What’s your choice?

why not both?

python of course, because it's useful beyond mschine learning.
R you can stll learning that later when you've become an expert in ml.

Python is bae <3

Py-fucking-thon 3.

Learning Scala :p

In R array starts from 1. So u know what R stands for ret..
If u have some prior programming language practice with array starting from 1 you may feel some problems in calculation. Not any sarcastic veiw but a warning

Matlab!

R-R-Retarded

Learn scala

I learned both. I preferred Python for a long while but recently took a deeper liking for R. Base R isn’t really my cup of tea but the TidyVerse family of packages implement a consistent design philosophy that allows for very quick and powerful workflows, great reuse of code, closely integrated plotting (I was actually using ggplot2 with Jupyter magic while I was using python anyway) and great Functional Programming utility.

My data pipeline is in Haskell anyway and I’m embedding my R scripts in Haskell systems, so arrays thankfully start at zero and data is fed to R only at the stage that it’s ready to go into a tibble and down some TidyVerse pipes for analysis.