[R] Do you use R for data manipulation?

[Frank E Harrell Jr]

Simon Pickett wrote:
> My institute uses SAS religiously, I am the only R "heathen".
> 
> I have resisted learning to use SAS because I dont see the point after 
> years of using R and I like being able to do everything using one 
> program. However, my colleagues maintain that SAS is "better" for 
> programming without really ever giving me a good reason why other than 
> memory issues.
> 
> dont want to hi-jack the thread but would be interested in hearing some 
> other views, especially since my organisation spends (wastes?) alot of 
> money every year on SAS licences...

Put quite simply, your colleagues' opinions are humbug.

Frank

> 
> Simon.
> 
> ----- Original Message ----- From: "Laura Arsanto" <ghina84 at hotmail.it>
> To: <jrkrideau at yahoo.ca>; <r-help at stat.math.ethz.ch>; <fjbuch at gmail.com>
> Cc: <ross.lazarus at gmail.com>; <gregory_warnes at urmc.rochester.edu>; 
> <greg at warnes.net>
> Sent: Wednesday, May 06, 2009 2:53 PM
> Subject: Re: [R] Do you use R for data manipulation?
> 
> 
> 
> 
> I used R for my master thesis (with big effort, anyway) and now I find 
> difficult to use R in my daily work, becasue it has really serious 
> problems with datasets of big dimension, both in the data manipulation 
> step and in the analysis step.
> 
> But I really would love to use it, as I like its transparence, compared 
> to other software.
> 
> Laura
> 
> ***********
> 
>> Date: Wed, 6 May 2009 06:42:45 -0700
>> From: jrkrideau at yahoo.ca
>> To: r-help at stat.math.ethz.ch; fjbuch at gmail.com
>> CC: ross.lazarus at gmail.com; gregory_warnes at urmc.rochester.edu; 
>> greg at warnes.net
>> Subject: Re: [R] Do you use R for data manipulation?
>>
>>
>>
>>
>> --- On Wed, 5/6/09, Farrel Buchinsky <fjbuch at gmail.com> wrote:
>>
>> > Is R an appropriate tool for data
>> > manipulation and data reshaping and data
>> > organizing? I think so but someone who recently joined our
>> > group thinks not.
>>
>> I only do small scale projects and am by no means a programmer. Isn't 
>> Perl something for earings?
>>
>> That said, I find R to be extremely useful at data manipulation and 
>> have used it exclusively in my last three projects.  The different 
>> data structures alone are worth their weight in gold, if for nothing 
>> else than making it harder to make stupid mistakes in coding.
>>
>> > The new recruit believes that python or another language is
>> > a far better tool for developing data manipulation scripts that can be
>> > then used by> several members of our research group. Her assessment is
>> > that R is useful> only when it comes to data analysis and working with
>> > statistical models.
>>
>> Any reason that she thinks this?  How well does she know R?  It is not 
>> exactly a language that one picks up in a week, especially if one is 
>> coming from using a stats package like SAS or SPSS. As an ex-SAS and 
>> SYSTAT user it took me weeks to just get comfortable with the power of 
>> subscripting and the ability to do all kinds of calculations "in-line".
>>
>> > So what do you think:
>> > 1)R is a phenomenally powerful and flexible tool and since you are 
>> going >  > to do analyses in R you might as well use it to read data 
>> in and merge
>> > it and reshape it to whatever you need.
>>
>> Definately. I am not a computer scientist or a statistician. I usually 
>> am working as a single contractor and normally with small datasets as 
>> part of a larger project.  R does what I want, usually very elegantly 
>> (albeit perhaps after a lot of headbanging and calls for help to the 
>> R-list) and it would be stupid for me to use more than one language 
>> when it is not needed.
>>
>> Another plus is that I can  easily leave my data analysis work and a 
>> working copy of R with the client.  He/she may have a problem seeing 
>> what I did but it is clearly readable & replicable by either the 
>> client or another consultant.
>>
>> > OR
>> > 2) Are you crazy? Nobody in their right mind uses R to pipe
>> > the data around their lab and assemble it for analysis.
>>
>> Well I don't work in a lab but why complicate things? If everyone is 
>> using the same tools then you have a good situation.  Others who do 
>> work in labs can address this point more cogently
>>
>> >From a personnel point of view do you expect everyone in the lab to 
>> be >proficient with R and, for example, Perl? What happens when/if you 
>> lose >your Perl expert(s)?  I've had occasions where I waited a week 
>> for data >simply because the division's MS Access "expert" was on 
>> holiday and the >only other "Access" person there only knew how to 
>> enter data and run the >monthly reports.  Anything more complicated 
>> required the "expert".
>>
>>
>>
>>
>>
>>       __________________________________________________________________
>> Make your browsing faster, safer, and easier with the new Internet 
>> Explorer® 8. Optimized for Yahoo! Get it Now for Free! at 
>> http://downloads.yahoo.com/ca/internetexplorer/
>>
>> ______________________________________________
>> 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.
> 
> _________________________________________________________________
> [[elided Hotmail spam]]
> 
> [[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.
>>
> 
> ______________________________________________
> 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.
> 


-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University