R help - Feb 2010 - xyplot, overlay two variables on one plot with group factors

All

I want to overlay two variables on the same plot following their appropriate
grouping.  I have attempted to use subscripting in panel with panel.xyplot,
but I can't get the grouping to follow into the panel...here is an
example...


dat<-data.frame(
y= log(1:10),
y2=10:19,
 x=1:10,
grp = as.factor(1)
)

dat2<-data.frame(
 y= log(10:19),
y2= 20:29,
x=1:10,
 grp = as.factor(c(2))
)

dat<-rbind(dat, dat2)

# Here I plot both response variables on y axis by the same x
xyplot(y2 ~ x, dat, groups=grp, type="l")
xyplot(y ~ x, dat, groups=grp, type="l")

# I want to overlay both in the same plot to compare
xyplot(y~ x, dat,
 groups = grp,
ylim = c(0,40),
panel=function(x,y,subscripts, groups,...){
 panel.xyplot(x, y, type="l")
panel.xyplot(dat$x[subscripts], dat$y2[subscripts],type="l")
 }
)


Thanks

-- 
Patrick Schmitz
Graduate Student
Plant Biology
1206 West Gregory Drive
RM 1500

	[[alternative HTML version deleted]]

Ok...I realized that if I don't wish to compare the groupings directly

i can plot the groups as separate panels, which is a much more simple
solution, but doesn't necessarily give the same effect.

xyplot(y+y2 ~ x | grp, dat)


On Sun, Feb 14, 2010 at 3:19 PM, Dennis Murphy <djmuser@gmail.com> wrote:
> Hi:
>
> Here's one approach: the idea is to essentially stack the responses in
both
> data sets, create a factor that identifies the stacked variables, merges
> the data
> together and then combines 'variable' and 'grp' from the
merged data frame
> into a single factor for use as the groups = element in xyplot. (Got all
> that?)
>
> library(lattice)
> library(reshape)
> # use the melt function of reshape to stack the responses and create an
> # identifier (variable) for them. The responses themselves are put into a
> # variable named 'value'. id.vars are kept separate, the others are
stacked
> # into a variable identifier called 'variable' and the vector of
values
> called
> # 'values'...
> Dat <- melt(dat, id.vars = c('x', 'grp'))
> Dat2 <- melt(dat2, id.vars = c('x', 'grp'))
> Dat3 <- rbind(Dat, Dat2)     # concatenate the two data frames
> Dat3$gv <- with(Dat3, paste(variable, grp, sep = ""))  #
combine factors
>
> xyplot(value ~ x, data = Dat3, groups = gv, type = "l")
>
> There are more elegant ways to create the grouping factor, which you may
> need if you intend to create a 'nice' legend, but at this level it
works.
>
> HTH,
> Dennis
>
> On Sun, Feb 14, 2010 at 12:44 PM, Pat Schmitz
<pschmitz@illinois.edu>wrote:
>
>> All
>>
>> I want to overlay two variables on the same plot following their
>> appropriate
>> grouping.  I have attempted to use subscripting in panel with
>> panel.xyplot,
>> but I can't get the grouping to follow into the panel...here is an
>> example...
>>
>>
>> dat<-data.frame(
>> y= log(1:10),
>> y2=10:19,
>>  x=1:10,
>> grp = as.factor(1)
>> )
>>
>> dat2<-data.frame(
>>  y= log(10:19),
>> y2= 20:29,
>> x=1:10,
>>  grp = as.factor(c(2))
>> )
>>
>> dat<-rbind(dat, dat2)
>>
>> # Here I plot both response variables on y axis by the same x
>> xyplot(y2 ~ x, dat, groups=grp, type="l")
>> xyplot(y ~ x, dat, groups=grp, type="l")
>>
>> # I want to overlay both in the same plot to compare
>> xyplot(y~ x, dat,
>>  groups = grp,
>> ylim = c(0,40),
>> panel=function(x,y,subscripts, groups,...){
>>  panel.xyplot(x, y, type="l")
>> panel.xyplot(dat$x[subscripts], dat$y2[subscripts],type="l")
>>  }
>> )
>>
>>
>> Thanks
>>
>> --
>> Patrick Schmitz
>> Graduate Student
>> Plant Biology
>> 1206 West Gregory Drive
>> RM 1500
>>
>>        [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help@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.
>>
>
>

-- 
Patrick Schmitz
Graduate Student
Plant Biology
1206 West Gregory Drive
RM 1500

	[[alternative HTML version deleted]]

On Feb 14, 2010, at 3:44 PM, Pat Schmitz wrote:
> All
>
> I want to overlay two variables on the same plot following their  
> appropriate
> grouping.  I have attempted to use subscripting in panel with  
> panel.xyplot,
> but I can't get the grouping to follow into the panel...here is an
> example...
>
>
> dat<-data.frame(
> y= log(1:10),
> y2=10:19,
> x=1:10,
> grp = as.factor(1)
> )
>
> dat2<-data.frame(
> y= log(10:19),
> y2= 20:29,
> x=1:10,
> grp = as.factor(c(2))
> )
>
> dat<-rbind(dat, dat2)
>
> # Here I plot both response variables on y axis by the same x
> xyplot(y2 ~ x, dat, groups=grp, type="l")
> xyplot(y ~ x, dat, groups=grp, type="l")
>
> # I want to overlay both in the same plot to compare
> xyplot(y~ x, dat,
> groups = grp,
> ylim = c(0,40),
> panel=function(x,y,subscripts, groups,...){
> panel.xyplot(x, y, type="l")
> panel.xyplot(dat$x[subscripts], dat$y2[subscripts],type="l")
> }
> )
>
My reading of the xyplot help page made me think that "subscripts" was
logical and not an indexing variable, although the opposite is  
suggested by panel.superpose's help page. (Just my continued confusion  
about how to use lattice functions properly.) It sounds as though you  
are trying to superimpose what would otherwise be separate panels. So  
I'm never quite sure that when I mess around with panel arguments that  
I'm doing it right,  but this produces what I think you are trying to  
do:

  xyplot(y~ x, dat,
  groups = grp,
  ylim = c(0,40),
  panel=function(x,y, groups,...){
       panel.xyplot(x, y, groups,  type="l", ...)
       panel.superpose(x, dat$y2, groups,  type="l", ...)
         })

My hypothesis after experimenting with trying to leave out the , ...  
argument is that it is necessary to accept the subscripts argument,  
even though you don't actually need to do anything with it. I think  
that is implied by the fact that "subscripts (with no default) appears  
before the ellipsis in the arguments list.
>
> Thanks
>
> -- 
> Patrick Schmitz
> Graduate Student
> Plant Biology
> 1206 West Gregory Drive
> RM 1500
>
> 	[[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.
David Winsemius, MD
Heritage Laboratories
West Hartford, CT