import streamlit as st
import pickle
import numpy as np

# Set page title and icon
st.set_page_config(page_title="Employee Promotion Prediction", page_icon=":guardsman:", layout="wide")

# Add a big heading
st.title("Employee Promotion Prediction")

# Change background color to black
st.markdown(
    """
    <style>
    .reportview-container {
        background: black
    }
    </style>
    """,
    unsafe_allow_html=True
)

model = pickle.load(open('rf.pkl', 'rb'))
scaling = pickle.load(open('sc.pkl', 'rb'))

# Define the department options
department_options = ['Analytics', 'Finance', 'HR', 'Legal', 'Operations', 'Procurement', 'R&D', 'Sales & Marketing', 'Technology']

# Define the education options
education_options = ["Bachelor's", "Master's & above", "Below Secondary"]

# Define the gender options
gender_options = ['Male', 'Female']

# Define the awards won options
awards_won_options = ['No', 'Yes']

# Define the maximum training score
max_training_score = 100

# Define the maximum number of trainings
max_trainings = 10

# Define the minimum and maximum age
min_age, max_age = 18, 60

# Define the minimum and maximum length of service
min_service_length, max_service_length = 0, 20

# Define the minimum and maximum previous year rating
min_prev_year_rating, max_prev_year_rating = 1, 5


# Define the input components using Streamlit

department = st.selectbox('---> Department', department_options)
education = st.selectbox('---> Education', education_options)
gender = st.radio('---> Gender', gender_options)
no_of_trainings = st.number_input('---> Number of Trainings', min_value=1, max_value=max_trainings, value=1)
age = st.slider('---> Age', min_value=min_age, max_value=max_age)
previous_year_rating = st.slider('---> Previous Year Rating', min_value=min_prev_year_rating, max_value=max_prev_year_rating)
length_of_service = st.number_input('---> Length of Service', min_value=min_service_length, max_value=max_service_length, value=0)
awards_won = st.selectbox('---> Awards Won?', awards_won_options)
avg_training_score = st.number_input('---> Average Training Score', max_value=max_training_score)


department_encoding = {'Analytics':0, 'Finance':1, 'HR':2, 'Legal':3, 'Operations':4, 'Procurement':5, 'R&D':6, 'Sales & Marketing':7, 'Technology':8}
education_encoding = {"Bachelor's":2, "Master's & above":3, "Below Secondary":1}
gender_encoding = {'Male':1, 'Female':0}
awards_won_encoding = {'No':0, 'Yes':1}

sum_metric = awards_won_encoding[awards_won]  + previous_year_rating
total_score = avg_training_score * no_of_trainings

inputs_to_model = np.array([(department_encoding[department], education_encoding[education],
                             gender_encoding[gender], no_of_trainings, age, previous_year_rating,
                             length_of_service, awards_won_encoding[awards_won], avg_training_score,
                             sum_metric, total_score)])

final_inputs = scaling.transform(inputs_to_model)

# Print the user inputs
if st.button('Submit'):
    # st.write('Department:', department)
    # st.write('Education:', education)
    # st.write('Gender:', gender)
    # st.write('Number of Trainings:', no_of_trainings)
    # st.write('Age:', age)
    # st.write('Previous Year Rating:', previous_year_rating)
    # st.write('Length of Service:', length_of_service)
    # st.write('KPIs Met >80%:', kpis_met)
    # st.write('Awards Won?:', awards_won)
    # st.write('Average Training Score:', avg_training_score)
    # st.write('sum metric', sum_metric)
    # st.write('total score', total_score)
    # st.write('model inputs', inputs_to_model)
    # st.write('final model inputs', final_inputs)

    prediction = model.predict(final_inputs)

    if prediction[0] == 1:
        st.header('The Employee should get Promotion')
        
    else:
        st.header('The Employee should not get promotion')