R help - Dec 2015 - Graphing a subset of data

I am trying to write a function that will graph a variable of a dataset or
a variable from a subset of the data. My function is as follows:

graphics<-function(dat, var, graph, varname, val, subset){
  if(subset==1) {
    data<-mySubset(dat=dat,varname=varname,val=val)
  }else if(subset!=1){
    data<-(dat)
  }
  if(graph==1) {
    with(data, boxplot(var), main = paste("Vertical box plot of",
myfunc(dat), "variable", myfunc(var), xlab = myfunc(var)))
  }else if(graph==2){
    with(data, boxplot(var), horizontal=TRUE, main = paste("Horizontal
box plot of", myfunc(dat), "variable", myfunc(var), xlab
myfunc(var)))
  }else if(graph==3){
    hist(var, main="Histogram of", myfunc(var))
  }}

I included "subset" as a parameter for the function, so that subset=1
would
tell the function to subset, and any other value will use the full dataset.
However, when I run the function with the following expression (bolts is
the dataset, SPEED1 is the variable, value=3)

graphics(bolts, bolts$SPEED1, graph=3, bolts$SPEED1, 3, 1)

I receive this error message: Error in eval(substitute(expr), data, enclos
= parent.frame()) : invalid 'envir' argument of type 'logical'

Does anyone know why this is, or if there is something wrong with my code?
Thanks!

-- 
*Alexandra Hua *
Yale University | MPH Candidate Class of 2016
Chronic Disease Epidemiology

	[[alternative HTML version deleted]]

Dear Alex (as you are signing yourself on StackOverflow);

It is considered poor manners to cross-post identical questions in multiple
venues.

http://stackoverflow.com/questions/34011669/graphing-function-based-on-subset-of-data

You should choose one or the other of SO and Rhelp. If you do not get a
satisfying answer in your first choice, you should wait an appropriate number of
hours before posting at the other venue. And when you do end up cross-posting,
your should stated where else the question was asked so that potential
respondents can check to see if you have already gotten an answer.

You should also read the the posting guide where it is clearly stated the rhelp
responders expect that you include a dataset built with R code for purposes of
illustration.
> On Nov 30, 2015, at 6:44 PM, Alexandra Hua <alexandra.hua at
yale.edu> wrote:
> 
> I am trying to write a function that will graph a variable of a dataset or
> a variable from a subset of the data. My function is as follows:
> 
> graphics<-function(dat, var, graph, varname, val, subset){
>  if(subset==1) {
>    data<-mySubset(dat=dat,varname=varname,val=val)
>  }else if(subset!=1){
>    data<-(dat)
>  }
>  if(graph==1) {
>    with(data, boxplot(var), main = paste("Vertical box plot of",
> myfunc(dat), "variable", myfunc(var), xlab = myfunc(var)))
>  }else if(graph==2){
>    with(data, boxplot(var), horizontal=TRUE, main = paste("Horizontal
> box plot of", myfunc(dat), "variable", myfunc(var), xlab
> myfunc(var)))
>  }else if(graph==3){
>    hist(var, main="Histogram of", myfunc(var))
>  }}
> 
Generally the use of `with` inside functions is ill-advised. Sometimes it
succeeds but nmany times it will fail with puzzling error messages.

> I included "subset" as a parameter for the function, so that
subset=1 would
> tell the function to subset, and any other value will use the full dataset.
> However, when I run the function with the following expression (bolts is
> the dataset, SPEED1 is the variable, value=3)
> 
> graphics(bolts, bolts$SPEED1, graph=3, bolts$SPEED1, 3, 1)
> 
> I receive this error message: Error in eval(substitute(expr), data, enclos
> = parent.frame()) : invalid 'envir' argument of type
'logical'
> 
> Does anyone know why this is, or if there is something wrong with my code?
> Thanks!
> 
> -- 
> *Alexandra Hua *
> Yale University | MPH Candidate Class of 2016
> Chronic Disease Epidemiology
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.
David Winsemius
Alameda, CA, USA

Hi Alexandra,
I think you are going about this in an excessively difficult way. Here is a
rough example:

graphit<-function(x,var,type,subset=NA,...){
 if(!is.na(subset[1])) x<-subset(x,subset)
 do.call(type,list(x=x[[var]],...))
}
# assume that the data are measurements of penetration by crossbow bolts at
specified speeds
bolts<-data.frame(SPEED1=rep(1:3,each=30),
 penetration=c(runif(30,10,15),runif(30,15,20),runif(30,20,25)))
# note the lazy way of adding arguments to the function call
graphit(bolts,"SPEED1",boxplot,main="Boxplot of SPEED1")
graphit(bolts,"penetration",boxplot,subset=bolts$SPEED1==3,
 main="Boxplot of SPEED1 == 3")
graphit(bolts,"SPEED1",hist,main="Histogram of SPEED1")

Jim

	[[alternative HTML version deleted]]