-
Notifications
You must be signed in to change notification settings - Fork 0
/
app.py
101 lines (82 loc) · 3.77 KB
/
app.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
from flask import Flask, render_template, request, flash
# from modules.MeanEncoder import MeanEncoder
from datetime import datetime
from modules.utils import *
from etl import single_distance_to
import pandas as pd
import numpy as np
import joblib
import yaml
import os
app = Flask(__name__)
app_logger = add_custom_logger('app', file_path='logs/app.log', streaming=True)
with open('./config.yaml', 'r') as file:
config = yaml.safe_load(file)
# Debug mode should be off if hosted on an external website
debug= config['local']
# Model version is determined by the config file, however if use_curr_datetime is set to True, then it will try to search for most recent model_version
model_version = config['use_model_version']
# Accounts for filepathing local and in pythonanywhere
if config['local']:
pass
else:
os.chdir(config['web_directory'])
# Load the model, scaler and encoders
prediction_model = joblib.load(f'./models/gbc_{model_version}.joblib')
scaler = joblib.load(f'./models/scaler_{model_version}.joblib')
mean_encoder = joblib.load(f'./models/mean_encoder_{model_version}.joblib')
# Alternative to pickling my own Class, set the encoder using a json
# mean_encoder = MeanEncoder()
# mean_encoder.set_from_json(f'models/encoding_dict_{model_version}.json')
# Flask Routing methods
@app.route("/")
def index():
return render_template('index.html')
@app.route('/dataset')
def dataset():
return render_template('dataset.html')
@app.route('/model')
def model():
return render_template('model.html')
@app.route('/predict', methods=('GET', 'POST'))
@timeit
def predict():
prediction=None
if request.method == 'POST':
# Build a Pandas DataFrame using the post info
data = {'floor_area_sqm': float(request.form['floor_area_sqm']),
'remaining_lease':float(request.form['remaining_lease']),
'avg_storey': float(request.form['avg_storey']),}
for_mean_encoding = pd.DataFrame({'town': request.form['town'],
'rooms':float(request.form['rooms'])},
index=[0])
df = pd.DataFrame(data, index=[0])
# Calculate distance to marina bay through OneMap API call
try:
df['dist_to_marina_bay'] = single_distance_to(request.form['address'], 'Marina Bay', verbose=1)
except Exception as error:
app_logger.error(error, exc_info=True)
flash('Unable to get location of address given, please try again.')
return render_template('predict.html')
# Mean encoding
df['mean_encoded'] = mean_encoder.transform(for_mean_encoding)
app_logger.info(f'Prediction for\n{df}')
df = scaler.transform(df)
# Prediction
try:
prediction = int(prediction_model.predict(df)[0])
rounded_prediction = round(prediction, -3)
app_logger.info(f'Prediction made at {datetime.now()}: {rounded_prediction} ({prediction})')
except ValueError as error:
app_logger.error(error, exc_info=True)
flash('No such type of flat found in Town specified, please try again.')
return render_template('predict.html')
return render_template('predict.html', prediction=rounded_prediction)
# GET request
elif request.method == 'GET':
return render_template('predict.html', prediction=prediction)
# Main()
if __name__ == "__main__":
# Getting the credentials for the session and database access
app.secret_key = os.environ['FLASK_KEY']
app.run(debug=debug)