[R-pkg-devel] Problems with :::

[Berry Boessenkool]

>     if (!inherits(x, "ftable")) 
>    if(class(x) == "ftable")  ## really really bad


Thank you Martin, for pointing that out!
I was not aware of it yet and changed all occurences in my package and added in note in my R course material.

(This message is mainly an encouragement that some people actually read the mails and act on it, even if not usually replying).

Regards,
Berry


> From: maechler at stat.math.ethz.ch
> Date: Mon, 1 Feb 2016 09:49:26 +0100
> To: kirill.mueller at ivt.baug.ethz.ch
> CC: r-package-devel at r-project.org
> Subject: Re: [R-pkg-devel] Problems with :::
> 
> >>>>> Kirill Müller <kirill.mueller at ivt.baug.ethz.ch>
> >>>>>     on Thu, 28 Jan 2016 17:42:12 +0100 writes:
> 
>     > On 28.01.2016 14:20, Duncan Murdoch wrote:
>     >> If the class is just "mytable", then you'll have to
>     >> change it to "ftable" before calling stats::format().
>     > Side note: The following will fail, because format.ftable
>     > double-checks the class of its argument:
> 
>     > format_ftable <- utils::getS3method("format", "ftable")
>     > format_ftable(mytable)
> 
>     > -Kirill
> 
> Well, I don't think it would fail as that does
> 
>     if (!inherits(x, "ftable")) 
>         stop("'x' must be an \"ftable\" object")
> 
> and the OP did mention that his objects were flat tables.
> 
> Note that the above is *good style* as opposed to the  "bad"
> code I've seen so often in packages (and otherwise) which would
> use something like
> 
>    if(class(x) == "ftable")  ## really really bad
> 
> because that code will work incorrectly for objects with S3
> class attribute, say   c("myFtable", "ftable")
> 
> Martin
> 
> ______________________________________________
> R-package-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
 		 	   		  
	[[alternative HTML version deleted]]