import streamlit as st
import pandas as pd
import numpy as np
import pickle
import json

# Load model and other files

with open('list_cat_cols.txt', 'r') as file_1:
    list_cat_col = json.load(file_1)

with open('list_num_cols.txt', 'r') as file_2:
    list_num_col = json.load(file_2)

with open("model.pkl", "rb") as file_3:
    model = pickle.load(file_3)

with open("scaler.pkl", "rb") as file_4:
    scaler = pickle.load(file_4)

with open("encoder.pkl", "rb") as file_5:
    encoder = pickle.load(file_5)

def run():
    with st.form('form_fifa_2022'):
        name = st.text_input('Name', value =' ') # value = default
        age = st.number_input('Age', value =25, min_value = 15, max_value = 60, help = 'isi dengan usia pemain')
        height = st.number_input('Height', value = 170, min_value = 100, help = 'in centimeter')
        weight = st.slider('Weight', value = 70, min_value = 50, max_value = 150, help = 'in kilogram')
        ValueEUR = st.number_input('ValueEUR', value = 0)
        st.markdown('---')
        attacking_work_rate = st.selectbox('AttackingWorkRate', ('Low', 'Medium', 'High'), index = 1) # index = default di selectbox/radio
        defensive_work_rate = st.radio('DefensiveWorkRate', ('Low', 'Medium', 'High'), index=1)
        pace = st.number_input('Pace', min_value = 0, max_value = 100, value = 50)
        shooting = st.number_input('Shooting Score', min_value = 0, max_value = 100, value = 50)
        passing = st.number_input('Passing Score', min_value = 0, max_value = 100, value = 50)
        dribbling = st.number_input('Dribbling Score', min_value = 0, max_value = 100, value = 50)
        defending = st.number_input('Defending Score', min_value = 0, max_value = 100, value = 50)
        physicality = st.number_input('Physicality Score', min_value = 0, max_value = 100, value = 50)
        
        # create submit button form
        submitted = st.form_submit_button('Predict')

    data_inf = {
        'Name' : name,
        'Age' : age,
        'Height' : height,
        'Weight' : weight,
        'ValueEUR' : ValueEUR,
        'AttackingWorkRate' : attacking_work_rate,
        'DefensiveWorkRate' : defensive_work_rate,
        'PaceTotal' : pace,
        'ShootingTotal' : shooting,
        'PassingTotal' : passing,
        'DribblingTotal' : dribbling,
        'DefendingTotal' : defending,
        'PhysicalityTotal' : physicality
    }

    data_inf = pd.DataFrame([data_inf])
    st.dataframe(data_inf) 

    # split between numerical and categorical columns
    data_inference_num = data_inf[list_num_col]
    data_inference_cat = data_inf[list_cat_col]

    data_inference_num_scaled = scaler.transform(data_inference_num)
    data_inference_cat_encoded = encoder.transform(data_inference_cat)
    data_inference_final = np.concatenate([data_inference_num_scaled, data_inference_cat_encoded], axis = 1)

    # predict using linear reg model
    predicted_score = model.predict(data_inference_final)

    st.write('## Rating: ', str(int(predicted_score)))

if __name__ == '__main__':
   run()