import gradio as gr
import pandas as pd
import torch
from fastai.tabular.all import load_learner

learn = load_learner('Diabetes.pkl')
coeffs = torch.load('Dianetes_Neural_Network.pkl')

import torch.nn.functional as F

def coeff(coeffs, indeps):
    layers,consts = coeffs
    n = len(layers)
    res = indeps
    for i,l in enumerate(layers):
        res = res@l + consts[i]
        if i!=n-1: res = F.relu(res)
    return torch.sigmoid(res)

"""
def predict(Pregnancies, Glucose, BloodPressure, SkinThickness, Insulin, BMI, DiabetesPedigreeFunction, Age):
    inputs = {'Pregnancies': Pregnancies,
              'Glucose': Glucose,
              'BloodPressure': BloodPressure,
              'SkinThickness': SkinThickness,
              'Insulin': Insulin,
              'Age': Age,
              'DiabetesPedigreeFunction': DiabetesPedigreeFunction,
              'BMI': BMI}

    df = pd.DataFrame([inputs])

    pred, _, _ = learn.predict(df.iloc[0])

    return pred.item()

iface = gr.Interface(fn=predict, 
                     inputs=["number", "number", "number", "number", "number", "number", "number", "number"], 
                     outputs="number",
                     description="Insira os dados para prever o resultado de diabetes")

iface.launch()
"""

def predict(Pregnancies, Glucose, BloodPressure, SkinThickness, Insulin, BMI, DiabetesPedigreeFunction, Age):
    inputs = torch.tensor([Pregnancies, Glucose, BloodPressure, SkinThickness, Insulin, BMI, DiabetesPedigreeFunction, Age]).float()

    pred = coeff(coeffs, inputs)

    return pred.item()

iface = gr.Interface(fn=predict, 
                     inputs=["number", "number", "number", "number", "number", "number", "number", "number"], 
                     outputs="number",
                     description="Insira os dados para prever o resultado de diabetes")

iface.launch()