I've hit a problem in ggplot2 which I can trace back to cut.Date ,
which is either a bug or (??) ggplot2 trying to do something it
shouldn't with cut.Date (although its use of cut.Date seems OK).
Apparently any (?) call of the form
cut(as.Date("2008-07-07"),"weeks")
where the date *begins the week*, gives the error
Error in 1:(1 + max(which(breaks < maxx))) :
result would be too long a vector
In addition: Warning message:
In max(which(breaks < maxx)) :
no non-missing arguments to max; returning -Inf
In particular, try:
d <- seq.Date(as.Date("2008-07-05"),as.Date("2008-07-08"),by="days")
weekdays(d)
cut(d[1],"weeks") ## OK
cut(d[2],"weeks") ## OK
cut(d[3],"weeks") ## Monday -- breaks
cut(d[4],"weeks") ## OK
cut(d[1],"weeks",start.on.monday=FALSE) ## OK
cut(d[2],"weeks",start.on.monday=FALSE) ## Sunday -- breaks
cut(d[3],"weeks",start.on.monday=FALSE) ## OK
cut(d[4],"weeks",start.on.monday=FALSE) ## OK
The particular line that fails within cut.Date is
breaks <- breaks[1:(1 + max(which(breaks < maxx)))]
but I haven't figured out enough about what's going on
to be able to suggest a fix ...
cheers
Ben Bolker
session info:
> sessionInfo()
R version 2.7.2 (2008-08-25)
i486-pc-linux-gnu
locale:
LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=C;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C
attached base packages:
[1] splines grid stats graphics grDevices utils datasets
[8] methods base
other attached packages:
[1] ggplot2_0.7 MASS_7.2-44 RColorBrewer_1.0-2 proto_0.3-8
[5] reshape_0.8.1 plyr_0.1
Here's the context in which I encountered the problem:
library(ggplot2)
tdat <- data.frame(date=seq.Date(as.Date("2008-07-07"),
as.Date("2008-10-10"),by="day"),
x=1:96)
q2 = qplot(date,x,data=tdat,geom="smooth")
print(q2+geom_point())
I can track the problem down to a call like
floor_date(as.Date("2008-07-07"),"weeks")
and digging further in, I can narrow it down to
the issue above.
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
