Spaces:
Sleeping
Sleeping
Commit
·
6f34772
1
Parent(s):
f54c316
inital commit
Browse files- .gitignore +2 -0
- app.py +174 -0
- bfs_municipality_and_tax_data.csv +0 -0
- random_forest_regression_extended.pkl +3 -0
- requirements.txt +5 -0
.gitignore
ADDED
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
1 |
+
disposition.md
|
2 |
+
week1/data_with_large_residuals.csv
|
app.py
ADDED
@@ -0,0 +1,174 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# %%
|
2 |
+
import gradio as gr
|
3 |
+
from sklearn.ensemble import RandomForestRegressor
|
4 |
+
import numpy as np
|
5 |
+
import pandas as pd
|
6 |
+
import pickle
|
7 |
+
# Define model filename
|
8 |
+
model_filename = "random_forest_regression_extended.pkl"
|
9 |
+
|
10 |
+
try:
|
11 |
+
# Try to load the model
|
12 |
+
with open(model_filename, 'rb') as f:
|
13 |
+
model_data = pickle.load(f)
|
14 |
+
if isinstance(model_data, dict) and 'model' in model_data and 'feature_names' in model_data:
|
15 |
+
random_forest_model = model_data['model']
|
16 |
+
feature_names = model_data['feature_names']
|
17 |
+
# Check scikit-learn version and handle feature information
|
18 |
+
if hasattr(random_forest_model, 'n_features_in_'):
|
19 |
+
print('Number of features: ', random_forest_model.n_features_in_)
|
20 |
+
else:
|
21 |
+
print('Number of features: ', len(feature_names))
|
22 |
+
print('Features are: ', feature_names)
|
23 |
+
else:
|
24 |
+
print("Error: Model file does not contain expected dictionary structure")
|
25 |
+
print("Expected keys: 'model' and 'feature_names'")
|
26 |
+
print(f"Found keys: {model_data.keys() if isinstance(model_data, dict) else 'not a dictionary'}")
|
27 |
+
exit(1)
|
28 |
+
except FileNotFoundError:
|
29 |
+
print(f"Error: Could not find model file '{model_filename}'")
|
30 |
+
print("Please run save_model.py first to create the model file.")
|
31 |
+
exit(1)
|
32 |
+
except Exception as e:
|
33 |
+
print(f"Error loading model: {str(e)}")
|
34 |
+
print(f"scikit-learn version: {sklearn.__version__}")
|
35 |
+
exit(1)
|
36 |
+
|
37 |
+
# Load and prepare BFS data
|
38 |
+
df_bfs_data = pd.read_csv('bfs_municipality_and_tax_data.csv', sep=',', encoding='utf-8')
|
39 |
+
df_bfs_data['tax_income'] = df_bfs_data['tax_income'].str.replace("'", "").astype(float)
|
40 |
+
df_bfs_data['proximity_to_public_transportation'] = 500 # Default value in meters
|
41 |
+
|
42 |
+
|
43 |
+
# %%
|
44 |
+
locations = {
|
45 |
+
"Zürich": 261,
|
46 |
+
"Kloten": 62,
|
47 |
+
"Uster": 198,
|
48 |
+
"Illnau-Effretikon": 296,
|
49 |
+
"Feuerthalen": 27,
|
50 |
+
"Pfäffikon": 177,
|
51 |
+
"Ottenbach": 11,
|
52 |
+
"Dübendorf": 191,
|
53 |
+
"Richterswil": 138,
|
54 |
+
"Maur": 195,
|
55 |
+
"Embrach": 56,
|
56 |
+
"Bülach": 53,
|
57 |
+
"Winterthur": 230,
|
58 |
+
"Oetwil am See": 157,
|
59 |
+
"Russikon": 178,
|
60 |
+
"Obfelden": 10,
|
61 |
+
"Wald (ZH)": 120,
|
62 |
+
"Niederweningen": 91,
|
63 |
+
"Dällikon": 84,
|
64 |
+
"Buchs (ZH)": 83,
|
65 |
+
"Rüti (ZH)": 118,
|
66 |
+
"Hittnau": 173,
|
67 |
+
"Bassersdorf": 52,
|
68 |
+
"Glattfelden": 58,
|
69 |
+
"Opfikon": 66,
|
70 |
+
"Hinwil": 117,
|
71 |
+
"Regensberg": 95,
|
72 |
+
"Langnau am Albis": 136,
|
73 |
+
"Dietikon": 243,
|
74 |
+
"Erlenbach (ZH)": 151,
|
75 |
+
"Kappel am Albis": 6,
|
76 |
+
"Stäfa": 158,
|
77 |
+
"Zell (ZH)": 231,
|
78 |
+
"Turbenthal": 228,
|
79 |
+
"Oberglatt": 92,
|
80 |
+
"Winkel": 72,
|
81 |
+
"Volketswil": 199,
|
82 |
+
