Pandas vs Polars vs DuckDB¶

pip install polars

import pandas as pd
import polars as pl
import duckdb

df_pd = pd.read_csv("EuropeSales.csv")
df_pd.head(3)

df_pl = pl.read_csv("EuropeSales.csv")
df_pl.head(3)

con = duckdb.connect()
df_duck = con.execute("SELECT * FROM 'EuropeSales.csv'").df()
df_duck.head(3)

fpd = df_pd[df_pd.Country == "Ukraine"]
fpd.head(10)

fpd = df_pl.filter(pl.col("Country") == "Ukraine")
fpd.head(10)

df_duck = con.execute("SELECT * FROM 'EuropeSales.csv' WHERE Country LIKE 'Ukraine'").df()
df_duck.head(10)

agg_pd = df_pd.groupby("Country")["Total Revenue"].sum().reset_index()
agg_pd.head(10)

agg_pl = df_pl.group_by("Country").agg([pl.sum("Total Revenue").alias("country_revenue")])
agg_pl.head(15)

agg_duck = con.execute("""
    SELECT Country, sum("Total Revenue") AS country_revenue
    FROM 'EuropeSales.csv'
    GROUP BY country
    ORDER BY country
""").df()
agg_duck.head(10)