Actually, thinking more about this, I think that Tassilo is right.
Therefore, from now on, this condition will throw an error.
I would like to make this #ERROR in the table cell, but
this is not easy, so for the time being, it really stops
the evaluation with an error message.
Thanks, Tassilo, for pointing to this issue.
- Carsten
On Dec 5, 2008, at 8:26 AM, Carsten Dominik wrote:
Hi Tassilo,
I can see that this could be confusing, and maybe it should throw an
error.
The use-ase fo this are running means as column formulas
that would be undefined for he first few rows in. Consider
the following table:
| i | nn | mean of last 3 |
|---+----+----------------|
| 1 | 4 | 4.00 |
| 2 | 7 | 5.50 |
| 3 | 2 | 4.33 |
| 4 | 4 | 4.33 |
| 5 | 5 | 3.67 |
| 6 | 7 | 5.33 |
| 7 | 9 | 7.00 |
| 8 | 22 | 12.67 |
#+TBLFM: $3=vmean([EMAIL PROTECTED]);%.2f
I am not sure how common this use is and who would
be hurt by turning this behavior into an error.
- Carsten
On Dec 4, 2008, at 11:49 AM, Tassilo Horn wrote:
Carsten Dominik <[EMAIL PROTECTED]> writes:
Hi Carsten,
@-1 references the row above the current. If that is a hline, if
actually references the current line.
Yes, I see that, but I wonder if that's actually the right thing to
do.
IMO defaulting to some cell/row/column which is not referenced
correctly
might obscure wrong formulas.
For example, in my case the first calculation gave 2 which is correct
and only the second recalculation shows me that the forumla is wrong.
So I'd prefer an #ERROR if a reference doesn't exist.
Bye,
Tassilo
_______________________________________________
Emacs-orgmode mailing list
Remember: use `Reply All' to send replies to the list.
Emacs-orgmode@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-orgmode
