S4 vs S3. New Package

This depends very often of on the developer and what he is comfortable with. I like S4 classes, as I come from C++ and S4 classes approximate C++ classes at least more than S3 classes do (Reference Classes would do so even more and I know very good R programmers liking these most).

1) I wrote a package for MCMC simulation with S4 classes carrying all simulated values - fast enough for me: in less than 1.5 secs I have my sample of 100.000 values together with several other 100T values like log-likelihoods, posterior hyper parameters, etc. I watch out for not copying too often an object but sometimes it is not avoidable. 

2) That is not true: 
    
Books:
http://www.amazon.de/Software-Data-Analysis-Programming-Statistics/dp/0387759352/ref=sr_1_1?ie=UTF8&qid=1384014486&sr=8-1&keywords=John+chambers+data
http://www.amazon.de/Programming-Data-Language-John-Chambers/dp/0387985034/ref=sr_1_4?ie=UTF8&qid=1384014486&sr=8-4&keywords=John+chambers+data

Online:
https://www.rmetrics.org/files/Meielisalp2009/Presentations/Chalabi1.pdf
https://www.stat.auckland.ac.nz/S-Workshop/Gentleman/S4Objects.pdf

And for a bunch of packages look into the Bioconductor packages.

Best

Simon

Hi,

I am working on a new credit portfolio optimization package. My question is
if it is more recommended to develop in S4 object oriented or S3.

It would be more naturally to develop in object oriented paradigm, but
there is many concerns regarding S4.

1) Performance of S4 could be an issue as a setter function, actually
changes the whole object behind the scenes.

2) Documentation. It has been really hard to find examples in S4. Most
books and articles consider straightforward S3 examples.

Thanks,

	[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.