[datatable-help] Making data.table shiny aware or vice-versa?

Matthew Dowle mdowle at mdowle.plus.com
Sat Dec 1 14:43:28 CET 2012


Adam,

Ah thanks for link. Yes that would have done it. Great.

For future reference, cedta.override is intended for calling 
environments we (or the data.table user) have no immediate control over 
(such as  gWidgetsWWW).  But if it's your environment or your package 
(and Depending or Importing data.table isn't possible) then you can set 
.datatable.aware=TRUE instead in your environment that's using 
data.table syntax. We couldn't add that to gWidgetsWWW because new 
variables can't be added to sealed environments, but we (and users) can 
change (the existing) cedta.override variable in the data.table 
namespace (being careful to add items to the existing vector) as an 
interim measure until a proper solution can be found.

Hopefully, Depending, Importing, cedta.override, and 
.datatable.aware=TRUE should cover all bases. Hadley has set 
.datatable.aware=TRUE in dplyr for example (it was his suggestion to add 
that actually, since dplyr has an _optional_ dependency on data.table).

Matthew


On 01.12.2012 05:50, Adam Caldwell wrote:
> Hmm, looks like this was a bug in shiny. I took a look and I was 
> still
> on 0.1.9, which wasnt showing as out of date with update.packages()
> since it was from an external repo.
>
> I think they fixed it a few days ago, here:
> 
> https://github.com/rstudio/shiny/commit/0b469f09df7e2ca3bbdb2ddadc8473a8126a9431
>
> Tested with 0.2.3 and everything works, with no workaround needed.
> Looks like this wasnt a data.table issue after all.
> _______________________________________________
> datatable-help mailing list
> datatable-help at lists.r-forge.r-project.org
> 
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help


More information about the datatable-help mailing list