"Kilchberg (ZH)": 135,
|
83 |
+
"Wetzikon (ZH)": 121,
|
84 |
+
"Zumikon": 160,
|
85 |
+
"Weisslingen": 180,
|
86 |
+
"Elsau": 219,
|
87 |
+
"Hettlingen": 221,
|
88 |
+
"Rüschlikon": 139,
|
89 |
+
"Stallikon": 13,
|
90 |
+
"Dielsdorf": 86,
|
91 |
+
"Wallisellen": 69,
|
92 |
+
"Dietlikon": 54,
|
93 |
+
"Meilen": 156,
|
94 |
+
"Wangen-Brüttisellen": 200,
|
95 |
+
"Flaach": 28,
|
96 |
+
"Regensdorf": 96,
|
97 |
+
"Niederhasli": 90,
|
98 |
+
"Bauma": 297,
|
99 |
+
"Aesch (ZH)": 241,
|
100 |
+
"Schlieren": 247,
|
101 |
+
"Dürnten": 113,
|
102 |
+
"Unterengstringen": 249,
|
103 |
+
"Gossau (ZH)": 115,
|
104 |
+
"Oberengstringen": 245,
|
105 |
+
"Schleinikon": 98,
|
106 |
+
"Aeugst am Albis": 1,
|
107 |
+
"Rheinau": 38,
|
108 |
+
"Höri": 60,
|
109 |
+
"Rickenbach (ZH)": 225,
|
110 |
+
"Rafz": 67,
|
111 |
+
"Adliswil": 131,
|
112 |
+
"Zollikon": 161,
|
113 |
+
"Urdorf": 250,
|
114 |
+
"Hombrechtikon": 153,
|
115 |
+
"Birmensdorf (ZH)": 242,
|
116 |
+
"Fehraltorf": 172,
|
117 |
+
"Weiach": 102,
|
118 |
+
"Männedorf": 155,
|
119 |
+
"Küsnacht (ZH)": 154,
|
120 |
+
"Hausen am Albis": 4,
|
121 |
+
"Hochfelden": 59,
|
122 |
+
"Fällanden": 193,
|
123 |
+
"Greifensee": 194,
|
124 |
+
"Mönchaltorf": 196,
|
125 |
+
"Dägerlen": 214,
|
126 |
+
"Thalheim an der Thur": 39,
|
127 |
+
"Uetikon am See": 159,
|
128 |
+
"Seuzach": 227,
|
129 |
+
"Uitikon": 248,
|
130 |
+
"Affoltern am Albis": 2,
|
131 |
+
"Geroldswil": 244,
|
132 |
+
"Niederglatt": 89,
|
133 |
+
"Thalwil": 141,
|
134 |
+
"Rorbas": 68,
|
135 |
+
"Pfungen": 224,
|
136 |
+
"Weiningen (ZH)": 251,
|
137 |
+
"Bubikon": 112,
|
138 |
+
"Neftenbach": 223,
|
139 |
+
"Mettmenstetten": 9,
|
140 |
+
"Otelfingen": 94,
|
141 |
+
"Flurlingen": 29,
|
142 |
+
"Stadel": 100,
|
143 |
+
"Grüningen": 116,
|
144 |
+
"Henggart": 31,
|
145 |
+
"Dachsen": 25,
|
146 |
+
"Bonstetten": 3,
|
147 |
+
"Bachenbülach": 51,
|
148 |
+
"Horgen": 295
|
149 |
+
}
|
150 |
+
def predict_apartment(rooms, area, town):
|
151 |
+
bfs_number = locations[town]
|
152 |
+
df = df_bfs_data[df_bfs_data['bfs_number']==bfs_number].copy()
|
153 |
+
df.reset_index(inplace=True)
|
154 |
+
df.loc[0, 'rooms'] = rooms
|
155 |
+
df.loc[0, 'area'] = area
|
156 |
+
df.loc[0, 'proximity_to_public_transportation'] = 500 # Default value
|
157 |
+
|
158 |
+
if len(df) != 1:
|
159 |
+
return -1
|
160 |
+
|
161 |
+
features = ['rooms', 'area', 'pop', 'pop_dens', 'frg_pct', 'emp', 'tax_income', 'proximity_to_public_transportation']
|
162 |
+
X = df[features].values # Convert to numpy array without feature names
|
163 |
+
prediction = random_forest_model.predict(X)
|
164 |
+
return np.round(prediction[0], 0)
|
165 |
+
|
166 |
+
# Create the Gradio interface
|
167 |
+
iface = gr.Interface(
|
168 |
+
fn=predict_apartment,
|
169 |
+
inputs=["number", "number", gr.Dropdown(choices=locations.keys(), label="Town", type="value")],
|
170 |
+
outputs=[gr.Number()],
|
171 |
+
examples=[[4.5, 120, "Dietlikon"], [3.5, 60, "Winterthur"]]
|
172 |
+
)
|
173 |
+
|
174 |
+
iface.launch()
|
bfs_municipality_and_tax_data.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|
random_forest_regression_extended.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b568b08ae4f77ccc8ca6410a8a29b1c36b0b59e9ffed320d810a6d48322613a7
|
3 |
+
size 1432809
|
requirements.txt
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
scikit-learn == 1.6.1
|
2 |
+
matplotlib == 3.10.1
|
3 |
+
pandas == 2.2.3
|
4 |
+
numpy == 2.2.3
|
5 |
+
geopy
|