Related Posts
More Posts
Jummah Mubarak Fam ☪️💚
Any tips or tricks for CSX certification?
Additional Posts in Consulting Exit Opportunities
Pros and Cons working for the fed bank?
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.







I keep hearing Python can do everything R can but it’s simply not true. You can execute analyses faster, in fewer lines of code in R. The whole language is built around data analysis. There isn’t a python IDE as nice as RStudio and Dash isn’t comparable to RStudio Connect.
Python lacks a wide range of statistics packages when compared to R (basically everything that is cutting edge/specialized). I’m note sure how easy it is to bind Python to C++ but in R it is crazy simple with Rcpp.
That said Python has its advantages in the big data space and NLP. Most ML libraries are released to python first. R doesn’t have a binding for Gluon yet. However, much of this only applies if you are working on NNs.
If I was learning a language today I might start with R and then learn Julia. Julia 1.0 was finally released and it is blazing fast when compared to python or R unless you awkwardly chain a bunch of libraries together. And both R and python will be slower for nontrivial code as Julia’s LLVM can optimize the whole program / function. The ability to stay in a single language is great. I have high hopes for JuliaDB in that you can compile any arbitrary function and pass it to the database. It’s all really cool stuff imo. However it is still immature
R but understand business use cases that utilized R as a back end. Then start thinking how you leverage R for your current and previous clients and value you could bring.
Learning the language is the easy part (:
Good luck OP!
I actually have experience with both and would say learn Python. It has all the same functionality as R, but is more universal (R is primarily data analysis, while python is that and more). In my opinion it’s also more intuitive.
Definitely python, if you choosing between languages. If you thinking whether to learn at all, learn basics and put it in CV, then you’ll quickly learn at the job. Usually python training is done by the L&D team tailored to company needs. Don’t learn just for resume, put it there and then figure it out.
1. For statistical analysis use R
2. For large scale data processing use python (ex : pandas, numpy, scikit )
Pandas and numpy are not for large scale analysis. Anything that fits in memory is small by definition and you can fit most things (or a large enough sample) in memory these days (eg VM instances with 2tb of RAM). If pandas is for large scale analysis then data.table in r is equally useful for large scale analysis.
The only way numpy gets large is if you use dask. And in that situation most folks would be using spark/presto/hive/mpi for analysis as few people need numerical computing at that scale.