October 2001 - PPL-devel - BUGSENG Mailman

ppl/doc definitions.dox
by Patricia Hill 22 Oct '01

22 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: hill(a)cs.unipr.it 2001-10-22 09:52:23 Modified files: doc : definitions.dox Log message: I was wrong to remove the "^\transpose" in Weil's theorem - restored. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/doc/definitions.dox.diff?cvsr…

1 0

generators
by P M Hill 22 Oct '01

22 Oct '01

Hi, I have two queries wrt the generators. As I understand it (and as stated on page 14 of the user manual, the linear expression used to define a vertex, ray or line should be homogeneous. After all, the constant part is irrelevant: we say that the constant part is disregarded. I think though that if it is really irrelevant and disregarded, we should not allow it. Surely, if the user includes a constant term, the user must have misunderstood the representation we are using. It seems to me that it is better that the user gets an error message than just ignore it. The second query is similar. For a ray and line the expression begins with n ^ or n | where n is any number. I am not clear a) why we let the user type anything here other than 1 (or even why the expression has to start with a number at all, all we need is a means of indicating that the following linear expression is to denote a line or a ray). Also, can this number be 1.234 - ie a rational or must it be an integer. Can it be 0? So all these questions I have (and therefore, maybe other ignorant users(!)), yet, if we just said it had to be 1 (and enforced that) then it is much simpler. ciao, Pat

1 0

ppl/src Polyhedron.defs.hh
by Elisa Ricci 22 Oct '01

22 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: ericci(a)cs.unipr.it 2001-10-22 09:25:31 Modified files: src : Polyhedron.defs.hh Log message: In two examples, erased the spaces between the factor and the variable. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Polyhedron.defs.hh.diff?c…

1 0

ppl/doc definitions.dox
by Patricia Hill 22 Oct '01

22 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: hill(a)cs.unipr.it 2001-10-22 09:04:55 Modified files: doc : definitions.dox Log message: Comment at the end of Weil's theorem had "convex" and "positive" and "rays" and "vertices" wrong way round. Minor change to wording introducing the two theorems. In final comment of user part, "blurb" removed. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/doc/definitions.dox.diff?cvsr…

1 0

ppl/doc definitions.dox
by Patricia Hill 22 Oct '01

22 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: hill(a)cs.unipr.it 2001-10-22 08:49:28 Modified files: doc : definitions.dox Log message: Removed spurious ^\transpose (^T) in Weil's theorem. Added "dual representation" as heading and emphasised in text at end of user definitions. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/doc/definitions.dox.diff?cvsr…

1 0

ppl/src ConSys.defs.hh GenSys.defs.hh Generato ...
by Enea Zaffanella 21 Oct '01

21 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2001-10-21 15:50:41 Modified files: src : ConSys.defs.hh GenSys.defs.hh Generator.defs.hh Generator.inlines.hh LinExpression.defs.hh LinExpression.inlines.hh Polyhedron.defs.hh Log message: Systems of constraints and generators do not define polyhedra; rather, they may "correspond to" a given polyhedron (the one that would be built by applying the adequate constructor). A system of generator does not need to include a vertex, in general. The vertex is needed if the system of generators is meant to correspond to a non-empty polyhedron. The detailed description of class Generator changed to highlight that vertices, lines and rays are represented as a type plus a direction in the space. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/ConSys.defs.hh.diff?cvsro… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/GenSys.defs.hh.diff?cvsro… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Generator.defs.hh.diff?cv… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Generator.inlines.hh.diff… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/LinExpression.defs.hh.dif… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/LinExpression.inlines.hh.… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Polyhedron.defs.hh.diff?c…

1 0

ppl Makefile.am
by Roberto Bagnara 21 Oct '01

21 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: roberto(a)cs.unipr.it 2001-10-21 14:00:43 Modified files: . : Makefile.am Log message: CREDITS added to the distribution. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/Makefile.am.diff?cvsroot=ppl&…

1 0

w3ppl/htdocs/Links Links.raw
by Roberto Bagnara 21 Oct '01

21 Oct '01

CVSROOT: /cvs/ppl Module name: w3ppl Changes by: roberto(a)cs.unipr.it 2001-10-21 13:52:52 Modified files: htdocs/Links : Links.raw Log message: Finished, for now. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/w3ppl/htdocs/Links/Links.raw.diff…

1 0

ppl/src Constraint.defs.hh Constraint.inlines.hh
by Enea Zaffanella 21 Oct '01

21 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2001-10-21 13:04:46 Modified files: src : Constraint.defs.hh Constraint.inlines.hh Log message: In the documentation, used n to denote the dimension of the space. In the renaming operator >>, the parameter p renamed into offset, consistently with the comment. In the examples, removed spaces between factor and variable when specifying a linear expression (e.g., "3 * x + 4 * y" now is written "3*x + 4*y"). Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Constraint.defs.hh.diff?c… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Constraint.inlines.hh.dif…

1 0

ppl/doc definitions.dox
by Enea Zaffanella 21 Oct '01

21 Oct '01

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2001-10-21 13:01:18 Modified files: doc : definitions.dox Log message: When specifying the coordinates of a point, indexes should range from 0 to n-1, if n is the dimension of the space. This is to be consistent with definition of class Variable. Change applied to the definitions of scalar-product, of polytope and to Weil's theorem. Note: I do not like to say a polyhedron is "open" in one direction; I prefer the terminology "unbounded", so I changed it. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/doc/definitions.dox.diff?cvsr…

1 0