model/app.py

import gradio as gr
import pandas as pd
import pickle
import os

MAIN_FOLDER = os.path.dirname(__file__)

# Define params names
PARAMS_NAME = [
    'orderAmount', 
    'orderState',
    'paymentMethodRegistrationFailure',
    'paymentMethodType',
    'paymentMethodProvider',
    'paymentMethodIssuer',
    'transactionAmount',
    'transactionFailed',
    'emailDomain',
    'emailProvider',
    'customerIPAddressSimplified',
    'sameCity',
]

# Load model
with open("model/modelo_proyecto_final.pkl", "rb") as f:
    model = pickle.load(f)

# Columnas
COLUMNS_PATH = "model/categories_ohe_without_fraudulent.pickle"
with open(COLUMNS_PATH, 'rb') as handle:
    ohe_tr = pickle.load(handle)



# ver como queda el encadenado de las carpetas
BINS_ORDER=os.path.join(MAIN_FOLDER,"model/saved_bins_order.pickle")
with open (BINS_ORDER, 'rb') as handle:
    new_saved_bins_order=pickle.load(handle)


BINS_TRANSACTION=os.path.join(MAIN_FOLDER, "model/saved_bins_transaction.pickle")
with open (BINS_TRANSACTION, 'rb') as handle:
    new_saved_bins_transaction=pickle.load(handle)   




def predict(*args):
    answer_dict = {}

    for i in range(len(PARAMS_NAME)):
        answer_dict[PARAMS_NAME[i]] = [args[i]]

    single_instance = pd.DataFrame.from_dict(answer_dict)


    single_instance["orderAmount"]=single_instance["orderAmount"].astype(float)
    single_instance["orderAmount"]=pd.cut(single_instance["orderAmount"],
                                                                        bins=new_saved_bins_order,
                                                                        include_lowest=True)
    single_instance["transactionAmount"]=single_instance["transactionAmount"].astype(int)
    single_instance["transactionAmount"]=pd.cut(single_instance["transactionAmount"],
                                                                        bins=new_saved_bins_order,
                                                                        include_lowest=True)





    
    # Reformat columns
    single_instance_ohe = pd.get_dummies(single_instance).reindex(columns = ohe_tr).fillna(0)
    
    prediction = model.predict(single_instance_ohe)

    type_of_fraud=int(prediction[0])
    
    response = {"tipo de fraude":type_of_fraud}

    # Adaptación respuesta
    response = "Error parsing value"
    if type_of_fraud == 0:
        response = "No fraud"
    if type_of_fraud == 1:
        response = "Fraud"
    if type_of_fraud == 2:
        response = "Revisar"


    return response


with gr.Blocks() as demo:
    gr.Markdown(
        """
        # DETECTION OF FRAUD 🔧🚜
        """
    )

    with gr.Row():
        with gr.Column():

            gr.Markdown(
                """
                ## Predecir Fraude.
                """
            )
            
            orderState = gr.Radio(
                label="estado orden",
                choices=["pending", "fulfilled", "failed"],
                value="pending"
                )

            paymentMethodRegistrationFailure=gr.Radio(
                label="failure payment method",
                choices=["True", "False"],
                value="True"
                )

            paymentMethodType=gr.Radio(
                label="metodo pago",
                choices=["card", "apple pay", "paypal", "bitcoin"],
                value="card"
                )


            paymentMethodProvider=gr.Radio(
                label="proveedor pago",
                choices=["JCB 16 digit","VISA 16 digit","Voyager","Diners Club / Carte Blanche","Maestro","VISA 13 digit","Discover","American Express","JCB 15 digit","Mastercard"],
                value="Mastercard"
                )

            paymentMethodIssuer=gr.Radio(
                label="emisor",
                choices=["Her Majesty Trust","Vertex Bancorp","Fountain Financial Inc.","His Majesty Bank Corp.","Bastion Banks","Bulwark Trust Corp.","Citizens First Banks","Grand Credit Corporation","Solace Banks","Rose Bancshares","B","e","c","r","n","x","o","a","p"],
                value="o"
                )

            transactionFailed=gr.Radio(
                label="transacion fallida",
                choices=["True", "False"],
                value="True"
                )                

            emailDomain=gr.Radio(
                label="dominio",
                choices=["weird","com","biz","org","net","info"],
                value="weird"
                )

            emailProvider=gr.Radio(
                label="proveedor",
                choices=["weird","ohter","gmail","yahoo","hotmail"],
                value="weird"
                )

            customerIPAddressSimplified=gr.Radio(
                label="ip",
                choices=["only letters","digits_and_letters"],
                value="only letters"
                )

            sameCity=gr.Radio(
                label="misma ciudad",  
                choices=["unknown","no","yes"],     
                value="yes"
                ) 
                
                
            

            orderAmount = gr.Slider(label="Order Amount", minimum=0, maximum=1000, step=1, randomize=True)

            transactionAmount=gr.Slider(label="Transaction Amount", minimum=0, maximum=1000, step=1, randomize=True)
            
            
            
            

            
            
            
        with gr.Column():

            gr.Markdown(
                """
                ## Predicción
                """
            )

            label = gr.Label(label="Fraud_detection")
            predict_btn = gr.Button(value="Evaluar")
            predict_btn.click(
                predict,
                inputs=[
                   orderAmount, 
                   orderState,
                   paymentMethodRegistrationFailure,
                   paymentMethodType,
                   paymentMethodProvider,
                   paymentMethodIssuer,
                   transactionAmount,
                   transactionFailed,
                   emailDomain,
                   emailProvider,
                   customerIPAddressSimplified,
                   sameCity,
                ],
                outputs=[label],
            )
    gr.Markdown(
        """
        <p style='text-align: center'>
            <a href='https://www.escueladedatosvivos.ai/cursos/bootcamp-de-data-science' 
                target='_blank'>Proyecto demo creado en el bootcamp de EDVAI 🤗
            </a>
        </p>
        """
    )

demo.launch()