October 2003 - PURRS-devel - BUGSENG Mailman

purrs/src Expr.defs.hh Expr.inlines.hh
by Tatiana Zolo 17 Oct '03

17 Oct '03

CVSROOT: /cvs/purrs Module name: purrs Changes by: zolo(a)cs.unipr.it 2003-10-17 16:14:35 Modified files: src : Expr.defs.hh Expr.inlines.hh Log message: Added the static function `Functor find_functor()'. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Expr.defs.hh.diff?cvsro… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Expr.inlines.hh.diff?cv…

1 0

purrs/src Symbol.defs.hh Symbol.inlines.hh
by Tatiana Zolo 17 Oct '03

17 Oct '03

CVSROOT: /cvs/purrs Module name: purrs Changes by: zolo(a)cs.unipr.it 2003-10-17 16:07:56 Modified files: src : Symbol.defs.hh Symbol.inlines.hh Log message: Added the explicit constructor `Symbol(const std::string& str)'. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Symbol.defs.hh.diff?cvs… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Symbol.inlines.hh.diff?…

1 0

[Fwd: Re: Type gen]
by Tatiana Zolo 17 Oct '03

17 Oct '03

-------- Original Message -------- Subject: Re: Type gen Date: Thu, 16 Oct 2003 18:01:00 +0200 From: "bernard.parisse" <parisse(a)mozart.ujf-grenoble.fr> To: Tatiana Zolo <zolo(a)cs.unipr.it> References: <3F8EC720.CC99C648(a)cs.unipr.it> Tatiana Zolo wrote: >Hi, >my problem is the following and it can be split in 3 parts: >1) transform my object (expressions of various types) in a giac object >2) factorize the giac object >3) come back to my object. > >I know how to do the problem 2): I used the type `gen' and the >functions in the file >sym2poly.{h,cc}. > >About the problems 1) and 2) I did something in both the directions but >I want ask you >the best method to decompose a type gen. >For example, if I have `g' of the type gen, where g = a + b + c + d, >how can I consider the single summands? The same problem with a product >of gen, >a power of gen, ... >How can I know if `g' is a sum of gen? Or a product of gen? >I did not find these methods in the file gen.h. >Is there documentation about these topics? > > > Hi! There is preliminary documentation in info format, if you installed giac by make install, you should have it in /usr/local/info (in emacs Ctrl-U Ctrl-H I followed by /usr/local/info/giac.info should open it). Alternatively the HTML translated file is on my web site. For your question, a gen g that is a sum or a product or any kind of symbolic function applied to arguments is such that g.type==_SYMB and g._SYMBptr->sommet is the operator or function, g._SYMBptr->feuille is the argument (if the function is unary) or the vector of arguments (if the function has more than 1 arg). You can check the function either by g._SYMBptr->sommet== at_plus (for +) or at_prod (for *) or any other operator, or by g.is_symb_of_sommet(at_plus) For exemple if g=a+b+c+d if (g.is_symb_of_sommet(at_plus)){ gen & arg=g._SYMBptr->feuille; if (arg.type==_VECT){ vecteur & v= *arg._VECTptr; int s=v.size(); // number of args ... } } bye!

1 0

Type gen
by Tatiana Zolo 16 Oct '03

16 Oct '03

Hi, my problem is the following and it can be split in 3 parts: 1) transform my object (expressions of various types) in a giac object 2) factorize the giac object 3) come back to my object. I know how to do the problem 2): I used the type `gen' and the functions in the file sym2poly.{h,cc}. About the problems 1) and 2) I did something in both the directions but I want ask you the best method to decompose a type gen. For example, if I have `g' of the type gen, where g = a + b + c + d, how can I consider the single summands? The same problem with a product of gen, a power of gen, ... How can I know if `g' is a sum of gen? Or a product of gen? I did not find these methods in the file gen.h. Is there documentation about these topics? Thanks, bye Tatiana

