Difference in behavior running in CLI and from a Java Program

Robert,

Thanks!

You were right... my program had inserted a space between the ":" and  
"=" which kept the definition of PolMat from being recognized.   
Fixing that solved the problem.

The lower case p was a type on my part.

This did bring up a more Maxima related question, though.  The result  
of integrating the zero elements is just the integral - it does not  
do the evaluation even though it does for the element whose value is  
"1".

Anyway, Thanks again and have a nice weekend,

Chris

On Feb 24, 2007, at 12:40 PM, Robert Dodier wrote:

> On 2/24/07, Chris Bevis <cfbevis at gmail.com> wrote:
>
>> When I enter the commands :
>>
>> PolMat(P) := matrix(  [ 1, cos(2*P), sin(2*P), 0 ], [ cos(2*P),
>> cos(P)*cos(P), sin(2*P)*cos(2*P), 0], [ sin(2*P), sin(2*P)*cos(2*P),
>> sin(2*P)*sin(2*P), 0],[ 0, 0, 0,0 ] ) ;
>> integrate(PolMat(P), P ) ;
>>
>> through this interface, it returns :
>>
>> integrate(PolMat(P), p)
>
> Chris, I don't see that behavior when I run the above input through
> the program RunProgram as attached to the message I posted Feb 10.
> I get the same output whether the input is entered through the command
> line interface or RunProgram. I tried RunProgram with Maxima + GCL.
> What does build_info() say?
>
> I'm guessing that there might be a spelling error -- either PolMat  
> is spelled
> differently or integrate. Another possibility is that the initial
> PolMat(P) := ...
> was lost somehow so PolMat is undefined when integrate is called.
>
> You might try putting display2d:false; writefile("tmp.log"); into the
> input stream. Does tmp.log then show the input you expected?
>
> I would be surprised if Maxima changed P to p (as shown in the  
> output).
> Did you copy the output verbatim?
>
> I didn't know that integrate distributes over matrix ... I learn
> something new every day.
>
> HTH
> Robert Dodier