-----"Bowo Prasetyo" <prazjp at gmail.com> wrote: -----
>To: "Barton Willis" <willisb at unk.edu>
>From: "Bowo Prasetyo" <prazjp at gmail.com>
>Date: 03/29/2007 05:03AM
>cc: maxima at math.utexas.edu
>Subject: Re: [Maxima] [realonly: true] doesn't work to [solve]
>
>Thank you Barton,
>
>On 29/03/07, Barton Willis <willisb at unk.edu> wrote:
>>
>> Yes, I believe it is by design. The user documentation for
>> 'realonly' doesn't say anything about the solve function:
>>
>
>I wonder because 'realonly: true' works to 'solve' in the following
>example..
>
>realonly: true;
>solve([4*x^2-y^2 = 12,x*y-x = 2],[x,y]);
>
>It gave me solutions which are free of %i..
For these equations, solve calls algsys (try trace(algsys)).
If solve calls algsys, the 'realonly' flag might make
a difference. I didn't think of this--sorry.
>> Note: Free of %i doesn't mean real. So be careful.
>
>In what case it is not real..?
>
(%i28) algsys([x=asin(45)],[x]), realonly : true;
(%o28) [[x=asin(45)]]
asin(45) is free of %i, but it isn't real. So be careful.
Barton