clean.py

import pandas as pd

def clean_upi_data(df):
    df['Date'] = pd.to_datetime(df['Date'], format="%d-%m-%Y")
    df['Category'] = df['Description'].apply(lambda x: categorize_transaction(x))
    return df

def categorize_transaction(description):
    keywords = {
        "Food": ["Zomato", "Swiggy", "Restaurant"],
        "Shopping": ["Amazon", "Flipkart", "Myntra"],
        "Bills": ["Electricity", "Gas", "Water"],
        "Travel": ["Uber", "Ola", "IRCTC"],
    }
    for category, words in keywords.items():
        if any(word.lower() in description.lower() for word in words):
            return category
    return "Other"