Why would you (not) choose R over Python for Data Science?

tidyverse, RStudio and native data frames.

Because Python is a well-thought-out language

Because often I have to embed analyses into existing python applications. The choice is actually quite obvious then ;)

Python is generic, people know it, and it works well enough. R is specialised, fewer people know it, it works *very* well at what it does, but for most of us the significant additional time to learn it just isn't worth the advantages it brings.

Python is general purpose.
R is generally used by researchers or people looking to do only analysis

Python is good for when you want to build an application or a system.

R is good for data exploration and research.

Depends on what kind of data you're working with. A one-off standard dataset means I'll use python/pandas/numpy because its quick and painless, but if i need to continuously process big data, I'm going R. It's faster with larger datasets and with more transformations.

I actually like R.
I can say that I find it nice and fun for extracting the most out of a data set.
I used it to help my job, but never as the core of my job because:
a) I am not doing a data analysis based job
b) I find it quite hard to program a complex yet clean code for big programs