[IPython-User] nbconvert converts to/from IPyNB or just from IPyNb?

[Matthias BUSSONNIER]

Le 24 févr. 2013 à 01:48, Nitin Borwankar a écrit :

> Matthias,
> 
> Re: your comment about knitr - yes knitr is similar to what I am saying but IPyNB metaphor is, IMHO, superior as it is not tied to a single language. Just my massively unashamedly biased opinion.

Knitr also support other languages I think 
http://yihui.name/knitr/demo/engines/

> 
> I am using knitr for a Coursera class right now - I like it a lot but I think pandoc + markdown + IPyNB ends up winning.  Right now R is way ahead due to very mature stats libs and easy to use for statistician non-programmers.  
> 
> I use %%R a lot for my content creation. Nbviewer views coming soon.

Great ! 


> In my opinion general purpose languages win in the longer term as they have law of large numbers (of users) on their side and they get better and better.  R object model and data types and data slicing can cause irreversible brain damage especially to an old time SQL user like myself. I spend hours resuscitating hundreds of thousands of neurons every day after my R sessions. :-)
> 
> But OTOH, ggplot is very cool and where R works it works really well.  So I have Stockholm Syndrome with R 
> :-)
> 
> Nitin Borwankar
> 

-- 
Matthias