prediction.py

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

# Load All Files
with open("grid_1_best.pkl", "rb") as file_1:
    grid_1_best = pickle.load(file_1)

with open("model_scaler.pkl", "rb") as file_2:
    model_scaler = pickle.load(file_2)

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

with open("list_num_cols.txt", "r") as file_5:
    list_num_cols = json.load(file_5)

with open("list_catn_cols.txt", "r") as file_6:
    list_catn_cols = json.load(file_6)

with open("list_cato_cols.txt", "r") as file_7:
    list_cato_cols = json.load(file_7)


def run():

  with st.form(key="form_fifa_2022"):
      name = st.text_input("Name", value="")
      age = st.number_input("Age", min_value=16, max_value=120, value=30, step=1, help="Usia")
      st.markdown("---")
      sex = st.selectbox("Gender", ("Female", "Male"), index=1)
      anemia = st.selectbox("History of Anemia", ("Yes", "No"), index=1)
      diabetes = st.selectbox("History of Diabetes", ("Yes", "No"), index=1)
      hbd = st.selectbox("History of High Blood Pressure", ("Yes", "No"), index=1)
      smoking = st.selectbox("History of Smoking", ("Yes", "No"), index=1)
      st.markdown("---")
      creatinine_phosphokinase = st.number_input("Creatinine Phosphokinase", min_value=10, max_value=8000, value=1000, step=1, help="creatinine_phosphokinase")
      ejection_fraction = st.number_input("Ejection Fraction", min_value=5, max_value=200, value=50, step=1, help="ejection_fraction")
      platelets = st.number_input("platelets", min_value=20000, max_value=1000000, value=100000, step=1, help="platelets")
      serum_creatinine = st.number_input("serum_creatinine", min_value=0, max_value=15, value=2, step=1, help="serum_creatinine")
      serum_sodium = st.number_input("serum_sodium", min_value=50, max_value=300, value=100, step=1, help="serum_sodium")
      time = st.number_input("time", min_value=1, max_value=365, value=100, step=1, help="time")


      st.markdown("---")



      submitted = st.form_submit_button("Predict")

  data_inf = {'sex': [1 if x == "Male" else 0 for x in anemia],
            'umur': int((age-1)/5),
            'anaemia': [1 if x == "Yes" else 0 for x in anemia],
            'diabetes': [1 if x == "Yes" else 0 for x in diabetes],
            'high_blood_pressure': [1 if x == "Yes" else 0 for x in hbd],
            'smoking': [1 if x == "Yes" else 0 for x in smoking],
            'creatinine_phosphokinase': creatinine_phosphokinase,
            'ejection_fraction': ejection_fraction,
            'platelets': platelets,
            'serum_creatinine': serum_creatinine,
            'serum_sodium': serum_sodium,
            'time': time
            }
  


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

  if submitted:
      

      # Split between Numerical Columns and Categorical Columns
      data_num = data_inf[list_num_cols] # Kolom numeric
      data_catn = data_inf[list_catn_cols] # Kolom Categorical Nominal
      data_cato = data_inf[list_cato_cols] # Kolom Categorical Ordinal


      # Feature Scaling and Feature Encoding
      data_num_scaled = model_scaler.transform(data_num) # Scalling Data Numeric
      data_catn_encoded = model_encodern.transform(data_catn).toarray() # Encoding Data Nominal
      data_cato_encoded =  data_cato# Encoding Data Ordinal
      data_final = np.concatenate([data_num_scaled, data_cato_encoded, data_catn_encoded], axis = 1) # Gabungan

      # Predict using Linear Regression
      y_pred = grid_1_best.predict(data_final)
      
      st.write("# Rating : ", str(int(y_pred)))

if __name__ == "__app__":
    run()