- GiNaC-list - ginac.de

[[1]]!=1 ??
by Javier Ros Ganuza 18 Mar '06

18 Mar '06

I suppose that it is a feature... (matrix(3,1,lst(0,0,1)).transpose()*matrix(3,1,lst(0,0,1))).evalm() gives a 1,1 matrix [[1]]. To convert it to a number or to a expression not containing a matrix, I have to take directly the 0,0 element of the matrix. Is there a more direct method? If not I'm facing errors like (in ginsh notation) > a=[[0,0,1]]*[[0],[0],[1]]; > a=evalm(a); [[1]] > 1+cos(a); add::eval(): sum of non-commutative objects has non-zero numeric term Thank you Javier Ros

2 1

(no subject)
by Javier Ros Ganuza 09 Mar '06

09 Mar '06

Hello! :) When performing matrix multiplications expressions get expanded like in the following Example. Is there a way to avoid expansion, that is to leave the parenthesis Thanks Javier Ros Example: ----------------- matrix aux1 [[-d331*sin(q3)*dq3-d332*cos(q3)*dq3],[d331*cos(q3)*dq3-d332*sin(q3)*dq3],[0]] matrix R23 [[cos(q3),-sin(q3),0],[sin(q3),cos(q3),0],[0,0,1]] matrix R12.mul(aux1) [[d332*sin(q2)*sin(q3)*dq3-d331*cos(q2)*sin(q3)*dq3-d332*cos(q3)*cos(q2)*dq3-d331*sin(q2)*cos(q3)*dq3],[-d332*cos(q2)*sin(q3)*dq3-d331*sin(q2)*sin(q3)*dq3+d331*cos(q3)*cos(q2)*dq3-d332*sin(q2)*cos(q3)*dq3],[0]] ----------------- What I want is to obtain something like cos(q3)*(-d331*sin(q3)*dq3-d332*cos(q3)*dq3)+.... and so on for the first (0,0) matrix element.

3 4

problem with function pow
by Saurav Pathak 08 Feb '06

08 Feb '06

hi, i am having trouble turning a string with function "pow" into an expression. an example of the code is given below along with the output. i cannot figure out what i am doing wrong. thanks for your help in advance. -------- #include <iostream> #include <string> #include <ginac/ginac.h> #include <ginac/power.h> using namespace std; using namespace GiNaC; int main() { lst all; all.append(symbol("x")); // no pow string str2 = "x*x"; ex e2(str2, all); cout << "nopow : " << e2 << endl; // with pow string str = "pow(x, 2)"; ex e(str, all); cout << "withpow: " << e << endl; return 0; } the output is: nopow : x^2 terminate called after throwing an instance of 'std::runtime_error' what(): no function 'pow' with 2 parameters defined Aborted -------------------- -- saurav

2 4

Re: [GiNaC-list] problem with function pow
by Saurav Pathak 31 Jan '06

31 Jan '06

Anish Muttreja [Mon, Jan 30, 2006 at 07:56:32PM -0500]: + Isn't the function called "power"? + http://www.ginac.de/reference/classGiNaC_1_1power.html hi ashish, there is a wrapper function called "pow". i have been able to use the following without any problem: ex e3 = pow(x,2); it seems if "pow" is in a string, it does not work. saurav + On Mon, Jan 30, 2006 at 07:31:20PM -0500, Saurav Pathak wrote: + >hi, + > + >i am having trouble turning a string with function "pow" into an + >expression. an example of the code is given below along with the + >output. i cannot figure out what i am doing wrong. + > + >thanks for your help in advance. + >-------- + >#include <iostream> + >#include <string> + >#include <ginac/ginac.h> + >#include <ginac/power.h> + >using namespace std; + >using namespace GiNaC; + > + >int main() + >{ + > lst all; + > all.append(symbol("x")); + > + > // no pow + > string str2 = "x*x"; + > ex e2(str2, all); + > cout << "nopow : " << e2 << endl; + > + > // with pow + > string str = "pow(x, 2)"; + > ex e(str, all); + > cout << "withpow: " << e << endl; + > + > return 0; + >} + > + >the output is: + >nopow : x^2 + >terminate called after throwing an instance of 'std::runtime_error' + > what(): no function 'pow' with 2 parameters defined + >Aborted + >-------------------- + >-- + >saurav + >_______________________________________________ + >GiNaC-list mailing list + >GiNaC-list(a)ginac.de + >https://www.cebix.net/mailman/listinfo/ginac-list -- saurav

1 0

let_op(size_t ) problem
by Kerson Rocha Jr. 20 Jan '06

20 Jan '06

The following code: symbol x("x"); symbol y("y"); ex test1 = x + y; for ( size_t i = 0 ; i != test1.nops() ; i++ ) { try { cout << test1.let_op(i) << " " ; } catch(std::range_error & ba) { cout << "i = " << i << endl; cout << ba.what() << endl; } } This results: i = 0 basic::let_op(): add has no operands i = 1 basic::let_op(): add has no operands Using op() runs ok but I want change subexpressions. I am using GiNaC 1.3.3 on Mandriva 2006.0. Thanks.

