[datatable-help] join vignette for data.table

[Arunkumar Srinivasan]

Sure. It’s on the list of tasks for this release as well, so I’ll have to work on it. Will try to put it up somewhere.

But the most useful thing you / anyone else could do is to provide / point to a data set (preferably in connection with flights, maybe weather?) that I can use as a 2nd table to explain joins. I’m not quite satisfied with the small artificial dataset that I’m using at the moment.

Also, the kind of questions that could be asked from that dataset using those two ( or more?) tables which requires joins (+ some other tasks) would be great! That’s the hardest part for me at the moment, not the explanations :-).

-- 
Arun

On 22 December 2015 at 14:42:38, stat quant (statquant at outlook.com) wrote:

I'd be happy to help.
Happy for me to have a look at it?

cheers 

On Tuesday, 22 December 2015, Arunkumar Srinivasan <aragorn168b at gmail.com> wrote:
In the works. I’ve not yet managed to finish.

require(data.table)
A = data.table(x=1:2, y=3:4)
B = data.table(x=2:3, y=3:4)

anti-join:

A[!B, on="x"] # equivalently A[!B, .SD, on="x"]

semi-join:

A[B, .SD, nomatch=0L, on="x"]

-- 
Arun

On 21 December 2015 at 17:09:22, statquant3 (statquant at outlook.com) wrote:

Hello, is there a join vignette for data.table ?
I realize that I am now more and more using dplyr just for semi_join and
anti_join.
Is there any plan to write those sugar function ?
Happy to do it here and get feedback.

C.



--
View this message in context: http://r.789695.n4.nabble.com/join-vignette-for-data-table-tp4715792.html
Sent from the datatable-help mailing list archive at Nabble.com.
_______________________________________________
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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/datatable-help/attachments/20151222/2627b760/attachment.html>