This worked for me. It is btw. quite confusing to name your y-variable x.
I think part of the problem arised from the date format.
xyplot(x + max.x ~ date, data = my.newdf, ylab = "x",
panel = function(x, y, x2, ...){
panel.xyplot(x, y, type = "l")
panel.loess(as.numeric(my.newdf$date), my.newdf$max.x, lty = 2)
#panel.xyplot(x, y2, type = "l")
})
2015-04-29 13:09 GMT+02:00 Naresh Gurbuxani <naresh_gurbux...@hotmail.com>:
> I want to plot multiple variables in xyplot, but plot loess trend for only
> one of these variables. My problem is that the last command below does not
> give the desired result.
> Any help will be gratefully received.
> Thanks,Naresh
> my.df <- data.frame(date = as.numeric(as.Date("2015-01-01")) + 0:49, x =
> rnorm(50))
> my.df$date <- as.Date(my.df$date, origin = as.Date("1970-01-01"))
>
>
> library(zoo)
> x <- zoo(my.df[,"x"], my.df[,"date"])
> max.x <- rollapply(x, 10, max, align = "right")
> x <- merge(x, max.x)
> my.newdf <- data.frame(x)
> my.newdf$date <- as.Date(row.names(my.newdf))
>
>
> library(lattice)# This works as expected
> xyplot(x + max.x ~ date, data = my.newdf, type = "l",
> auto.key = list(columns = 2, points = FALSE, lines = TRUE), ylab = "x")
> # This does not work
> xyplot(x ~ date, data = my.newdf, y2 = max.x, ylab = "x",
> panel = function(x, y, x2, ...){
> panel.xyplot(x, y, type = "l")
> panel.loess(x, y, lty = 2)
> panel.xyplot(x, y2, type = "l")
> })
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@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.
>
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
