Project

General

Profile

Bug #1140

ImplicitModular: too many bad primes

Added by John Abbott 4 months ago. Updated 3 months ago.

Status:
Closed
Priority:
Normal
Category:
bug
Target version:
Start date:
14 Dec 2017
Due date:
% Done:

100%

Estimated time:
1.01 h
Spent time:

Description

In the example below ImplicitModular decides that it has found a "better prime", but after that all primes seem to be "bad".
Very suspicious!

---------------------------
source("../IdealsModP.cpkg5");
---------------------------
SetVerbosityLevel(90);

R ::= QQ[x,y,z,v,w];
Use S ::= QQ[s,t];
f1:= t^5*s^2 -s -1;  f2:= s^3*t^2 +2*t^2 -t;  f3:= t^2 +s*t -1;  f4:=s^3-2*t+s; f5:= s^3-t^2;
L := [f1, f2, f3,f4, f5];
T := CpuTime();  G:= implicit(R,L); TimeFrom(T);
-- 4.643
T := CpuTime();  GM:= ImplicitModular(R,L,100); TimeFrom(T);
-- 6.820

This is just Renzo's example 3su6.cocoa5 (from IdealsModp) with starting prime 100.


Related issues

Related to CoCoALib - Feature #587: port to CoCoALib: Homomorphism pkg (ker, IsInjective, IsSurjective..)In Progress2014-07-14

History

#1 Updated by John Abbott 4 months ago

The critical prime appears to be 223.
If I start the calculation from p=223, it reports that all other primes are BAD.
But starting from 227 works OK.

#2 Updated by John Abbott 4 months ago

The defn of IsPrecedingNoCheck looks suspicious to me (compared to the defn in the article).
But changing the defn to what I think it should be has disastrous consequences for ex-bad.cocoa5

#3 Updated by Anna Maria Bigatti 4 months ago

  • Status changed from New to Closed
  • Assignee set to Anna Maria Bigatti
  • % Done changed from 0 to 100
  • Estimated time set to 1.01 h

this is related with implicit: it should (by default) return the ideal in a ring with weights, so that the generators are a GBasis.

#4 Updated by Anna Maria Bigatti 4 months ago

  • Related to Feature #587: port to CoCoALib: Homomorphism pkg (ker, IsInjective, IsSurjective..) added

#5 Updated by John Abbott 3 months ago

  • Target version changed from CoCoA-5.?.? to CoCoA-5.2.2

Also available in: Atom PDF