[R] R or weka

Christian Schulz ozric at web.de
Tue Feb 8 15:07:53 CET 2005


it's dependend of your task and the amount of data.
R-project have a lot more possibilities  manipulate the data and  many 
different plot possibilities.
Weka have the nice possibilty use the experimenter which could test 
several algorithms
on different dataset's for comparison, further you find 
association-algorithms which
doesn't exist in R-project if i'm correct.
Neverthless you could program  a R-project script which doing the same
like the experimenter in weka, but it's really more work for a beginner 
and a advanced user, too.

IMHO  if you want doing really data-mining ( i.e. more than ~ 200.000 
rows and 20 attributes)
with many recodings, i recgonzie for both software-packages
the best way is doing all recodings in a database ( i.e. MySQL,RMySQL)  
and  only mine (train , test , predict etc.)
in weka or R-project, because you get crazy when you didn't have a 
machine with a lot of Speed and  RAM.

hope it helps,
regards, Christian

WeiWei Shi wrote:

>Hi, guys:
>These days I keep using R and Weka to do data mining. I think my next
>step is open the source codes so that I can "customrize them" and make
>them better server my purpose. But now I kinda hesitate to do so b/c I
>am really not sure which is better to start with. You know, both
>require some time and I cannot clone myself to work on both:) If here
>are some persons who used both of them, please provide some
>suggestions from your experience.
>R-help at stat.math.ethz.ch mailing list
>PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

More information about the R-help mailing list