app.py

#Bismillahir Rahmaanir Raheem
#Almadadh Ya Gause RadiAllahu Ta'alah Anh - Ameen


import gradio as gr
import pandas as pd
from pycaret.classification import load_model, predict_model


# load the trained model for predictions
model = load_model('tuned_blend_specific_model_19112021')


# define the function to call 
def predict(model, input_df):
	predictions_df = predict_model(estimator=model, data=input_df)
	predict_label = predictions_df["Label"][0] # either 1 (amputation yes) or 0 (amputation no)
	predict_score = predictions_df["Score"][0] # the prediction (accuracy)
	amputation_risk = ""
	if predict_label == 1:
		amputation_risk = "YES"
		amputation_risk_output = "Amputation Risk: " + amputation_risk
		score_output = "Score: "+str(predict_score)
		
		html = "<div style='background-color:rgb(153, 0, 0);color:white;font-size:20px;'>" + amputation_risk_output + "<br>" + score_output + "<br>" + "</div>"
	else:
		amputation_risk = "NO"
		amputation_risk_output = "Amputation Risk: " + amputation_risk
		score_output = "Score: "+str(predict_score)
		html = "<div style='background-color:rgba(16, 185, 129, var(--tw-bg-opacity));color:white;font-size:20px;'>" + amputation_risk_output + "<br>" + score_output + "<br>" + "</div>"
	
	return html#"AMPUTATION RISK: " + amputation_risk + " SCORE: "+str(predict_score)  


# the parameters in this function, actually gets the inputs for the prediction 
def predict_amputation(age, gender, race, diabetes_type):
	diabetes_class = "Type "+str(diabetes_type)+" diabetes"
	gender = gender[0]
	input_dict = {"AGE": age, "GENDER": gender, "RACE": race, "DIABETES_CLASS":diabetes_class, "AMPUTATION":''}
	
	input_df = pd.DataFrame([input_dict])
	
	# output 
	return str(predict(model=model, input_df=input_df)) # calls the predict function when 'submit' is clicked


title = "DIabetes-related Amputation Risk Calculator (DIARC)"

description = "A diabetes-related amputation machine learning model trained on the diabetes dataset from the Inkosi Albert Luthuli Central Hospital (IALCH) in Durban, KwaZulu-Natal, South Africa."

article = "<p style='text-align: center'><span style='font-size: 15pt;'>Copyright &copy; DIARC. 2021. All Rights Reserved. Contact Us: <a href='mailto:smtshali@wol.co.za'>Dr Sifiso Mtshali</a> or <a href='mailto:mahomedo@ukzn.ac.za'>Dr Ozayr Mahomed</a></span></p>"


iface = gr.Interface(
		fn=predict_amputation, 
		title=title, 
		description=description, 
		article=article,
		inputs=[gr.inputs.Slider(minimum=0,maximum=100, step=1, default=0, label="Age"), 
		        gr.inputs.Dropdown(["Female", "Male"], default="Female", label="Gender"), 
				gr.inputs.Dropdown(["Asian", "Black", "Coloured", "White", "Other"], default="Asian", label="Race"), 
				gr.inputs.Dropdown(["1", "2"], default="1", label="Diabetes Type")], 
		outputs="html",
		theme="grass",
		examples=[
			[77, "Female", "Asian", 2],
			[28, "Male", "Black", 1],
			[75, "Male", "White", 2],
			[59, "Male", "Coloured", 1],
			[73, "Female", "Other", 1],
			[4, "Female", "Black", 2],
			[65, "Male", "Coloured", 2],
		],
)


iface.test_launch()		
if __name__ == "__main__":
	iface.launch()