[R] Subsetting data frames without a loop

[MCCANN, JACKSON]

Hi,

I've been working on data with a similar issue and ended up with the
following code. Given the test data in your post it seems to work.

	maxvals <- tapply(v,factor(stn),max)
	subset(my.df, v >= maxvals[stn])

However I've only been using R for a couple of weeks so comments on the
appropriateness of this solution or better/faster code please!

Regards Jackson

-----Original Message-----
From: Uwe Ligges [mailto:ligges at statistik.uni-dortmund.de]
Sent: 16 January 2002 08:44
To: Michael_Nielsen/Syd/Synergy.SYNERGY at synergy.com.au
Cc: r-help at stat.math.ethz.ch
Subject: Re: [R] Subsetting data frames without a loop



Michael_Nielsen/Syd/Synergy.SYNERGY at synergy.com.au wrote:
> 
> I KNOW this should be easy, but I'm stuck.
> 
> My data frame consists of multiple observations from each of a number of
> stations, and what I would like to do is create another data frame that
> contains all the variables of the first, but only rows where a certain
> variable is at its maximum for the station.
> 
> So, for example:
> 
> > my.df
>    stn obs           v
> 1    1   1  0.26400396
> 2    2   1 -0.79194397
> 3    3   1  0.11924528
> 4    4   1  0.42596859
> 5    5   1 -0.50528235
> 6    1   2 -1.57524853
> 7    2   2  0.17762482
> 8    3   2 -0.83013770
> 9    4   2 -0.53203400
> 10   5   2 -2.71397275
> 11   1   3  0.26902053
> 12   2   3  2.01147908
> 13   3   3  0.73301643
> 14   4   3 -0.67333384
> 15   5   3 -1.36219773
> 16   1   4 -2.20342109
> 17   2   4  0.18941702
> 18   3   4  0.51492032
> 19   4   4  0.03597370
> 20   5   4 -1.43502366
> 21   1   5 -1.34589392
> 22   2   5  1.00389195
> 23   3   5 -0.21233041
> 24   4   5 -1.35141044
> 25   5   5 -0.02052348
> 
> > tapply(v,factor(stn),max)
>           1           2           3           4           5
>  0.26902053  2.01147908  0.73301643  0.42596859 -0.02052348
> 
> so my new data frame should contain (possibly multiple rows per station)
> 
>    stn obs           v
> 1    1   3  0.26902053
> 2    2   3  2.01147908
> 3    3   3  0.73301643
> 4    4   1  0.42596859
> 5    5   5 -0.02052348


As a first idea:

  my.df[tapply(v,factor(stn), function(x) which(v==max(x))),]


Uwe
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.
-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._.
_._

************
This email and any accompanying documents are intended only for the named recipient, are confidential and may be privileged. If you are not the intended recipient please notify us immediately by mailto:admin at britannic.co.uk and you must not copy, disclose or otherwise use this message. Unauthorised use is strictly prohibited and may be unlawful. The content of this email represents the view of the individual and not the company. The company reserves the right to monitor the content of all emails in accordance with lawful business practice.

Whilst attachments are virus checked before transmission, Britannic Assurance plc does not accept any liability in respect of any virus which is not detected.

Britannic Assurance plc, No.3002 is registered in England and maintains its registered office at 1 Wythall Green Way, Wythall, Birmingham B47 6WG.
Telephone: 0870 887 0001
Fax: 0870 887 0002
Website: www.britannicassurance.com

Britannic Assurance plc, Britannic Unit Linked Assurance Limited, Britannic ISA Managers Limited and Britannic Unit Trust Managers Limited are regulated by the Financial Services Authority. Each of these companies is a member of the Britannic marketing group which only advises on and sells its own life assurance, pension, unit trust and ISA products.
************
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._