2 1

operations modulo 2
by Kostas Patsakis 04 Jan '06

04 Jan '06

sorry for bothering the list again but I would appreciate some help. I'm trying to do operation(addition,multiplications) with multivariate polynomials. The variables cat take values 0 and 1 and I'm working with about 30 variables k0,k1... So since 0^2=0 and 1^1=1 I make the following substistution(written for ginsh) subs(e,$0^$1==$0). Since 2*x=0 mod 2 I need to do parse all the coeffiients mod 2. The problem is that even if I expand the expressions ginac(or at least my code to be honest) does not appear to expand my expressions halting immediately all my calculations e.g. this list of expressions is an output that was supposed to have been expanded and have been reduced mod 2 temp={1+k4+k5,(1+k5)*(1+k4)+k6,(1+k5)*(1+k6)*(1+k4)+k7,(1+k7)*(1+k5)*(1+k6)*(1+k4)+k8,1+k9+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4),1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4)),k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10,1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4)))+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10,k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4)))+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10,k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10)*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))),k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10)*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))),k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10)*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))),k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10)*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))),1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10*(1+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))*k10)*(1+k10+k9*(1+(1+k7)*(1+k8)*(1+k5)*(1+k6)*(1+k4))),0,0,1,0,1,0,1,0,0,1,1,1,0,k0,k1,k2,k3,1+k4} Using ginsh is obvious that this has not happened. Cutting and pasting code from manual has done some work but not much:( This is what I've got so far: const symbol & g_smbl(const string & s){ static map<string, symbol> directory; map<string, symbol>::iterator i = directory.find(s); if (i != directory.end()) return i->second; else return directory.insert(make_pair(s, symbol(s))).first->second; } struct map_rem_quad_evens : public map_function { ex var; ex tmp,tmp1; numeric coeff; map_rem_quad_evens(const ex & var_) : var(var_) {} ex operator()(const ex & e){ if (is_a<add>(e)) return e.map(*this); if (is_a<mul>(e)) return e.map(*this); else if (is_a<power>(e) && e.op(0).is_equal(var)) return e.op(0); else if (is_a<numeric>(e)){ coeff=ex_to<numeric>(e); if (coeff>0) return mod(coeff,2); else return e.map(*this); } else return e; } }; ex simpl(const ex & e){ ex tmp; tmp=subs(expand(e),pow(wild(),wild(1))==wild()); return tmp; } ex expr_simpl(const ex & e){ ex a; a=expand(e); map_rem_quad_evens rem_quad_evens(g_smbl("k0")); a=rem_quad_evens(e); a=simpl(a); a=rem_quad_evens(e); return a; } lst expr_simpl(const lst & a){ int i; lst tmp; tmp= 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0; for (i=0;i<word_len;i++) tmp[i]=expr_simpl(a[i]); return tmp; }

2 1

using modulo and expanding
by Kostas Patsakis 01 Jan '06

01 Jan '06

is there a way to use modulo over an expression? for example: 3*k1^2+4*k1*k2+1231231*k1 mod 2= k1^2+k1 all I really want is to work with polynomials over Z_2 so with subs I can do subs(a,$1^2=$1) but some times it's not enough cause I might run into something with power more than 2 would subs(a,$1^$2=$1) be enough??? I'm not using ginsh but c++ any help???

2 1

GiNaC library for FSC construction
by Vladimir Kisil 28 Dec '05

28 Dec '05

Dear All, I wrote a GiNaC library (http://arxiv.org/abs/cs.MS/0512073) which implements the Fillmore-Springer-Cnops construction (see abstract bellow). Best wishes to all for the coming New Year 2006! Vladimir -- Vladimir V. Kisil email: kisilv(a)maths.leeds.ac.uk -- www: http://maths.leeds.ac.uk/~kisilv/ Computer Science, abstract cs.MS/0512073 From: Vladimir V Kisil [view email] Date: Sat, 17 Dec 2005 15:09:11 GMT (149kb) Fillmore--Springer--Cnops Construction Implemented in GiNaC Authors: Vladimir V. Kisil Comments: LaTeX, 53 p; 9 PS graphics in one figure, the full source files are included, see last section of the paper Report-no: LEEDS-MATH-PURE-2005-29 Subj-class: Mathematical Software; Computational Geometry; Symbolic Computation This paper presents an implementation of the Fillmore--Springer--Cnops construction (FSCc) along with illustrations of its usage. FSCc linearises the linear-fraction action of the Moebius group in R^n. This has clear advantages in several theoretical and applied fields including engineering. Our implementation is based on the Clifford algebra capacities of the GiNaC computer algebra system (http://www.ginac.de/) which were described in cs.MS/0410044. The core of this realisation of FSCc is done for an arbitrary dimension of R^n with a metric given by an arbitrary bilinear form. We also present a subclass for two dimensional cycles (i.e. circles, parabolas and hyperbolas), which add some 2D specific routines including a visualisation to PostScript files through the MetaPost (http://www.tug.org/metapost.html) or Asymptote (http://asymptote.sourceforge.net/) packages. This software is the backbone of many results published in math.CV/0512416 and we use its applications there for demonstration. It can be ported (with various level of required changes) to other CAS with Clifford algebras capabilities similar to GiNaC.

