import numpy as np
import gradio as gr

def ovarian(Shadow, Solid, Menopause, CA125):

    if Shadow == "Yes":
      Shadow = 1
    else:
      Shadow = 0
    if Solid == "Yes":
      Solid = 1
    else:
      Solid = 0
    if Menopause == "Yes":
      Menopause = 1
    else:
      Menopause = 0
    if CA125 == "Yes":
      CA125 = 1
    else:
      CA125 = 0
    
    #print(Shadow, Solid, Menopause, CA125)
    logit_P = -3.771109 - 2.293585*Shadow + 3.877268*Solid + 1.76309*Menopause + 1.320551*CA125
    e_logit_P = 2.71828182846 ** logit_P
    prob = e_logit_P/(1+e_logit_P) * 100
    return "Risk of Ovarian cancer in this patient : \n" + str('%.3f' %(prob)) + " %" + "\n[Following up on the Predictive Model for Ovarian Cancer.]"

demo = gr.Interface(
    ovarian,
    [
        gr.Radio(["Yes", "No"]),
        gr.Radio(["Yes", "No"]),
        gr.Radio(["Yes", "No"]),
        gr.Radio(["Yes", "No"])
    ],
    "text",
    examples=[
        ["Yes", "Yes", "Yes", "No"],
    ],
    title="Predictive model of Ovarian cancer",
    description="Here is a predictive model for ovarian cancer. Choose any four of the risk factors listed below for your patient. This model allows you to calculate the risk of ovarian cancer based on these selected factors.",
)

demo.launch()