import streamlit as st
import joblib
import pandas as pd
import json
from io import StringIO

with open('LinearRegression_model.pkl', 'rb') as file_1:
  lin_reg = joblib.load(file_1)

# Load the MinMaxScaler
with open('Scaler.pkl', 'rb') as file_2:
  scaler = joblib.load(file_2)

# Load the OneHotEncoder
with open('OHEncoder.pkl', 'rb') as file_3:
  encoder = joblib.load(file_3)


with open('num_features.txt', 'r') as file_4:
  num_features = json.load(file_4)

with open('cat_features.txt', 'r') as file_5:
  cat_features = json.load(file_5)


st.title('Prediksi Harga Taksi Online')
name = st.selectbox('Service Name',tuple(encoder.categories_[0].tolist()))
#name = st.selectbox('Service Name',('Black','Black SUV','Lux','Lux Black','Lux Black XL','Lyft','Lyft XL','UberPool','UberX','UberXL','WAV'))
distance = st.slider('Distance (miles)',0,20)
sm = st.selectbox('Surge Multiplier',(1.0, 1.25, 2.5 , 2.0  , 1.75, 1.5 , 3.0  ))

new_data = pd.DataFrame({
    'name': [name],
    'distance': [distance],
    'surge_multiplier': [sm]
})

'''
new_data = pd.DataFrame([[name,distance,sm]],columns=['name','disntance','surge_multiplier'])
'''

new_cat_encoded = encoder.transform(new_data[cat_features])
cat_cols_df = pd.DataFrame(new_cat_encoded, columns=encoder.get_feature_names_out(cat_features))
X = pd.concat([cat_cols_df, new_data['surge_multiplier']], axis=1)
num_cols_scaled = scaler.transform(new_data[num_features])
X['distance'] = num_cols_scaled
col = X.pop('distance')
X.insert(0, 'distance', col)
col = X.pop('surge_multiplier')
X.insert(1, 'surge_multiplier', col)

if st.button('Predict'):
  price_pred = lin_reg.predict(X)
  st.subheader('Harga taksi: ${:.2f}'.format(price_pred[0]))