|
import pickle |
|
import numpy as np |
|
|
|
__model =None |
|
__location_encoder = None |
|
__location_list = None |
|
|
|
|
|
def load_assests(): |
|
global __model |
|
global __location_encoder |
|
global __location_list |
|
|
|
with open('assets/banglore_price_prediction_model.pickle', 'rb') as f: |
|
__model = pickle.load(f) |
|
with open('assets/location_encoder.pickle', 'rb') as ld: |
|
__location_encoder= pickle.load(ld) |
|
__location_list = __location_encoder.categories_[0] |
|
|
|
def get_estimated_price(location,bhk,tsqft,bath): |
|
try: |
|
x = __location_encoder.transform([[location]]).toarray()[0] |
|
except: |
|
x = np.zeros(len(__location_list)) |
|
|
|
x = np.append(x[1:], np.array([bhk, tsqft, bath])) |
|
return __model.predict(x.reshape(1, -1))[0] |
|
|
|
|
|
|
|
|
