What a simple way to do what I want to do! Thanks.
But if there is missing data in variable "y", then the averaged lines is broken
where the missing data is present. For example:
dat$y[c(10,185,200,400,450)]<-NA
then using type='a' option will result in broken lines.
xyplot(y~day|sex, groups=trt, data=dat, type=c('p','g', 'a'))
I was able to solve this using panel.superpose() and panel.average(). But it's
too much syntax. Is there a simpler way to do this?
xyplot(y~day|sex, groups=trt, data=dat, type=c('p','g'), panel=function(...)
{panel.xyplot(...);
panel.superpose(..., col=c(1:3), panel.groups=function(...)
panel.average(..., fun=function(x) mean(x,na.rm=T),horizontal=F))})
Thanks
John
________________________________
From: Dennis Murphy <djmu...@gmail.com>
Cc: r-help@r-project.org
Sent: Fri, August 20, 2010 3:25:06 PM
Subject: Re: [R] xyplot plot with groups
Hi:
Like this?
xyplot(y~day|sex, groups=trt, data=dat, type=c('p','g', 'a'))
The 'a' stands for average line, or 'connect the averages at different x
values'.
By using the groups = trt argument, you get one line per treatment group in
each panel.
A complete list of valid types is given on p. 75 of the Lattice book (Sarkar,
2008).
Related types are:
'l' : join points by lines
'r': least squares line
'smooth': loess smooth
HTH,
Dennis
Hi, I am a beginner of xyplot() (or lattice package). On one hand, I immediately
>realized it's a very powerful utility. On the other hand, there are too many
>things for me to learn. Still haven't figure out a generalization of the syntax
>and usage under many different circumstances.
>
>Let me give an example dataset:
>
>
>xxx<-cbind(expand.grid(day=0:7,trt=c('A','B','C'),sex=c('M','F')),subject=rep(1:60,each=8))
>
>
>xxx<-xxx[sort.list(xxx$trt),]
>y<-c(rnorm(160,15),rnorm(160,10),rnorm(160,5))
>
>dat<-cbind(y=y,xxx)
>
>str(dat)
>'data.frame': 480 obs. of 5 variables:
> $ y : num 13.4 15.3 14.4 16.2 15.1 ...
> $ day : int 0 1 2 3 4 5 6 7 0 1 ...
> $ trt : Factor w/ 3 levels "A","B","C": 1 1 1 1 1 1 1 1 1 1 ...
> $ sex : Factor w/ 2 levels "M","F": 1 1 1 1 1 1 1 1 2 2 ...
> $ subject: int 1 1 1 1 1 1 1 1 4 4 ...
>
>So this is a typical longitudinal dataset: each subject has measurements for 8
>days (0 to 7), there are 3 treatments (A, B and C), and sex factor (M and F). I
>would like to have 2 plots, one for each sex. Within each plot, I want to plot
3
>averaged lines (average of y for each day), one for each treatment group. How
>can I do this?
>
>I tried the following using panel.average(), but it seems that the groups
>information was not passed automatically to panel.average().
>
>
>xyplot(y~day|sex, groups=trt, data=dat, type=c('p','g'), panel=function(...)
>{panel.xyplot(...); panel.average(..., fun=function(x)
>mean(x,na.rm=T),horizontal=F,col=1:3)})
>
>Appreciate if anyone can help.
>
>John
>
>______________________________________________
>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.
>
[[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.
