app.py

import warnings
warnings.filterwarnings('ignore')
import pandas as pd
import numpy as np
from sklearn import preprocessing
from sklearn.ensemble import GradientBoostingClassifier
import pickle
import gradio as gr

# Load the model and encoder
with open("model.pkl", "rb") as model_file:
    gbc = pickle.load(model_file)

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

def predict_employability(Age, Accessibility, EdLevel, Employment, Gender, MentalHealth, MainBranch, YearsCode, PreviousSalary, ComputerSkills, Continent):
    data = {
        'Age': Age, 'Accessibility': Accessibility, 'EdLevel': EdLevel, 'Employment': Employment, 
        'Gender': Gender, 'MentalHealth': MentalHealth, 'MainBranch': MainBranch, 
        'YearsCode': YearsCode, 'PreviousSalary': PreviousSalary, 'ComputerSkills': ComputerSkills, 
        'Continent': Continent
    }
    df = pd.DataFrame([list(data.values())], columns=[
        'Age', 'Accessibility', 'EdLevel', 'Employment', 'Gender', 'MentalHealth', 
        'MainBranch', 'YearsCode', 'PreviousSalary', 'ComputerSkills', 'Continent'
    ])
    category_col = ['Age', 'Accessibility', 'EdLevel', 'Gender', 'MentalHealth', 'MainBranch', 'Continent']
    for cat in encoder_dict:
        for col in df.columns:
            le = preprocessing.LabelEncoder()
            if cat == col:
                le.classes_ = np.array(encoder_dict[cat], dtype=object)
                for unique_item in df[col].unique():
                    if unique_item not in le.classes_:
                        df[col] = ['Unknown' if x == unique_item else x for x in df[col]]
                df[col] = le.transform(df[col].astype(str))
    features_list = df.values.tolist()
    prediction = gbc.predict(features_list)
    return "Employable" if prediction[0] == 1 else "Less Employable"

iface = gr.Interface(
    fn=predict_employability,
    inputs=[
        gr.Dropdown(['<35', '>35'], label="Age"),
        gr.Dropdown(['No', 'Yes'], label="Accessibility"),
        gr.Dropdown(['Master', 'NoHigherEd', 'Other', 'PhD', 'Undergraduate'], label="EdLevel"),
        gr.Slider(0, 1, step=1, label="Employment"),
        gr.Dropdown(['Man', 'NonBinary', 'Woman'], label="Gender"),
        gr.Dropdown(['No', 'Yes'], label="MentalHealth"),
        gr.Dropdown(['Dev', 'NotDev'], label="MainBranch"),
        gr.Slider(0, 50, step=1, label="YearsCode"),
        gr.Number(label="PreviousSalary"),
        gr.Slider(0, 100, step=1, label="ComputerSkills"),
        gr.Dropdown(['Africa', 'Asia', 'Europe', 'North_America', 'Oceania', 'Others', 'South_America'], label="Continent")
    ],
    outputs="text",
    title="Employability Prediction",
    description="Predict employability based on various features."
)

iface.launch()