I am running some simulations at work, and I wanted to experiment with different approaches for performing a groupby and then running the simulation. The idea is to run a simulation of each group (identified by an id column) using multiple datasets all keyed by the same id column. There are about 5 …

That’s a very interesting post. I book marked it for future reference as I work to solve some different applications. There’s so many different scenarios for groupby and join.

Timings for different groupby approaches

Specific Domains Data

xiaodai January 23, 2020, 1:50am 2

This post is still very relevant Group-by performance benchmarks and recommendations - #9 by xiaodai

Topic		Replies	Views
Group-by performance benchmarks and recommendations Data	12	3608	September 2, 2019
A minor group-by benchmark - DataFrames.jl plenty fast General Usage	5	511	August 27, 2020
Tables package for fast grouping and filtering? Performance package	18	1670	December 8, 2019
DataFrames operation scales badly Performance	21	2856	December 10, 2018
Various by-group strategies compared Data	36	4168	January 30, 2018

Timings for different groupby approaches

Related topics