Dear David,
I would like to use summarize(Hmisc) and print.xtable(xtable) in a
single Sweave document, but a conflict with the 'label' function
prohibits this at the moment!
Would you be able to correct the conflicting code? I will gladly test
the new package!
I have tried latex(Hmisc) to
One possible solution without renaming the functions is to add namespace to
either xtable or Hmisc. Given the size of Hmisc, it probably would be much
easier to do that with xtable.
With namespace in xtable, you can do xtable:::label() to refer to the
label() in xtable specifically.
Andy
Even without a namespace one could explicitly reference the label
in xtable via:
xtable.label - get(label, package:xtable)
On 5/16/05, Liaw, Andy [EMAIL PROTECTED] wrote:
One possible solution without renaming the functions is to add namespace to
either xtable or Hmisc. Given the size of
Hi Andy and Gabor,
Thanks for your help so far! I am discovering another R dimension.
Trying to put my head around all thisthe conflict actually exposes
itself when calling summarize(Hmisc). Summarize(Hmisc) calls label
internally, so I can not call it explicitly. Simply calling
Do you need label in both xtable and Hmisc? If you only need
it in Hmisc and not in xtable then just be sure you have loaded
xtable first and Hmisc second. This:
search()
will give you the search path. It will find the first one on the
search path so if Hmisc is before xtable (which would
You need to add the namespace to the source package, by adding a NAMESPACE
file. There's an R News article by Prof. Tierney on how to do this. Also
see the `Writing R Extensions' manual. You should get the package
maintainer to do that, as that constitute a change in the package source
code.
I tried to follow your suggestions, but without success:
Error: couldn't find function xtable.mylabel-
... resulting from the code below.
Any suggestions?
Thanks,
Sander.
library(xtable)
xtable.mylabel - get(label, package:xtable)
library(Hmisc) # provides summarize
set.seed(1)
temperature -
Sander,
Thanks for pointing out the conflict between Hmisc and xtable. I am not
sure I have a good solution.
My understand of the namespace solution is that packages can specify
which variables to export for use by the package users. The label
function is not an internal function, rather one
Hi David,
Thanks for creating and supporting xtable. Glad I can contribute by
pointing out problems!
An earlier response from Frank Harrell Jr. suggests that the
responsibility to resolve the conflict lies primarily with you, as Hmisc
'was there first'.
Not sure how disputes over package
For fixing the problem you have, you need:
xtable.mylabel- - get(label-.xtable, package:xtable)
I.e., you need the replacement function.
One possibility of resolving the conflict, as I communicated with Frank, is
to make label() and label-() in Hmisc the S3 default methods, as xtable
defines
That seems like a good solution.
The one problem is that both packages would need their own
label generic functions if they are to operate independently
of each other so if they were both loaded the last one loaded
would give a warning that the label generic from the first loaded
package is
Using label as an lvalue (i.e. on the left hand side of the assignment)
causes it to refer to a different function, not label itself.
Any any rate, looking at your example,
it seems that you don't actually need to use Hmisc and xtable at the same time
so just make sure that whichever you want
Sander Oom wrote:
Dear R users,
The Sweave code below runs fine, as it is. However, an error occurs when
the line 'library(xtable)' is uncommented:
Error: chunk 1
Error in label-(`*tmp*`, value = month) :
no applicable method for label-
Is anybody aware of this and knows a workaround?
Dear Frank,
I have a Sweave document in which I export anova (aov) tables to Latex
and calculate some summary statistics with summarize{Hmisc} for a graph
(as in the example below).
I currently use the following code for the aov tables:
results=tex=
tmp - datGrassHC[datGrassHC$Loc 0
14 matches
Mail list logo
