Re: [O] BUG babel :colnames

[Andreas Leha]

Hi Eric,

thanks for looking into this.  And thanks for the confirmation already!

Good to know, that it is specific to R -- I had not checked.  Of course,
I do not want to use R to do mere subsetting, but rather some more
complex processing, which I do not want to do in shell (or elisp,....)

I have some more data on this bug if this is helpful:

(1)
On normal code blocks it depends on the order of the arguments.  It
seems to work for the last argument:

--8<---------------cut here---------------start------------->8---
#+name: testtable2
| a | b |
|---+---|
| 1 | 3 |
| 3 | 3 |
| 4 | 1 |
| 4 | 2 |

#+header: :var testtable=testtable2
#+header: :var dummy=20
#+begin_src R :colnames yes
  testtable[2:3,]
#+end_src

#+results:
| X1 | X3 |
|----+----|
|  4 |  1 |
|  4 |  2 |


#+header: :var dummy=20
#+header: :var testtable=testtable2
#+begin_src R :colnames yes
  testtable[2:3,]
#+end_src

#+results:
| a | b |
|---+---|
| 3 | 3 |
| 4 | 1 |
--8<---------------cut here---------------end--------------->8---

(2)
It is also buggy in call lines:

--8<---------------cut here---------------start------------->8---
#+name: testtable2
| a | b |
|---+---|
| 1 | 3 |
| 3 | 3 |
| 4 | 1 |
| 4 | 2 |


#+name: process_testtable
#+header: :var dummy=20
#+header: :var testtable=testtable2
#+begin_src R :colnames yes
  testtable[2:3,]
#+end_src

#+results: process_testtable
| a | b |
|---+---|
| 3 | 3 |
| 4 | 1 |

#+call: process_testtable() :colnames yes

#+results:
| a | b |
|---+---|
| 3 | 3 |
| 4 | 1 |

#+call: process_testtable(dummy=24) :colnames yes

#+results:
| a | b |
|---+---|
| 4 | 1 |
| 4 | 2 |

#+call: process_testtable(dummy=24, testtable=testtable2) :colnames yes

#+results:
| X1 | X3 |
|----+----|
|  4 |  1 |
|  4 |  2 |

#+call: process_testtable(testtable=testtable2) :colnames yes

#+results:
| X1 | X3 |
|----+----|
|  4 |  1 |
|  4 |  2 |
--8<---------------cut here---------------end--------------->8---


Regards,
Andreas


Eric Schulte <address@hidden> writes:

> This bug looks to be specific to Babel's R support.  E.g., it is not
> present with shell code blocks.
>
> #+name: testtable2
> | a | b |
> |---+---|
> | 1 | 3 |
> | 3 | 3 |
> | 4 | 1 |
> | 4 | 2 |
>
> #+header: :var testtable=testtable2
> #+begin_src sh :colnames yes
>   echo "$testtable"|tail -2
> #+end_src
>
> #+results:
> | a | b |
> |---+---|
> | 4 | 1 |
> | 4 | 2 |
>
> #+header: :var testtable=testtable2
> #+header: :var dummy=20
> #+begin_src sh :colnames yes
>   echo "$testtable"|tail -2
> #+end_src
>
> #+results:
> | a | b |
> |---+---|
> | 4 | 1 |
> | 4 | 2 |
>
> Best,
>
> Andreas Leha <address@hidden> writes:
>
>> Hi all,
>>
>> there is a bug when it comes to babel not respecting :colnames when more
>> variables are passed.
>>
>> Here is the MWE:
>>
>> --8<---------------cut here---------------start------------->8---
>> #+name: testtable2
>> | a | b |
>> |---+---|
>> | 1 | 3 |
>> | 3 | 3 |
>> | 4 | 1 |
>> | 4 | 2 |
>>
>> #+header: :var testtable=testtable2
>> #+begin_src R :colnames yes
>>   testtable[2:3,]
>> #+end_src
>>
>> #+results:
>> | a | b |
>> |---+---|
>> | 3 | 3 |
>> | 4 | 1 |
>>
>> #+header: :var testtable=testtable2
>> #+header: :var dummy=20
>> #+begin_src R :colnames yes
>>   testtable[2:3,]
>> #+end_src
>>
>> #+results:
>> | X1 | X3 |
>> |----+----|
>> |  4 |  1 |
>> |  4 |  2 |
>> --8<---------------cut here---------------end--------------->8---
>>
>>
>> Forgot to say in my last post:
>> Org-mode version 8.2.5h (release_8.2.5h-581-g7b11ba.dirty @ 
>> /home/andreas/local/emacs/org-mode-install/lisp/)
>> GNU Emacs 24.3.1 (x86_64-pc-linux-gnu, X toolkit, Xaw3d scroll bars) of 
>> 2013-12-22 on brahms, modified by Debian
>>
>>
>> Regards,
>> Andreas
>>
>>