R help - Aug 2008 - Obtaining the first /or last record of a subject in a longitudinal study

Dear R users,

I was wondering if anyone knows how to obtain(subset) the first and/or the
last record of a subject in a longitudinal setup.

Normally in SAS one uses first.variable1 and last.variable1. So my question
is that is there an R way of doing this.

Regards,

-- 
Luwis Diya, Phd student (Biostatistics),
Biostatistical Center,
School Of Public Health,
Catholic University of Leuven,
U.Z. St Raphael,
Kapucijnenvoer 35,
B-3000 Leuven,
Belgium,

Cell: +32(0)497 22 94 83
Phone:+32(0)16 32 68 86 [Office]
Phone:+32(0)16 32 98 76 [Home]
Fax: +32(0)16 33 70 15

Email: luwis.diya@med.kuleuven.be
http://med.kuleuven.be/biostat/staff/LD.htm

	[[alternative HTML version deleted]]

On 8/7/2008 6:21 AM, Luwis Tapiwa Diya wrote:
> Dear R users,
> 
> I was wondering if anyone knows how to obtain(subset) the first and/or the
> last record of a subject in a longitudinal setup.
> 
> Normally in SAS one uses first.variable1 and last.variable1. So my question
> is that is there an R way of doing this.
> 
> Regards,
   How about a combination of aggregate() and head() or tail()?

 > library(nlme)

 > aggregate(Oxboys[,-1], list(Subject = Oxboys$Subject), head, n=1)
    Subject age height Occasion
1       10  -1  126.2        1
2       26  -1  132.2        1
3       25  -1  135.5        1
4        9  -1  132.7        1
5        2  -1  136.9        1
6        6  -1  142.4        1
7        7  -1  141.3        1
8       17  -1  134.9        1
9       16  -1  142.8        1
10      15  -1  137.5        1
11       8  -1  141.7        1
12      20  -1  146.5        1
13       1  -1  140.5        1
14      18  -1  145.5        1
15       5  -1  145.8        1
16      23  -1  144.5        1
17      11  -1  142.5        1
18      21  -1  143.9        1
19       3  -1  150.0        1
20      24  -1  147.8        1
21      22  -1  147.4        1
22      12  -1  149.9        1
23      13  -1  148.9        1
24      14  -1  151.6        1
25      19  -1  156.9        1
26       4  -1  155.7        1

 > aggregate(Oxboys[,-1], list(Subject = Oxboys$Subject), tail, n=1)

..

-- 
Chuck Cleland, Ph.D.
NDRI, Inc. (www.ndri.org)
71 West 23rd Street, 8th floor
New York, NY 10010
tel: (212) 845-4495 (Tu, Th)
tel: (732) 512-0171 (M, W, F)
fax: (917) 438-0894

Had the same problem recently. I used a loop. Assume your data frame is
called "data" and your units of observation are "Subject". 

first.round.index=NULL
final.round.index=NULL
for(i in unique(Subject)){
    first.round.index[i]=min(which(Subject==i))
    final.round.index[i]=max(which(Subject==i))
    }

first.round.index=first.round.index[is.na(first.round.index)==F]
final.round.index=final.round.index[is.na(final.round.index)==F]

##This gives you the row numbers for first, respectively last, observation
for each subject

##Then you can select the data rows of your data according to first- or
final.row.index, e.g.:

data.first.round=data[first.round.index, ]
data.final.round=data[final.round.index, ]


-------------------------
cuncta stricte discussurus
-------------------------

-----Urspr?ngliche Nachricht-----
Von: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] Im
Auftrag von Luwis Tapiwa Diya
Gesendet: Thursday, August 07, 2008 6:22 AM
An: R-help at r-project.org
Betreff: [R] Obtaining the first /or last record of a subject in
alongitudinal study

Dear R users,

I was wondering if anyone knows how to obtain(subset) the first and/or the
last record of a subject in a longitudinal setup.

Normally in SAS one uses first.variable1 and last.variable1. So my question
is that is there an R way of doing this.

