R help - Feb 2008 - Subsetting a data.frame degenerates at one column?

Greetings.

At the moment, I'm applying R to some AIX 'nmon' output, trying to
get
a handle on some disk performance metrics.  In case anyone's
interested:

http://docs.osg.ufl.edu/tsm/pdf/

some of them are more edifying than others. (ahem)

I'm trying to develop a somewhat general framework for plotting these
measures, in the hopes that it's of some use to people other than me.
One obstacle I encounter is that, when I select one column out of a
data.frame, the result is no longer a data.frame.  So, say I've got,
in data frame 'input'

      disk1 disk2 disk3 disk4
T0000     0     1     0     4
T0001     0     1     0     5
T0002     0     1     0     5
T0003     0     2     0     4
T0004     0     2     0     3
T0005     0     1     0     3
T0006     0     0     0     3

and somewhere I've noted a list

targets <- c('disk2','disk3')

I can say 

input[,targets]
      disk2 disk3
T0000     1     0
T0001     1     0
T0002     1     0
T0003     2     0
T0004     2     0
T0005     1     0
T0006     0     0

but if 

targets <- c('disk2')
input[,targets]
[1] 1 1 1 2 2 1 0

Ick.

I've been reading through the indexing and data.frame docs, and remain
unsatisfied so far.  Where is my thinking going wrong? 



- Allen S. Rout

try:

input[,targets, drop=FALSE]

see:

?"["

for an explanation.


On 2/8/08, Allen S. Rout <asr at ufl.edu> wrote:
>
> Greetings.
>
> At the moment, I'm applying R to some AIX 'nmon' output, trying
to get
> a handle on some disk performance metrics.  In case anyone's
> interested:
>
> http://docs.osg.ufl.edu/tsm/pdf/
>
> some of them are more edifying than others. (ahem)
>
> I'm trying to develop a somewhat general framework for plotting these
> measures, in the hopes that it's of some use to people other than me.
> One obstacle I encounter is that, when I select one column out of a
> data.frame, the result is no longer a data.frame.  So, say I've got,
> in data frame 'input'
>
>      disk1 disk2 disk3 disk4
> T0000     0     1     0     4
> T0001     0     1     0     5
> T0002     0     1     0     5
> T0003     0     2     0     4
> T0004     0     2     0     3
> T0005     0     1     0     3
> T0006     0     0     0     3
>
> and somewhere I've noted a list
>
> targets <- c('disk2','disk3')
>
> I can say
>
> input[,targets]
>      disk2 disk3
> T0000     1     0
> T0001     1     0
> T0002     1     0
> T0003     2     0
> T0004     2     0
> T0005     1     0
> T0006     0     0
>
> but if
>
> targets <- c('disk2')
> input[,targets]
> [1] 1 1 1 2 2 1 0
>
> Ick.
>
> I've been reading through the indexing and data.frame docs, and remain
> unsatisfied so far.  Where is my thinking going wrong?
>
>
>
> - Allen S. Rout
>
> ______________________________________________
> 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.
>

-- 
Jim Holtman
Cincinnati, OH
+1 513 646 9390

What is the problem you are trying to solve?

I think I understand this.  Try
str(input) and you will see that input$disk2 in a
vector.  You can force a data.frame using 
data.frame(input(,targets))


--- "Allen S. Rout" <asr at ufl.edu> wrote:
> 
> Greetings.
> 
> At the moment, I'm applying R to some AIX 'nmon'
> output, trying to get
> a handle on some disk performance metrics.  In case
> anyone's
> interested:
> 
> http://docs.osg.ufl.edu/tsm/pdf/
> 
> some of them are more edifying than others. (ahem)
> 
> I'm trying to develop a somewhat general framework
> for plotting these
> measures, in the hopes that it's of some use to
> people other than me.
> One obstacle I encounter is that, when I select one
> column out of a
> data.frame, the result is no longer a data.frame. 
> So, say I've got,
> in data frame 'input'
> 
>       disk1 disk2 disk3 disk4
> T0000     0     1     0     4
> T0001     0     1     0     5
> T0002     0     1     0     5
> T0003     0     2     0     4
> T0004     0     2     0     3
> T0005     0     1     0     3
> T0006     0     0     0     3
> 
> and somewhere I've noted a list
> 
> targets <- c('disk2','disk3')
> 
> I can say 
> 
> input[,targets]
>       disk2 disk3
> T0000     1     0
> T0001     1     0
> T0002     1     0
> T0003     2     0
> T0004     2     0
> T0005     1     0
> T0006     0     0
> 
> but if 
> 
> targets <- c('disk2')
> input[,targets]
> [1] 1 1 1 2 2 1 0
> 
> Ick.
> 
> I've been reading through the indexing and
> data.frame docs, and remain
> unsatisfied so far.  Where is my thinking going
> wrong? 
> 
> 
> 
> - Allen S. Rout
> 
> ______________________________________________
> 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.
>

Try this also:

input[targets]

On 08/02/2008, Allen S. Rout <asr at ufl.edu>
wrote:
>
> Greetings.
>
> At the moment, I'm applying R to some AIX 'nmon' output, trying
to get
> a handle on some disk performance metrics.  In case anyone's
> interested:
>
> http://docs.osg.ufl.edu/tsm/pdf/
>
> some of them are more edifying than others. (ahem)
>
> I'm trying to develop a somewhat general framework for plotting these
> measures, in the hopes that it's of some use to people other than me.
> One obstacle I encounter is that, when I select one column out of a
> data.frame, the result is no longer a data.frame.  So, say I've got,
> in data frame 'input'
>
>       disk1 disk2 disk3 disk4
> T0000     0     1     0     4
> T0001     0     1     0     5
> T0002     0     1     0     5
> T0003     0     2     0     4
> T0004     0     2     0     3
> T0005     0     1     0     3
> T0006     0     0     0     3
>
> and somewhere I've noted a list
>
> targets <- c('disk2','disk3')
>
> I can say
>
> input[,targets]
>       disk2 disk3
> T0000     1     0
> T0001     1     0
> T0002     1     0
> T0003     2     0
> T0004     2     0
> T0005     1     0
> T0006     0     0
>
> but if
>
> targets <- c('disk2')
> input[,targets]
> [1] 1 1 1 2 2 1 0
>
> Ick.
>
> I've been reading through the indexing and data.frame docs, and remain
> unsatisfied so far.  Where is my thinking going wrong?
>
>
>
> - Allen S. Rout
>
> ______________________________________________
> 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.
>

-- 
Henrique Dallazuanna
Curitiba-Paran?-Brasil
25? 25' 40" S 49? 16' 22" O

"jim holtman" <jholtman at gmail.com> writes:
> try:
> input[,targets, drop=FALSE]
> see:
> ?"["
> for an explanation.

Thanks, you who responded; this was exactly helpful, and a good
reference to the part of the FM I was missing.  To unpack (and
demonstrate some comprehension gained.. ;) the subsetting operations
on data frames, by default, use the most basic data type capable of
representing the answer.  

Either the drop=FALSE or the inputs[targets] solution give me the
result I had in mind.  I mildly prefer the [targets] statement from a
visual perspective.





- Allen S. Rout