Re: [BABEL] evaluation of R code in export

["Eric Schulte" <schulte.eric@gmail.com>]

Hi Rainer,

I general it is useful to have the results of Babel-evaluated code
appear in the session history as it can help for debugging.

Maybe (if you are specifying the session on a subtree or file wide
bases) it would be possible to change the session name to a different
name so that your normal working session won't be polluted.

This is butting up against a more general feature which has been dancing
in the twilight of Babel's glow for some time now.  Namely the ability
to specify header arguments which are conditional on the export state,
where export state can be something like
- nil :: not currently export, this is the default
- t :: an export is taking place
- html :: exporting to html
- latex :: exporting to latex
- ascii :: exporting to ascii
etc...

How exactly this should best be folded into the existing header argument
resolution schema is not immediately clear, so this development may take
some time.

If anyone has suggestions on syntax or behavior for this functionality
please do share.

Thanks -- Eric

Rainer M Krug <r.m.krug@gmail.com> writes:

> Hi
>
> assuming I have the following in an org buffer (and enabled evaluation
> of R code), and export it to html.
>
> The exported html looks perfect, but when I go into the *R_test* buffer,
> I see all the commands evaluated in the export (which is not such a
> problem), but they are also inserted in the history.
>
> When evaluating a line of code in ESS, this line is evaluated, but NOT
> added to the history. Would that be possible for evaluation of R code
> during export as well?
>
> My reasoning is, that I have a report in an .org file, which is using an
> existing R session in which I conduct simulations. When creating the
> report from a finished simulation, my command history in the *R* buffer
> contains all the commands evaluated for the export, and if I want to
> change one parameter for the simulation, I either have to go back
> several commands to find it, or type it in again. If the evaluated
> commands would not appear in the history, I could simply go one command
> back.
>
> If it is not to much work, it would be great if this could be changed.
>
> Cheers and thanks for an excellent tool,
>
> Rainer
>
> ############################################
>
> #+BABEL: session *R_test*
> ** test
> #+begin_src R :exports both
>   x <- 5
>   x
> #+end_src
>
> ############################################