Bug #1779
Radical error with lex (again)
Description
This problem resurfaced now that radical is in C++.
/**/ use QQ[x,y],lex; /**/ I := ideal(x*y +x, x^2 +x); /**/ radI := radical(I); --> triggered error --> ERROR: Ring is not standard graded --> [CoCoALib] dim
I'm very grateful to
test/exbugs
and to its reference to redmine with the solution ;-)
Related issues
History
#1 Updated by Anna Maria Bigatti 3 months ago
- Related to Bug #1577: radical: lex order, ERROR: Ring is not standard graded added
#2 Updated by Anna Maria Bigatti 3 months ago
- Subject changed from Radical error with lex to Radical error with lex (again)
#3 Updated by Anna Maria Bigatti 3 months ago
- % Done changed from 0 to 30
first bug fixed.
New bug:
/**/ use ZZ/(5)[x,y,z]; /**/ SetVerbosityLevel(99); /**/ radI := radical(ideal(x^2+z,y+z)); --> gave error in factorizer --> ERROR: Poly must be non-constant --> [CoCoALib] UnivariateIndetIndex
in the code there is
factorization<RingElem> Q = factor(one(RingOf(I))); // sichtbar machen
which is a trick. Investigating why and how to avoid it
#4 Updated by Anna Maria Bigatti 3 months ago
- Related to Slug #1646: radical: could be more clever added
#5 Updated by John Abbott 3 months ago
- Status changed from New to In Progress
It would be easy to allow the factorize to return a result if the input is a non-zero constant.
I suppose I made it trigger an error because it is fairly pointless factorizing a constant.
Should I change the code to allow factorizing a constant?
UPDATE after skype discussion, we have decided to leave factorizing a deg 0 poly as an error
#6 Updated by Anna Maria Bigatti 3 months ago
- Related to Feature #1780: radical for ideals in SparsePolyRing: code in C++ added
#7 Updated by Anna Maria Bigatti 3 months ago
John Abbott wrote:
It would be easy to allow the factorize to return a result if the input is a non-zero constant.
I suppose I made it trigger an error because it is fairly pointless factorizing a constant.
Should I change the code to allow factorizing a constant?
Right now I'm chasing another problem. But I'll try to understand why there was that line, and then see whether that initialization made sense in that context.
#8 Updated by Anna Maria Bigatti 3 months ago
- Status changed from In Progress to Feedback
- % Done changed from 30 to 80
Anna Maria Bigatti wrote:
John Abbott wrote:
It would be easy to allow the factorize to return a result if the input is a non-zero constant.
I suppose I made it trigger an error because it is fairly pointless factorizing a constant.
Should I change the code to allow factorizing a constant?
I don't know why there was that initialization.
Now I call directly
factorization<RingElem> Q = SqFreeFactor(G); // redmine #1779.#3
And I haven't met any problem with the tests I have.
It seems all fine.
I have another set of tests (CoCoALib tests) which need some polishing. Then I'll close this issue.
#9 Updated by Anna Maria Bigatti 2 months ago
- Related to Support #1782: Make CoCoALib test file for radical added
#10 Updated by Anna Maria Bigatti 2 months ago
- % Done changed from 80 to 90
I made a new issue for implementing and running the tests.
Closing this one ... hoping non errors will come from the tests.
#11 Updated by John Abbott 2 months ago
- % Done changed from 90 to 80
#12 Updated by Anna Maria Bigatti 2 months ago
- Status changed from Feedback to Closed
- % Done changed from 80 to 100