1 0

small documentation fix
by varg＠theor.jinr.ru 25 Dec '05

25 Dec '05

Hello! Documentation on writing GiNaC classes contains code which won't even compile. Here is a fix: Index: doc/tutorial/ginac.texi =================================================================== RCS file: /home/cvs/GiNaC/doc/tutorial/ginac.texi,v retrieving revision 1.159.2.13 diff -r1.159.2.13 ginac.texi 7774c7774 < cout << e.bp->class_name() << endl; --- > cout << ex_to<basic>(e).class_name() << endl; -- All science is either physics or stamp collecting.

2 1

solving linear systems with ginac
by Kent-Andre Mardal 19 Dec '05

19 Dec '05

First of all, I would like to say that the GiNaC package is great! I have used it a lot lately! Now, I am trying to solve a linear system, which consists of 18 equations and 18 variables. The matrix and right-hand side do not contain floating point numbers. I have done similar things with smaller matrices and I very much appreciate that GiNaC gives the solution in non-floating point numbers. However, in this concrete case (which is list below). GiNaC is not able to solve the system on my machine with 1 Gb RAM. It runs out of memory! It is of course possible to solve this linear system by using some other library, but it also seems a bit strange that a 18x18 matrix should be that hard to invert. The matrix is non-singular and non of the eigenvalues are "close to 0". The matrix and right-hand side are as follows: Matrix: [ [0,0,0,0,-1/5,0,0,0,0,0,0,0,0,0,-1/30,-1/10,0,0], [0,0,0,0,-1/20,0,0,0,0,0,0,0,0,0,-1/20,-1/15,0,0], [0,0,0,0,-1/30,0,0,0,0,0,0,0,0,0,-1/5,-1/10,0,0], [0,0,0,0,0,0,1/30,1/5,0,0,0,0,0,1/10,0,0,0,0], [0,0,0,0,0,0,1/20,1/20,0,0,0,0,0,1/15,0,0,0,0], [0,0,0,0,0,0,1/5,1/30,0,0,0,0,0,1/10,0,0,0,0], [1/30*sqrt(2),1/10*sqrt(2),0,0,1/30*sqrt(2),0,0,1/5*sqrt(2),1/10*sqrt(2),0,1/5*sqrt(2),1/30*sqrt(2),0,0,0,0,1/10*sqrt(2),1/5*sqrt(2)], [1/20*sqrt(2),1/15*sqrt(2),0,0,1/20*sqrt(2),0,0,1/20*sqrt(2),1/15*sqrt(2),0,1/20*sqrt(2),1/20*sqrt(2),0,0,0,0,1/15*sqrt(2),1/20*sqrt(2)], [1/5*sqrt(2),1/10*sqrt(2),0,0,1/5*sqrt(2),0,0,1/30*sqrt(2),1/10*sqrt(2),0,1/30*sqrt(2),1/5*sqrt(2),0,0,0,0,1/10*sqrt(2),1/30*sqrt(2)], [1/60,1/30,0,1/360,0,1/180,1/60,1/20,0,1/180,0,1/120,1/60,1/30,0,0,1/90,1/120], [1/20,1/30,0,1/180,0,1/60,1/60,1/60,0,1/180,0,1/30,1/30,1/60,0,0,1/60,1/180], [1/60,1/60,0,1/120,0,1/90,1/20,1/60,0,1/180,0,1/120,1/30,1/30,0,0,1/180,1/360], [0,0,1/30,1/180,1/60,1/180,0,0,1/30,1/60,1/20,1/180,0,0,1/60,1/60,1/60,1/30], [0,0,1/60,1/360,1/20,1/180,0,0,1/30,1/180,1/60,1/120,0,0,1/60,1/30,1/90,1/120], [0,0,1/30,1/120,1/60,1/180,0,0,1/60,1/90,1/60,1/360,0,0,1/20,1/30,1/180,1/120], [0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0], [0,0,0,-2,0,0,0,0,0,2,0,0,0,0,0,0,0,0], [0,0,0,-2,0,2,0,0,0,0,0,0,0,0,0,0,0,0]] rhs: [[1],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0],[0]] Kent-Andre Mardal Simula Research Laboratory

2 6