app/utils/data_preparation.py

import pandas as pd
import joblib


def load_model_and_data():
    with open("app/data/lightgbm_deuda.pkl", "rb") as file:
        model = joblib.load(file)

    df = pd.read_csv("app/data/data_finance.csv")
    df = df.set_index("invoiceId")
    return df, model


def prepare_data(df: pd.DataFrame = None) -> pd.DataFrame:
    """
    Prepare data.
    """
    df = df.drop(["Unnamed: 0", "overdueDays"], axis=1)
    df = df.drop(["businessId", "payerId"], axis=1)

    df = df[
        [
            "receiptAmount",
            "relationDays",
            "relationRecurrence",
            "issuerInvoicesAmount",
            "issuerCancelledInvoices",
            "activityDaysPayer",
            "clients12Months",
        ]
    ]
    return df.copy()