \"Pfäffikon\": 177,\n",
" \"Ottenbach\": 11,\n",
" \"Dübendorf\": 191,\n",
" \"Richterswil\": 138,\n",
" \"Maur\": 195,\n",
" \"Embrach\": 56,\n",
" \"Bülach\": 53,\n",
" \"Winterthur\": 230,\n",
" \"Oetwil am See\": 157,\n",
" \"Russikon\": 178,\n",
" \"Obfelden\": 10,\n",
" \"Wald (ZH)\": 120,\n",
" \"Niederweningen\": 91,\n",
" \"Dällikon\": 84,\n",
" \"Buchs (ZH)\": 83,\n",
" \"Rüti (ZH)\": 118,\n",
" \"Hittnau\": 173,\n",
" \"Bassersdorf\": 52,\n",
" \"Glattfelden\": 58,\n",
" \"Opfikon\": 66,\n",
" \"Hinwil\": 117,\n",
" \"Regensberg\": 95,\n",
" \"Langnau am Albis\": 136,\n",
" \"Dietikon\": 243,\n",
" \"Erlenbach (ZH)\": 151,\n",
" \"Kappel am Albis\": 6,\n",
" \"Stäfa\": 158,\n",
" \"Zell (ZH)\": 231,\n",
" \"Turbenthal\": 228,\n",
" \"Oberglatt\": 92,\n",
" \"Winkel\": 72,\n",
" \"Volketswil\": 199,\n",
" \"Kilchberg (ZH)\": 135,\n",
" \"Wetzikon (ZH)\": 121,\n",
" \"Zumikon\": 160,\n",
" \"Weisslingen\": 180,\n",
" \"Elsau\": 219,\n",
" \"Hettlingen\": 221,\n",
" \"Rüschlikon\": 139,\n",
" \"Stallikon\": 13,\n",
" \"Dielsdorf\": 86,\n",
" \"Wallisellen\": 69,\n",
" \"Dietlikon\": 54,\n",
" \"Meilen\": 156,\n",
" \"Wangen-Brüttisellen\": 200,\n",
" \"Flaach\": 28,\n",
" \"Regensdorf\": 96,\n",
" \"Niederhasli\": 90,\n",
" \"Bauma\": 297,\n",
" \"Aesch (ZH)\": 241,\n",
" \"Schlieren\": 247,\n",
" \"Dürnten\": 113,\n",
" \"Unterengstringen\": 249,\n",
" \"Gossau (ZH)\": 115,\n",
" \"Oberengstringen\": 245,\n",
" \"Schleinikon\": 98,\n",
" \"Aeugst am Albis\": 1,\n",
" \"Rheinau\": 38,\n",
" \"Höri\": 60,\n",
" \"Rickenbach (ZH)\": 225,\n",
" \"Rafz\": 67,\n",
" \"Adliswil\": 131,\n",
" \"Zollikon\": 161,\n",
" \"Urdorf\": 250,\n",
" \"Hombrechtikon\": 153,\n",
" \"Birmensdorf (ZH)\": 242,\n",
" \"Fehraltorf\": 172,\n",
" \"Weiach\": 102,\n",
" \"Männedorf\": 155,\n",
" \"Küsnacht (ZH)\": 154,\n",
" \"Hausen am Albis\": 4,\n",
" \"Hochfelden\": 59,\n",
" \"Fällanden\": 193,\n",
" \"Greifensee\": 194,\n",
" \"Mönchaltorf\": 196,\n",
" \"Dägerlen\": 214,\n",
" \"Thalheim an der Thur\": 39,\n",
" \"Uetikon am See\": 159,\n",
" \"Seuzach\": 227,\n",
" \"Uitikon\": 248,\n",
" \"Affoltern am Albis\": 2,\n",
" \"Geroldswil\": 244,\n",
" \"Niederglatt\": 89,\n",
" \"Thalwil\": 141,\n",
" \"Rorbas\": 68,\n",
" \"Pfungen\": 224,\n",
" \"Weiningen (ZH)\": 251,\n",
" \"Bubikon\": 112,\n",
" \"Neftenbach\": 223,\n",
" \"Mettmenstetten\": 9,\n",
" \"Otelfingen\": 94,\n",
" \"Flurlingen\": 29,\n",
" \"Stadel\": 100,\n",
" \"Grüningen\": 116,\n",
" \"Henggart\": 31,\n",
" \"Dachsen\": 25,\n",
" \"Bonstetten\": 3,\n",
" \"Bachenbülach\": 51,\n",
" \"Horgen\": 295\n",
"}"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"# Define the core prediction function\n",
"def predict_apartment(rooms, area, town):\n",
" bfs_number = locations[town]\n",
" df = df_bfs_data[df_bfs_data['bfs_number']==bfs_number]\n",
" \n",
" if len(df) != 1: # if there are more than two records with the same bfs_number reutrn -1\n",
" return -1\n",
" \n",
" input_data = np.array([rooms, area, df['pop'].iloc[0], df['pop_dens'].iloc[0], df['frg_pct'].iloc[0], df['emp'].iloc[0], df['tax_income'].iloc[0]])\n",
" input_data = input_data.reshape(1, 7)\n",
" prediction = model.predict(input_data)\n",
" return float(np.round(prediction[0][0], 0))"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running on local URL: http://127.0.0.1:7863\n",
"\n",
"To create a public link, set `share=True` in `launch()`.\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": []
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 91ms/step\n"
]
}
],
"source": [
"# Create the Gradio interface\n",
"iface = gr.Interface(\n",
" fn=predict_apartment,\n",
" inputs=[\"number\", \"number\", gr.Dropdown(choices=locations.keys(), label=\"Town\", type=\"value\")],\n",
" outputs=gr.Number(),\n",
" examples=[[4.5, 120, \"Dietlikon\"], [3.5, 60, \"Winterthur\"]]\n",
")\n",
"\n",
"iface.launch()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "venv_new",
"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": "3.11.3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}