inline src_R also in latex src block

inline src_R also in latex src block

[Michael Bach]

Hello,

I am using org-mode version 7.8.02

I can successfully do inline `src_R' e.g. via

--8<---------------cut here---------------start------------->8---
\begin{equation}
  src_R{4 + 4}
\end{equation}
--8<---------------cut here---------------end--------------->8---

where `8' is put into the equation environment on export.  Sadly, there
is no LaTeX font lock in the org file this way.  But when I do it inside
a latex src block providing font lock, e.g.

--8<---------------cut here---------------start------------->8---
#+name: test
#+begin_src latex
  \begin{equation}
    src_R{4 + 4}
  \end{equation}
#+end_src
--8<---------------cut here---------------end--------------->8---

the `src_R' is put in literally on latex export.

Is there a way of preprocessing this latex src block with babel for R
prior to exporting?

Thanks in advance,
Michael

Re: inline src_R also in latex src block

[Eric Schulte]

Michael Bach <phaebz@gmail.com> writes:

> Hello,
>
> I am using org-mode version 7.8.02
>
> I can successfully do inline `src_R' e.g. via
>
> \begin{equation}
>   src_R{4 + 4}
> \end{equation}
>
> where `8' is put into the equation environment on export.  Sadly, there
> is no LaTeX font lock in the org file this way.  But when I do it inside
> a latex src block providing font lock, e.g.
>
> #+name: test
> #+begin_src latex
>   \begin{equation}
>     src_R{4 + 4}
>   \end{equation}
> #+end_src
>
> the `src_R' is put in literally on latex export.
>
> Is there a way of preprocessing this latex src block with babel for R
> prior to exporting?
>

I don't know of a way to evaluate inline code blocks which are already
inside of regular code blocks.  You could try direct begin_latex code
blocks, but in-buffer font lock support may simply not be possible in
this situation.

Best,

>
> Thanks in advance,
> Michael
>
>

-- 
Eric Schulte
http://cs.unm.edu/~eschulte/

Re: inline src_R also in latex src block

[Michael Bach]

Eric Schulte <eric.schulte@gmx.com> writes:


> I don't know of a way to evaluate inline code blocks which are already
> inside of regular code blocks.  

Thanks for your thoughts on this, Eric.  This would be *really*
convenient for me.  Too bad the christmas wishlist was processed
already.

> You could try direct begin_latex code blocks, but in-buffer font lock
> support may simply not be possible in this situation.
>

Sadly, direct begin_latex code blocks get fontified with `org-block' and
also export the inline `src_R' snippets literally.

What I was looking for is latex code fontified as AUCTeX and inline
`src_R' snippets evaluated prior to exporting.  The only working
combination I got right now is inline `src_R' evaluation and no AUCTeX
font-lock.  This is only a minor itch and not too bad considering what
org-mode already does for me.

Best Regards,
Michael

Re: inline src_R also in latex src block

[Andreas Leha]

Michael Bach <phaebz@gmail.com> writes:

> Eric Schulte <eric.schulte@gmx.com> writes:
>
>
>> I don't know of a way to evaluate inline code blocks which are already
>> inside of regular code blocks.  
>
> Thanks for your thoughts on this, Eric.  This would be *really*
> convenient for me.  Too bad the christmas wishlist was processed
> already.
>
>> You could try direct begin_latex code blocks, but in-buffer font lock
>> support may simply not be possible in this situation.
>>
>
> Sadly, direct begin_latex code blocks get fontified with `org-block' and
> also export the inline `src_R' snippets literally.
>
> What I was looking for is latex code fontified as AUCTeX and inline
> `src_R' snippets evaluated prior to exporting.  The only working
> combination I got right now is inline `src_R' evaluation and no AUCTeX
> font-lock.  This is only a minor itch and not too bad considering what
> org-mode already does for me.
>
> Best Regards,
> Michael


Hi Michael,

would this noweb-based approach work for you:

#+name: inner_test
#+begin_src R
  4 + 4
#+end_src

#+name: test
#+begin_src latex :noweb yes
  \begin{equation}
    <<inner_test()>>
  \end{equation}
#+end_src

Cheers,
Andreas

Re: inline src_R also in latex src block

[Michael Bach]

Andreas Leha <andreas.leha@med.uni-goettingen.de> writes:

> would this noweb-based approach work for you:
>
> #+name: inner_test
> #+begin_src R
>   4 + 4
> #+end_src
>
> #+name: test
> #+begin_src latex :noweb yes
>   \begin{equation}
>     <<inner_test()>>
>   \end{equation}
> #+end_src
>

Yes, that works.  Thanks for your thoughts on this, Andreas.

