G'day Bill,
On 18/05/2011, at 10:36 AM, <bill.venab...@csiro.au> wrote:
> If you
>
> ?plot.stl
>
> you will see that that the second argument, set.pars, is a list of argument
> settings for par(), including a (variable) default setting for mfrow. I.e.
> plot.stl overrides your external setting (which will also override any
> layout() setting).
>
> It looks like to override it back, you may need to take charge yourself.
> Here is the gist of a partial way round it, perhaps, sort of. ...
>
> ################
> par(mar = c(0, 6, 0, 6), oma = c(6, 0, 4, 0),
> tck = -0.01, mfcol = c(4, 2))
>
> plot(stl.1, set.pars = list())
> plot(stl.2, set.pars = list())
> ################
Perfect - thank you very much - I had played with the set.pars, but it never
occurred to me to send it a null list (both times)
> Things might be a bit more flexible if you were to use xyplot.stl in the
> latticeExtra package. With lattice the operations of making the plot and
> displaying it are more cleanly separated.
I'll look into it - thanks for the tip.
cheers
Ben
> I am looking at monthly reports, and have three series of monthly data from
> 2007 to 2009. I would like to show the season decomposition of these two
> series side by side on the one device, however using plot doesn't seem to
> respect any use of layout(matrix(1:3, ncol=3)) or par(mfcol=c(1,3)).
>
> I'm guessing that this means that the plot(stl) perhaps uses them, but I
> can't find anywhere the / a plot.stl() - ie, I can't work out where the
> plot() call is going?
>
> I've attached a small example of data and some probably overly verbose code.
> There are only two sets of example data, not three as mentioned above.
>
> # load the data
> vals.1 <-
> structure(list(year = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L,
> 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
> 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("2007",
> "2008", "2009"), class = c("ordered", "factor")), month = structure(c(1L,
> 2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L,
> 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L, 5L, 6L, 7L,
> 8L, 9L, 10L, 11L, 12L), .Label = c("Jan", "Feb", "Mar", "Apr",
> "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"), class = c("ordered",
> "factor")), count = c(105, 100, 64, 44, 49, 65, 88, 90, 99, 92,
> 93, 88, 212, 146, 96, 131, 220, 143, 137, 138, 395, 362, 349,
> 222, 294, 268, 298, 310, 426, 348, 287, 101, 66, 112, 105, 4)), .Names =
> c("year",
> "month", "count"), row.names = c("1", "2", "3", "4", "5", "6",
> "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17",
> "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28",
> "29", "30", "31", "32", "33", "34", "35", "36"), class = "data.frame")
> vals.2 <-
> structure(list(year = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L,
> 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
> 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("2007",
> "2008", "2009"), class = c("ordered", "factor")), month = structure(c(1L,
> 2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L,
> 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L, 5L, 6L, 7L,
> 8L, 9L, 10L, 11L), .Label = c("Jan", "Feb", "Mar", "Apr", "May",
> "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"), class = c("ordered",
> "factor")), count = c(45, 34, 17, 6, 7, 16, 12, 11, 14, 17, 11,
> 20, 27, 12, 10, 14, 22, 23, 92, 144, 385, 274, 320, 252, 240,
> 146, 222, 142, 122, 117, 163, 89, 51, 89, 108)), .Names = c("year",
> "month", "count"), row.names = c("1", "2", "3", "4", "5", "6",
> "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17",
> "18", "19", "20", "21", "22", "23", "24", "25", "26", "27", "28",
> "29", "30", "31", "32", "33", "34", "35"), class = "data.frame")
>
> # create the time series
> series.1 <- ts(vals.1$count, start=c(2007,1), freq=12)
> series.2 <- ts(vals.2$count, start=c(2007,1), freq=12)
>
> # apply the seasonal decomposition
> stl.1 <- stl(series.1, "per", robust=TRUE)
> stl.2 <- stl(series.2, "per", robust=TRUE)
>
> # set up the device for side by side display
> layout(matrix(1:2, ncol=2))
>
> # little check
> layout.show(2)
>
> # plot the first series and second series
> plot(stl.1, labels = c('Count by month','Seasonal
> Component','Trend','Remainder'), main='Series.1 Decomposition')
> plot(stl.2, labels = c('Count by month','Seasonal
> Component','Trend','Remainder'), main='Series.2 Decomposition')
>
> # hmm, used whole device twice, try again
>
> par(mfcol=c(1,2))
>
> # and now the second
>
> plot(stl.1, labels = c('Count by month','Seasonal
> Component','Trend','Remainder'), main='Series.1 Decomposition')
> plot(stl.2, labels = c('Count by month','Seasonal
> Component','Trend','Remainder'), main='Series.2 Decomposition')
>
> # oh what about split.screen()
>
> split.screen(c(1,2))
> screen(1)
> # now plot
> plot(stl.1, labels = c('Count by month','Seasonal
> Component','Trend','Remainder'), main='Series.1 Decomposition')
>
> # something wrong with the plot, not seeing original series at the top.
> screen(2)
>
> # Error in par(split.screens[[n]]) : parameter "j" in "mfg" is out of range
--
Ben Madin
AusVet Animal Health Services
P.O. Box 5467
Broome WA 6725
Australia
AusVet's website: http://www.ausvet.com.au
______________________________________________
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.
