Josiah Jul 6, 2023

Stop staying you can't put R in prod.  

I made a blog post with my thoughts and reactions.

#rstats #rust #putRinprod

https://josiahparry.com/posts/2023-07-06-r-is-still-fast.html

Josiah Parry - R is still fast: a salty reaction to a salty blog post

Show threadorizuruJul 6, 2023

@josiah How about maintainability?

Never met a developer who liked R, and to put things into production, you have to work with devs.
The new people coming into data science all know python. Even among older data scientists, you rarely find one that prefers R (unless they're a statistician working in academia).

Who is going to maintain that code after the person who wrote it leaves for another company?

Show threadJoris Meys
@orizuru @josiah funny, I meet them all the time.
Jul 9, 2023 at 10:12amWeb
Show threadorizuruJul 9, 2023

@JorisMeys @josiah outside of academia, where stuff is normally put into production, this is very rare. In these fields (which seems to be the focus of the blog post), R is a dying language. Picking a language to put into production has to take into account the potential pool of people who you could hire to maintain it.

If you can present convincing data otherwise, I'm all ears.

Show threadJoris MeysJul 9, 2023

@orizuru @josiah pharmaceutics is a huge industry. As is biotech in general. Both fields use a ton of R code in production.

As does BBC, Financial Times,... for infographics etc. Nobody in his right mind uses matplotlib for that.

Show threadorizuruJul 9, 2023

@JorisMeys @josiah

I'm repeating myself, by "putting into production", I mean running in a kube cluster, not performing some analysis.

The example in the blogpost is a webserver. I have never seen one written in R.

Show threadJoris MeysJul 9, 2023
@orizuru @josiah please make yourself familiar with how that industry works. "Some analysis" is highly automated production pipelines with data entry and management from different test sites ( in different continents sometimes) on huge systems that do run R code. You're being silly, mate.