import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import gradio as gr

df = pd.read_csv('https://huggingface.co/spaces/aksrad/GPA/resolve/main/SAT%20GPA.csv')

from sklearn.model_selection import train_test_split

X = df.drop(['*GPA (4.0 Scale)*'], axis=1)
y = df['*GPA (4.0 Scale)*']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

from sklearn.ensemble import RandomForestRegressor

modelrf = RandomForestRegressor(n_estimators=100, random_state=42)
modelrf.fit(X_train, y_train)

#function for prediction
#function for prediction
def predict_gparf(sat_score):
  gpa = modelrf.predict([[sat_score]])
  return gpa[0]

#gradio app
new = gr.Interface(fn=predict_gparf, 
                   inputs= [gr.Number (label= 'SAT_Score') ],
                   title= 'GPA Predictor',  
                   outputs= [gr.Number (label= 'GPA')])
new.launch()