# PANDAS
df_pd = pd.read_csv("data.csv")
df_pd["date"] = pd.to_datetime(df_pd["date"])

df_pd_res = (
    df_pd
        .groupby([df_pd["date"].dt.year, df_pd["location"]])
        .agg(
            total_sales=("sales", "sum"),
            avg_sales=("sales", "mean")
        )
        .reset_index()
        .sort_values(by=["date", "total_sales"], ascending=[True, False])
)

# POLARS
df_pl_res = (
    pl.read_csv("data.csv")
        .with_columns(pl.col("date").str.to_date("%Y-%m-%d %H:%M:%S"))
        .group_by([pl.col("date").dt.year(), pl.col("location")])
        .agg(
            pl.sum("sales").alias("total_sales"),
            pl.mean("sales").alias("avg_sales")
        )
        .sort(by=["date", "total_sales"], descending=[False, True])
)

# POLARS WITH LAZY EVALUATION
df_pl_lazy_query = (
    pl.scan_csv("data.csv")
        .with_columns(pl.col("date").str.to_date("%Y-%m-%d %H:%M:%S"))
        .group_by([pl.col("date").dt.year(), pl.col("location")])
        .agg(
            pl.sum("sales").alias("total_sales"),
            pl.mean("sales").alias("avg_sales")
        )
        .sort(by=["date", "total_sales"], descending=[False, True])
)

df_pl_lazy_res = df_pl_lazy_query.collect()