1 0

purrs/src Constant.defs.hh Constant.cc approxi ...
by Tatiana Zolo 16 Oct '03

16 Oct '03

CVSROOT: /cvs/purrs Module name: purrs Changes by: zolo(a)cs.unipr.it 2003-10-16 09:12:11 Modified files: src : Constant.defs.hh Constant.cc approximate_impl.hh Log message: Removed the not used constant `Euler'. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Constant.defs.hh.diff?c… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Constant.cc.diff?cvsroo… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/approximate_impl.hh.dif…

1 0

purrs/src Expr.defs.hh Expr.inlines.hh
by Tatiana Zolo 15 Oct '03

15 Oct '03

CVSROOT: /cvs/purrs Module name: purrs Changes by: zolo(a)cs.unipr.it 2003-10-15 17:52:47 Modified files: src : Expr.defs.hh Expr.inlines.hh Log message: Added the method `Expr::get_function_name()'. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Expr.defs.hh.diff?cvsro… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/src/Expr.inlines.hh.diff?cv…

1 0

[Fwd: Re: Question about giac.]
by Tatiana Zolo 15 Oct '03

15 Oct '03

-------- Original Message -------- Subject: Re: Question about giac. Date: Wed, 15 Oct 2003 12:49:27 +0200 (MET DST) From: Bernard Parisse <Bernard.Parisse(a)ujf-grenoble.fr> To: Tatiana Zolo <zolo(a)cs.unipr.it> > Dear Bernard, > I'm studying how it is possible to use your library for the > factorization. > I have some of questions: > Hello! > 1) I want to install only the necessary libraries for my aims. > I already have GMP and NTL and so I think would be enough > GIAC in order to the factorization on rationals, is it true? > (I do not work with floating point numbers and I am not interested > on graphic's tools.) > The installation's process is finished well, even if I have again > a doubt: what is DL support which by default is setted to `yes'? > DL support is for dynamic insertion of libraries in a running session of giac/xcas, you can keep giac configure guess safely > 2) The second question is about the input of an object of the type > `gen': > in the little program attached you can see that the use of `n*5' > instead of `5*n' caused an ambiguity for operator*. > It seems very strange for me this behavior... Is it wanted? > the reason is that I did not overload gen*int, only int*gen, you can use n*gen(5) to force the conversion int -> gen bye, Bernard Parisse

1 0

Question about giac.
by Tatiana Zolo 15 Oct '03

15 Oct '03

Dear Bernard, I'm studying how it is possible to use your library for the factorization. I have some of questions: 1) I want to install only the necessary libraries for my aims. I already have GMP and NTL and so I think would be enough GIAC in order to the factorization on rationals, is it true? (I do not work with floating point numbers and I am not interested on graphic's tools.) The installation's process is finished well, even if I have again a doubt: what is DL support which by default is setted to `yes'? 2) The second question is about the input of an object of the type `gen': in the little program attached you can see that the use of `n*5' instead of `5*n' caused an ambiguity for operator*. It seems very strange for me this behavior... Is it wanted? Thanks in advance, all the best Tatiana Zolo #include <giac/giac.h> using namespace std; using namespace giac; int main() { identificateur h("h"); identificateur n("n"); #if 1 gen p = pow(h,4)+pow(h,3)*(2*n+4)+pow(h,2)*(pow(n,2)+5*n+5) +h*(pow(n,2)+3*n+2); #else gen p = pow(h,4)+pow(h,3)*(2*n+4)+pow(h,2)*(pow(n,2)+n*5+5) +h*(pow(n,2)+3*n+2); #endif gen q = factor(p); cout << p << " -> " << q << endl << endl; return 0; }

1 0

purrs config.h.in
by Roberto Bagnara 03 Oct '03

03 Oct '03

CVSROOT: /cvs/purrs Module name: purrs Changes by: roberto(a)cs.unipr.it 2003-10-03 09:47:10 Removed files: . : config.h.in Log message: Automatically generated: should not be in the repository. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/config.h.in.diff?cvsroot=pu…

1 0