src/data_preprocessing.py

import pandas as pd

def load_parquet_data(split="train_sample"):
    df = pd.read_csv(f"{split}.csv")
    return df

def preprocess_parquet(df):
    # Skip filtering 'event_type' because it's already been preprocessed
    df = df[["user_id", "product_id", "rating"]]
    df = df.groupby("user_id").filter(lambda x: len(x) >= 5)
    return df