I am just thinking of another benefit:  Is it correct that I can use
caching with this approach?  So that the call to <<inner_test()>> does
not evaluate the `inner_test' block but instead takes its results
directly from the cached `#+results[6099...]: inner_test' block?  This
would be speedier than src_R usage...

Re: inline src_R also in latex src block

[Sebastien Vauban]

Hi Michael,

Michael Bach wrote:
> Andreas Leha <andreas.leha-A1rZ2h3LdSKGMSlLMZIubhS11BummzK+@public.gmane.org> writes:
>
>> would this noweb-based approach work for you:
>>
>> #+name: inner_test
>> #+begin_src R
>>   4 + 4
>> #+end_src
>>
>> #+name: test
>> #+begin_src latex :noweb yes
>>   \begin{equation}
>>     <<inner_test()>>
>>   \end{equation}
>> #+end_src
>
> Yes, that works.  Thanks for your thoughts on this, Andreas.
>
> I am just thinking of another benefit: Is it correct that I can use caching
> with this approach? So that the call to <<inner_test()>> does not evaluate
> the `inner_test' block but instead takes its results directly from the
> cached `#+results[6099...]: inner_test' block? This would be speedier than
> src_R usage...

That's perfectly correct: as long as your input code block does not change
(options on the "meta" line and/or contents), your results will be taken "as
is", instead of being recomputed every time.

Best regards,
  Seb

-- 
Sebastien Vauban

Re: inline src_R also in latex src block

[Andreas Leha]

Michael Bach <phaebz@gmail.com> writes:

> Andreas Leha <andreas.leha@med.uni-goettingen.de> writes:
>
>> would this noweb-based approach work for you:
>>
>> #+name: inner_test
>> #+begin_src R
>>   4 + 4
>> #+end_src
>>
>> #+name: test
>> #+begin_src latex :noweb yes
>>   \begin{equation}
>>     <<inner_test()>>
>>   \end{equation}
>> #+end_src
>>
>
> Yes, that works.  Thanks for your thoughts on this, Andreas.
>
> I am just thinking of another benefit:  Is it correct that I can use
> caching with this approach?  So that the call to <<inner_test()>> does
> not evaluate the `inner_test' block but instead takes its results
> directly from the cached `#+results[6099...]: inner_test' block?  This
> would be speedier than src_R usage...

yes, that should work.  Try to export the following several times.  The
first number should not change.
(note: maybe you'd have to execute the inner_test_cached once on your
system)

#+name: inner_test_cached
#+begin_src R :cache yes :exports none
  rnorm(1)
#+end_src

#+results[dbeb7280be41fdc949815808cf601a6d01a400bb]: inner_test_cached
: 1.12399967834262

#+name: inner_test_not_cached
#+begin_src R :exports none
  rnorm(1)
#+end_src

#+results: inner_test_not_cached
: -0.562660468468834

#+name: test
#+begin_src latex :noweb yes
  \begin{equation}
    <<inner_test_cached()>>
    !=
    <<inner_test_not_cached()>>
  \end{equation}
#+end_src

Cheers,
Andreas

Re: inline src_R also in latex src block

[Michael Bach]

"Sebastien Vauban"
<wxhgmqzgwmuf-geNee64TY+gS+FvcfC7Uqw@public.gmane.org> writes:

> That's perfectly correct: as long as your input code block does not change
> (options on the "meta" line and/or contents), your results will be taken "as
> is", instead of being recomputed every time.

Very nice, Thanks for confirmation!  

Michael

Re: inline src_R also in latex src block

[Michael Bach]

Andreas Leha <andreas.leha@med.uni-goettingen.de> writes:

> yes, that should work.  Try to export the following several times.  The
> first number should not change.
> (note: maybe you'd have to execute the inner_test_cached once on your
> system)
>
> #+name: inner_test_cached
> #+begin_src R :cache yes :exports none
>   rnorm(1)
> #+end_src
>
> #+results[dbeb7280be41fdc949815808cf601a6d01a400bb]: inner_test_cached
> : 1.12399967834262
>
> #+name: inner_test_not_cached
> #+begin_src R :exports none
>   rnorm(1)
> #+end_src
>
> #+results: inner_test_not_cached
> : -0.562660468468834
> #+name: test
> #+begin_src latex :noweb yes
>   \begin{equation}
>     <<inner_test_cached()>>
>     !=
>     <<inner_test_not_cached()>>
>   \end{equation}
> #+end_src

Thanks for the testing code.  Having R and LaTeX inside org this way is
just pure fun.

Michael