R and Python Unite!

[GRULSTMRNN]

Depends on what you mean with "deployment". Exposing a deep learning inference engine is easy, there is no magic regardless of which language one uses. Developing and training useful models and presenting them to end users in ways users are willing to pay money for takes thousands of lines of code, PhDs, senior developers with many years of experience. It's kind of naive to suggest that deployment of one api is easy. Yes it is but it's also just one of dozens of apis and considerations in the entire product development, testing, and deployment, and monetization process.

Depends on the application. I write PySpark pipelines for big data all the time and deploy models in 50 lines of code or less. You don't have to write much code for model deployment any more, as much of the work has been subsumed by high-level ML APIs or products such as DataRobot.

https://developers.google.com/machine-learning/guides/rules-of-ml/

Regards,

PTR

 

[kiwi_trader]

FWIW ... Python is based primarily on C not C++. Where extra speed is required modules are usually written in C or Cython including modules like Numpy & Pandas.

(Not a fan of R but do like Python, C, and Nim )

 

[blueraincap]

Saeed amen also primarily uses python

 

[blueraincap]

www.timothymasters.info there is internet access in a deep country?

 

[NoiseTraderr]

I would just add jupyter notebooks work nicely with R and is easy to get setup.
I mostly use python but there are just so many interesting packages for R.

 

[El Trado]

An alternative is Rpy2. It works fairly ok, and let you use the few R functions you might not find already in place in Python: https://rpy2.bitbucket.io/