[R] "subset" or "condition" as argument to a function

[William Dunlap]

> -----Original Message-----
> From: r-help-bounces at r-project.org 
> [mailto:r-help-bounces at r-project.org] On Behalf Of Santosh
> Sent: Tuesday, December 01, 2009 7:39 AM
> To: r-help at r-project.org
> Subject: Re: [R] "subset" or "condition" as argument to a function
> 
> Dear R gurus..
> I had tried out some suggestions sent to me privately..and 
> unfortunately,
> they did not work..
> 
> To use a condiition in a subset, the associated dataframe needs to be
> attached and detached, which I found cumbersome to use if 
> using more than 1
> dataframe (with different dimensions) with same condition. I 
> would highly
> appreciate your suggestions to overcome this problem.. Please 
> see my example
> of usage I am trying to implement. Please note that the 
> "cond' takes in the
> character instead of logical values...
> 
> cond1 <- 'group==1&sales>200'
> cond2 <- 'group==2&sales>200'
> cond3 <- 'group==3&sales>200'
> 
> d1 <- subset(dat,subset=cond1)
> plot(y~x,data=dat,subset=cond1,type='b')

The subset argument to plot (and many similar functions)
must be given as a literal expression, not a string that
could be parsed into an expression nor the name of an object
containing an expression nor a function call that evaluates
to an expression.  This design is handy for interactive
use but painful in programatic use.  One way to deal with
it is to use do.call, which evaluates all the arguments to
the function and then calls the function with the arguments
given literally.  Replace the above plot call with
  do.call("plot", list(y~x,data=dat,subset=parse(text=cond1),type='b'))
and see if you get what you want.

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com 

> lines(y~x,data=dat,subset=cond2)
> points(y~x,data=dat,subset=paste(cond1,cond2,sep='&'),col='blue')
> points(y~x,data=d1,subset=cond3,col='red')
> 
> If I try the above, I get the following error message:
> *Error in subset.data.frame(dat, cond) : 'subset' must 
> evaluate to logical*
> 
> If you could also suggest references implementing similar 
> code, I would
> highly appreciate it.
> Thanks so much,
> 
> Santosh
> 
> On Mon, Nov 23, 2009 at 6:38 PM, Santosh 
> <santosh2005 at gmail.com> wrote:
> 
> > Thanks... I would try it out..
> >  -santosh
> >
> >
> > On Sat, Nov 21, 2009 at 12:04 AM, Bernardo Rangel Tura <
> > tura at centroin.com.br> wrote:
> >
> >> On Fri, 2009-11-20 at 17:40 -0800, Santosh wrote:
> >> > Dear Rxperts!
> >> >
> >> > I was wondering if it is possible to write a function 
> which can take in
> >> > argument of a subset or condition.. Of course, I am aware of the
> >> alternate
> >> > methods like coplot, par.plot, xyplot etc... I am specifically
> >> interested in
> >> > using conditions/subsets with "plot"..
> >> >
> >> > A simple fragmented example is shown here..
> >> >
> >> > pltit <- function(y,x,dat,dat1,dat2,sbst) {
> >> > plot(y~x, data=dat, subset=sbst)
> >> > lines(y~x,data=dat1, subset=subst)
> >> > points(y~x,data=dat2,subset=subst)
> >> > }
> >> >
> >> > pltit(profit,weeks,dat=zone1, sbst='group==1&sales>200')
> >> >
> >> > Could you also suggest pointers/references/examples on 
> efficient ways to
> >> > plot simulated data overlaid with observed data?
> >> >
> >> > Have a good weekend!
> >>
> >>
> >>
> >
> 
> 	[[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.
>