Re: Complex numbers

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

Renier Marchand <renierm@gmail.com> writes:

> It is a cool idea yes, but I quickly learned that it also have its
> problems, i.e. you cant really create nice column formulas because the
> reference is not aware of "the current row" etc.
>
> The biggest problem as I see it is that emacs-lisp does not support
> the complex data type as lisp (natively) do. I am going to look into a
> way of representing it.. maybe a two element lisp list (re im). I am
> not really versed in lisp so it is a bit of a problem to modify code,
> but I am looking into it.
>
> One problem that I have encountered is that when a two element list is
> referenced in the following scenario
>
> | (1 2) | #ERROR |
> #+TBLFM: $2='(sbe "myfunc" (data $1))
>
> an error is caused because lisp is trying to execute this list
> appearing struct, but when I do
>
> | (1 2) |
> | (2 3) |
> | resul |
> #+TBLFM: $1@3='(sbe "myfunc" (data @1$1..@2$1))
>
> no error is reported because it is assumed to be a list of data for my
> python code. i.e. it is sent as [[1,2],[2,3]] which is what I want. Is
> there a way to get more consistent behavior?
>

If you represent the complex number as a vector instead of a list, then
elisp will not assume that the first element is a function and try to
evaluate the data, e.g.

| [1 2]            |
|------------------|
| real:1 complex:2 |
#+TBLFM: @2$1='(sbe real (it @1$1))

#+source: real(it='())
#+begin_src emacs-lisp
  (format "real:%d complex:%d" (aref it 0) (aref it 1))
#+end_src

also, a quoted list '(1 2) would work...

