Comparison Between Python Pandas and Julia DataFrames GroupBy Operations

groupByTiming.jl

using DataFrames

keys = rand(1:100000, 500000);
values = randn(length(keys));

df = DataFrame();
df[:KEY] = keys;
df[:VALUE] = values;

@time by(df, :KEY, x -> sum(x[:VALUE]));

groupByTiming.py

import pandas as pd
import numpy as np
import timeit

keys = np.random.randint(0, 100000, 500000)
values = np.random.normal(size=len(keys))

df = pd.DataFrame()
df["KEY"] = keys
df["VALUE"] = values

def group_func():
    return df.groupby("KEY").sum()

print timeit.timeit(group_func, number=1)