Regards,

--
Luwis Diya, Phd student (Biostatistics), Biostatistical Center, School Of
Public Health, Catholic University of Leuven, U.Z. St Raphael,
Kapucijnenvoer 35, B-3000 Leuven, Belgium,

Cell: +32(0)497 22 94 83
Phone:+32(0)16 32 68 86 [Office]
Phone:+32(0)16 32 98 76 [Home]
Fax: +32(0)16 33 70 15

Email: luwis.diya at med.kuleuven.be
http://med.kuleuven.be/biostat/staff/LD.htm

	[[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.

A very fast way using tapply may be found in section 4.3.8 of 
http://biostat.mc.vanderbilt.edu/twiki/pub/Main/RS/sintro.pdf

Frank


Daniel Malter wrote:
> Had the same problem recently. I used a loop. Assume your data frame is
> called "data" and your units of observation are
"Subject".
> 
> first.round.index=NULL
> final.round.index=NULL
> for(i in unique(Subject)){
>     first.round.index[i]=min(which(Subject==i))
>     final.round.index[i]=max(which(Subject==i))
>     }
> 
> first.round.index=first.round.index[is.na(first.round.index)==F]
> final.round.index=final.round.index[is.na(final.round.index)==F]
> 
> ##This gives you the row numbers for first, respectively last, observation
> for each subject
> 
> ##Then you can select the data rows of your data according to first- or
> final.row.index, e.g.:
> 
> data.first.round=data[first.round.index, ]
> data.final.round=data[final.round.index, ]
> 
> 
> -------------------------
> cuncta stricte discussurus
> -------------------------
> 
> -----Urspr?ngliche Nachricht-----
> Von: r-help-bounces at r-project.org [mailto:r-help-bounces at
r-project.org] Im
> Auftrag von Luwis Tapiwa Diya
> Gesendet: Thursday, August 07, 2008 6:22 AM
> An: R-help at r-project.org
> Betreff: [R] Obtaining the first /or last record of a subject in
> alongitudinal study
> 
> Dear R users,
> 
> I was wondering if anyone knows how to obtain(subset) the first and/or the
> last record of a subject in a longitudinal setup.
> 
> Normally in SAS one uses first.variable1 and last.variable1. So my question
> is that is there an R way of doing this.
> 
> Regards,
> 
> --
> Luwis Diya, Phd student (Biostatistics), Biostatistical Center, School Of
> Public Health, Catholic University of Leuven, U.Z. St Raphael,
> Kapucijnenvoer 35, B-3000 Leuven, Belgium,
> 
> Cell: +32(0)497 22 94 83
> Phone:+32(0)16 32 68 86 [Office]
> Phone:+32(0)16 32 98 76 [Home]
> Fax: +32(0)16 33 70 15
> 
> Email: luwis.diya at med.kuleuven.be
> http://med.kuleuven.be/biostat/staff/LD.htm
-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University

It would have helped to see an example to clarify the input and
output to the problem but !duplicated(x) will be TRUE for the first
of each set of equal values in x and !duplicated(x, fromLast = TRUE)
for the last.  See ?duplicated for details.

On Thu, Aug 7, 2008 at 6:21 AM, Luwis Tapiwa Diya <siwulayid at gmail.com>
wrote:
> Dear R users,
>
> I was wondering if anyone knows how to obtain(subset) the first and/or the
> last record of a subject in a longitudinal setup.
>
> Normally in SAS one uses first.variable1 and last.variable1. So my question
> is that is there an R way of doing this.
>
> Regards,
>
> --
> Luwis Diya, Phd student (Biostatistics),
> Biostatistical Center,
> School Of Public Health,
> Catholic University of Leuven,
> U.Z. St Raphael,
> Kapucijnenvoer 35,
> B-3000 Leuven,
> Belgium,
>
> Cell: +32(0)497 22 94 83
> Phone:+32(0)16 32 68 86 [Office]
> Phone:+32(0)16 32 98 76 [Home]
> Fax: +32(0)16 33 70 15
>
> Email: luwis.diya at med.kuleuven.be
> http://med.kuleuven.be/biostat/staff/LD.htm
>
>        [[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.
>