Best -- Eric

>
> Thank you.
> Renier
>
> On Wed, Apr 13, 2011 at 6:35 PM, Eric Schulte <schulte.eric@gmail.com> wrote:
>> Renier Marchand <renierm@gmail.com> writes:
>>
>>> I have found the following way to reference my data correctly without
>>> having to quote it or have data rewritten.
>>>
>>> I do:
>>>
>>>       #+tblname: my-data
>>>       | hmin |
>>>       |------|
>>>       |      |
>>>       | 0.05 |
>>>       |  0.2 |
>>>       |  0.2 |
>>>       #+TBLFM: @5$1='(sbe "myfunc" (data "my-data[3:4,0]"))
>>>
>>> This also works for complex data as you suggested.
>>>
>>
>> Very cool, I would never have thought to use a reference within a table
>> formula.
>>
>>>
>>> But is there a way to refer in the formula to the current table
>>> instead of a specific table? Because this would obviously not work for
>>> multiple tables with the same name. (i.e. copy and paste for another
>>> dataset)
>>>
>>
>> No, there is no support for that sort of usage, and adding such a
>> position dependent reference would be a fairly large change from the
>> existing reference resolution mechanisms.
>>
>> Best -- Eric
>>
>>>
>>> Regards,
>>>
>>> Renier
>>>
>>> On Wed, Apr 13, 2011 at 5:52 AM, Eric Schulte <schulte.eric@gmail.com> wrote:
>>>> Hi Renier,
>>>>
>>>> The Org-mode table machinery is interpreting the values of your table
>>>> cells as emacs lisp (given that the table formula is an elisp, rather
>>>> than a calc formula).  Due to the "," the result is a weird nested list
>>>> which confuses your python code block.  Some options here include...
>>>>
>>>> 1. wrapping these cells in quotes so that they are passed to the python
>>>>   block as strings...
>>>>
>>>>   #+source: parameter-variation(data=0)
>>>>   #+begin_src python :result values
>>>>     return 'text'
>>>>   #+end_src
>>>>
>>>>   |---------------------------------------|
>>>>   | "(0.0331901438056,0.000535222885197)" |
>>>>   | "(0.0333434157791,0.000537930174356)" |
>>>>   | "(0.0345727512157,0.000559346040457)" |
>>>>   | "(0.0353146483908,0.000571501584524)" |
>>>>   | "(0.0355522909393,0.000574387067408)" |
>>>>   | "(0.0356575682336,0.000574851263615)" |
>>>>   | "(0.0357806926897,0.000575051685084)" |
>>>>   |---------------------------------------|
>>>>   | text                                  |
>>>>   #+TBLFM: @8$1='(sbe parameter-variation (nums @1$1..@7$1))
>>>>
>>>> 2. referencing the table from an external code block, rather than inside
>>>>   of a table formula.  This is probably the easier solution, but it
>>>>   doesn't insert the result into your table, unless you do something
>>>>   tricky like give the code block and the table the same name so that
>>>>   the results of the code block replace the table...
>>>>
>>>>   #+results: complex-data
>>>>   |-------------------------------------|
>>>>   | (0.0331901438056,0.000535222885197) |
>>>>   | (0.0333434157791,0.000537930174356) |
>>>>   | (0.0345727512157,0.000559346040457) |
>>>>   | (0.0353146483908,0.000571501584524) |
>>>>   | (0.0355522909393,0.000574387067408) |
>>>>   | (0.0356575682336,0.000574851263615) |
>>>>   | (0.0357806926897,0.000575051685084) |
>>>>   #+TBLFM: @8$1='(sbe parameter-variation (nums @1$1..@7$1))
>>>>
>>>>   #+begin_src python :var data=complex-data
>>>>     return data
>>>>   #+end_src
>>>>
>>>> Hope this helps -- Eric
>>>>
>>>> Renier Marchand <renierm@gmail.com> writes:
>>>>
>>>>> Hi.
>>>>>
>>>>> I have been playing around with complex data that has been returned
>>>>> from Python. This is obviously not in calc.el format but if I change
>>>>> them to the correct format I can manipulate them using calc.
>>>>>
>>>>> but
>>>>>
>>>>> When I want to pass the complex numbers (python format) to python I
>>>>> get an error. If I pass real number everything works as expected
>>>>>
>>>>> For example:
>>>>>
>>>>>       #+source: parameter-variation(data=0)
>>>>>       #+begin_src python :result values
>>>>>         return 'text'
>>>>>       #+end_src
>>>>>
>>>>>
>>>>>       |   |  hmin |                                     |
>>>>>       |---+-------+-------------------------------------|
>>>>>       |   |       |                                     |
>>>>>       |   |  0.05 | (0.0331901438056,0.000535222885197) |
>>>>>       |   |   0.1 | (0.0333434157791,0.000537930174356) |
>>>>>       |   |   0.3 | (0.0345727512157,0.000559346040457) |
>>>>>       |   |   0.6 | (0.0353146483908,0.000571501584524) |
>>>>>       |   |   0.9 | (0.0355522909393,0.000574387067408) |
>>>>>       |   |   1.2 | (0.0356575682336,0.000574851263615) |
>>>>>       |   |  10.0 | (0.0357806926897,0.000575051685084) |
>>>>>       | $ | x=0.1 | y=0.1                               |
>>>>>       |   |  text |                                     |
>>>>>       #+TBLFM: @11$2='(sbe "parameter-variation" (data
>>>>> @3$2..@9$2))::@11$3='(sbe "parameter-variation" (data @3$3..@9$3))
>>>>>
>>>>> i.e. I get the word 'text' returned for column 2 where there are real
>>>>> numbers but I don't get anything returned where there are complex
>>>>> numbers. As you can see, there are no actual calculation performed on
>>>>> the data I am just returning 'text' so I am expecting it to work in
>>>>> both instances.
>>>>>
>>>>> The debug sessions show the following for the real column:
>>>>>
>>>>> Substitution history of formula
>>>>> Orig:   '(sbe "parameter-variation" (data @3$2..@9$2))
>>>>> $xyz->  '(sbe "parameter-variation" (data @3$2..@9$2))
>>>>> @r$c->  '(sbe "parameter-variation" (data #("0.05" 0 4 (fontified t
>>>>> face org-table)) #("0.1" 0 3 (fontified t face org-table)) #("0.3" 0 3
>>>>> (fontified t face org-table)) #("0.6" 0 3 (fontified t face
>>>>> org-table)) #("0.9" 0 3 (fontified t face org-table)) #("1.2" 0 3
>>>>> (fontified t face org-table)) #("10.0" 0 4 (fontified t face
>>>>> org-table))))
>>>>> $1->    '(sbe "parameter-variation" (data #("0.05" 0 4 (fontified t
>>>>> face org-table)) #("0.1" 0 3 (fontified t face org-table)) #("0.3" 0 3
>>>>> (fontified t face org-table)) #("0.6" 0 3 (fontified t face
>>>>> org-table)) #("0.9" 0 3 (fontified t face org-table)) #("1.2" 0 3
>>>>> (fontified t face org-table)) #("10.0" 0 4 (fontified t face
>>>>> org-table))))
>>>>> Result: text
>>>>> Format: NONE
>>>>> Final:  text
>>>>>
>>>>> and for the complex column:
>>>>>
>>>>> Substitution history of formula
>>>>> Orig:   '(sbe "parameter-variation" (data @3$3..@9$3))
>>>>> $xyz->  '(sbe "parameter-variation" (data @3$3..@9$3))
>>>>> @r$c->  '(sbe "parameter-variation" (data
>>>>> #("(0.0331901438056,0.000535222885197)" 0 35 (fontified t face
>>>>> org-table)) #("(0.0333434157791,0.000537930174356)" 0 35 (fontified t
>>>>> face org-table)) #("(0.0345727512157,0.000559346040457)" 0 35
>>>>> (fontified t face org-table)) #("(0.0353146483908,0.000571501584524)"
>>>>> 0 35 (fontified t face org-table))
>>>>> #("(0.0355522909393,0.000574387067408)" 0 35 (fontified t face
>>>>> org-table)) #("(0.0356575682336,0.000574851263615)" 0 35 (fontified t
>>>>> face org-table)) #("(0.0357806926897,0.000575051685084)" 0 35
>>>>> (fontified t face org-table))))
>>>>> $1->    '(sbe "parameter-variation" (data
>>>>> #("(0.0331901438056,0.000535222885197)" 0 35 (fontified t face
>>>>> org-table)) #("(0.0333434157791,0.000537930174356)" 0 35 (fontified t
>>>>> face org-table)) #("(0.0345727512157,0.000559346040457)" 0 35
>>>>> (fontified t face org-table)) #("(0.0353146483908,0.000571501584524)"
>>>>> 0 35 (fontified t face org-table))
>>>>> #("(0.0355522909393,0.000574387067408)" 0 35 (fontified t face
>>>>> org-table)) #("(0.0356575682336,0.000574851263615)" 0 35 (fontified t
>>>>> face org-table)) #("(0.0357806926897,0.000575051685084)" 0 35
>>>>> (fontified t face org-table))))
>>>>> Result:
>>>>> Format: NONE
>>>>> Final:
>>>>>
>>>>>
>>>>> Thank you very much for your help.
>>>>>
>>>>> Renier
>>>>>
>>>>>
>>>>
>>>> --
>>>> Eric Schulte
>>>> http://cs.unm.edu/~eschulte/
>>>>
>>>
>>
>> --
>> Eric Schulte
>> http://cs.unm.edu/~eschulte/
>>
>

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