Weirdness re: inclusion of figures

Weirdness re: inclusion of figures

[Paul Magwene]

Hi All,

I'm trying to get up to speed with org-mode and babel for doing
reproducible computational research.  I'm just starting to play around
with simple examples, and I'm baffled by the following.

This first example, when exported to HTML or LaTeX produces the
expected result -- a simply code block with one embedded figure.

# Example 1.

This is my R example:

#+begin_src R :file z.png :results output graphics
plot(matrix(rnorm(100), ncol=2), type="l")
#+end_src

Some intervening text...

#+results:
[[file:z.png]]


However, this almost identical example, minus the intervening text
between the code and the results, doesn't include the figure:

# Example 2

This is my R example:

#+begin_src R :file z.png :results output graphics
plot(matrix(rnorm(100), ncol=2), type="l")
#+end_src

#+results:
[[file:z.png]]


What gives here? Do I always need to have intervening text between the
source code and results in order to get a figure in the exported
document?

Thanks,
Paul

Re: Weirdness re: inclusion of figures

[Paul]

Paul Magwene <pmmagic <at> gmail.com> writes:

> 
> Hi All,
> 
> I'm trying to get up to speed with org-mode and babel for doing
> reproducible computational research.  I'm just starting to play around
> with simple examples, and I'm baffled by the following.
> 
>

Oh, and I forgot to mention, my setup is:

Org version 7.8.03 with Emacs version 24

Re: Weirdness re: inclusion of figures

[Nick Dokos]

Paul Magwene <pmmagic@gmail.com> wrote:

> Hi All,
> 
> I'm trying to get up to speed with org-mode and babel for doing
> reproducible computational research.  I'm just starting to play around
> with simple examples, and I'm baffled by the following.
> 
> This first example, when exported to HTML or LaTeX produces the
> expected result -- a simply code block with one embedded figure.
> 
> # Example 1.
> 
> This is my R example:
> 
> #+begin_src R :file z.png :results output graphics
> plot(matrix(rnorm(100), ncol=2), type="l")
> #+end_src
> 
> Some intervening text...
> 
> #+results:
> [[file:z.png]]
> 
> 
> However, this almost identical example, minus the intervening text
> between the code and the results, doesn't include the figure:
> 
> # Example 2
> 
> This is my R example:
> 
> #+begin_src R :file z.png :results output graphics
> plot(matrix(rnorm(100), ncol=2), type="l")
> #+end_src
> 
> #+results:
> [[file:z.png]]
> 
> 
> What gives here? Do I always need to have intervening text between the
> source code and results in order to get a figure in the exported
> document?
> 

You need an :exports both header:

#+begin_src R :file z.png :results output graphics :exports both
plot(matrix(rnorm(100), ncol=2), type="l")
#+end_src

I don't know how or why the intervening text affects anything, but
it just smells like a red herring to me...

Nick :-)

Re: Weirdness re: inclusion of figures

[Yu]

Hello!

The behaviour is actually as expected. To understand this, you have to
consider, that org-mode currently only seems to recognize the relation
between source block and #+result paragraph if one of these two
conditions holds:
  o The "#+result:" line follows the source code without any text in between.
  o Both code block and "#+result:" line are labeled with the same name.

Also consider, as said by Nick Dokos, the significance of the
":exports" header argument:
  o By default the result is computed but not exported.
  o If the (unnamed) result is delimited from the source code by
intervening text, it is no longer considered any source blocks result:
The exporter just ignores the "#+result:" line and includes the image
into the exported file.

Labelling a source code block can be done with a "#+NAME:" line
preceding the code block. When evaluating the code block, the
"#+results"-line automatically gets named too. When reevaluating a
code block then the contents of this "#+results" entry is correctly
refreshed even after intervening code.

What you want thus probably is:

  : #+name: code-block
  : #+begin_src R :file z.png :results output graphics
  : plot(matrix(rnorm(100), ncol=2), type="l")
  : #+end_src
  :
  : Intervening text
  :
  : #+results: code-block
  : [[file:z.png]]

I did some tests to verify I'm right about this though, having
installed a recent version from git (at most 2 days since the last
pull):
  o No intervening text, no ":exports": Image is not exported.
  o No intervening text, ":exports both": Image is exported once; When
exporting the code is rerun.
  o Intevening text, no ":exports": The image is exported once. The
code block is not rerun on export.
  o Intervening text, ":exports both": The code is run twice (!), the
image is exported both before and after the text.

king regards, Yu


2012/2/16 Paul Magwene <pmmagic@gmail.com>:
> Hi All,
>
> I'm trying to get up to speed with org-mode and babel for doing
> reproducible computational research.  I'm just starting to play around
> with simple examples, and I'm baffled by the following.
>
> This first example, when exported to HTML or LaTeX produces the
> expected result -- a simply code block with one embedded figure.
>
> # Example 1.
>
> This is my R example:
>
> #+begin_src R :file z.png :results output graphics
> plot(matrix(rnorm(100), ncol=2), type="l")
> #+end_src
>
> Some intervening text...
>
> #+results:
> [[file:z.png]]
>
>
> However, this almost identical example, minus the intervening text
> between the code and the results, doesn't include the figure:
>
> # Example 2
>
> This is my R example:
>
> #+begin_src R :file z.png :results output graphics
> plot(matrix(rnorm(100), ncol=2), type="l")
> #+end_src
>
> #+results:
> [[file:z.png]]
>
>
> What gives here? Do I always need to have intervening text between the
> source code and results in order to get a figure in the exported
> document?
>
> Thanks,
> Paul
>