[GiNaC-devel] help

[Qin An]

Hi, 
 
Thank you for your answering :) 
 
I have worked out this problem. Yes, that is the reason that the exponent 2 of the expression must 
be defined to integer type. I am sorry that I provided it is double type, 2.0 . However, the error 
message do not cast light on me clearly enough. 
 
by the way, I found the compile result is different between the kdevelop and the console compile 
command. For example, 
 
// test.cpp 
#include <iostream> 
#include <ginac/ginac.h> 
using namespace std; 
using namespace GiNaC; 
 
int main(int argc, char *argv[]) 
{ 
 symbol x("x"); 
 
 ex exp = sin(x); 
 cout << exp << endl; 
 
 return EXIT_SUCCESS; 
}; 
 
that is ok when you typed the command: c++ test.cpp -o test -lcln -lginac , the result showed 
->sin(x); 
but on the kdevelop, the compiler will show you the error messages: 
 
/usr/local/include/ginac/container.h: In member function 
`GiNaC::container_storage<C>::STLT GiNaC::container<<anonymous template 
template parameter> >::evalchildren(int) const [with C = std::vector]': 
/usr/local/include/ginac/container.h:557: instantiated from `GiNaC::ex GiNaC::container<<anonymous 
template template parameter> >::eval(int) const [with C = std::vector]' 
/usr/local/include/ginac/container.h:130: instantiated from here 
/usr/local/include/ginac/container.h:708: exception handling disabled, use 
-fexceptions to enable 
 
That is same for asin,cos,acos, and so on. 
 
I think the error may be parameter type of the sin founction, because you definded the parameter of 
sin() is numeric type(that is smoothful when given the numeric type to x ). If I definded the x as 
numeric type, I have not given the ginac symbolic computing performance for trigonometric 
functions... That is so pity. I hope the symbolic computing performance will include the trigonometric 
functions. 
 
Could you give me some advice? 
Thanks again. 
 
(My computer is PIII 550,RedHat linux9, ginac-1.3.0, gcc-3.0, and the developing tool is kdevelop ) 
 
-Qin An 

======= 2004-12-22 05:48:17 您在来信中写道：=======

>Hi,
>
>I cannot reproduce your findings.
>
>On Mon, 20 Dec 2004, Qin An wrote:
>> some standard functions do not run smothfully everytime
>>
>>
>> the function -- expand:
>> when I used it as follow:
>> #include <iostream>
>> #include <ginac/ginac.h>
>> using namespace std;
>> using namespace GiNaC;
>>
>> main()
>> {
>>   ex e = expand(power(x+1,2));
>>   cout << e << endl;
>> }
>> the result given by ginac is: (x+1)^2
>
>You have not declared x.  When I do so, the output is 1+x^2+2*x, as it
>should.
>
>> but when I give him the parameters as follows:
>> main()
>> {
>>   ex e = expand((x+1)*(x+1));
>>   cout << e << endl;
>> }
>> the result is right: 1.0 + x^2.0 + 2.0*x
>
>I get the same as above.  Again, as it should.  But with integers, rather
>than with floating point values.
>
>> the difference above is conspicuous, when the above function  has been
>> used in other C++ class files that have not the main function.
>>
>> could anyone give some advice?  thanks!
>
>Please provide a self-contained example program, if necessary together
>with instructions how to compile and invoke it.  Please provide the output
>you obtain and the output you think would be correct.  Otherwise nobody
>will be able to help you.
>
>> by the way, my computer is PIII 550, REDHAT Linux 9,gcc is 3.0, and the
>> developing tool is kdevelop.
>
>Please include the version of GiNaC you are using.
>
>Regards
>    -richy.
>-- 
>Richard B. Kreckel
><http://www.ginac.de/~kreckel/>
>
>_______________________________________________
>GiNaC-devel mailing list
>GiNaC-devel at ginac.de
>http://thep.physik.uni-mainz.de/mailman/listinfo/ginac-devel
>

= = = = = = = = = = = = = = = = = = = =
			

　　　　　　　　致
礼！
 
				 
　　　　　　　　Qin An
　　　　　　　　qin.an at 163.com
　　　　　　　　　　2004-12-22