[sage-devel] Re: Bug in hilbert numerator of a big ideal?
Hi! So, the workaround I found was to wrap cocoa using a subprocess... It works quite ok for my purposes... Thanks for your help! JP Le samedi 23 mai 2015 08:05:51 UTC+3, john_perry_usm a écrit : > > > Ok, I'll have a look at frobby! Just to check, I installed Macaulay2 and >> did the same computation and it gave me the right answer... >> > > CoCoA might also compute the correct answer, as I believe CoCoA switches > silently to bigint whenever it notices the need. I don't think Sage > supports CoCoA anymore though. (I'd like to fix that one day, but there are > lots of things I'd like to do one day...) > > >> So I could interface my code, but this is not the optimal way. If it is a >> bug, it should be looked at and repaired... >> > > I'm willing to be corrected here, but my reaction is that this is "not a > bug." Sage uses Singular as its commutative algebra engine, and Singular > advertises this limitation. As an analogy, it's not a bug if a computer > algebra system that cautions users that it works only modulo a prime p > gives you 1-2=p-1 instead of -1. > > That doesn't mean we can't address it. Hilbert series & polynomials work > with int, so maybe a workaround would be to check output from Singular, and > see if 2^32 divides it; if so, cancel the corresponding term. I don't know > if we can guarantee that this is always correct. > > Alternately, we can ask upstream. > > john perry > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Bug in hilbert numerator of a big ideal?
> Ok, I'll have a look at frobby! Just to check, I installed Macaulay2 and > did the same computation and it gave me the right answer... > CoCoA might also compute the correct answer, as I believe CoCoA switches silently to bigint whenever it notices the need. I don't think Sage supports CoCoA anymore though. (I'd like to fix that one day, but there are lots of things I'd like to do one day...) > So I could interface my code, but this is not the optimal way. If it is a > bug, it should be looked at and repaired... > I'm willing to be corrected here, but my reaction is that this is "not a bug." Sage uses Singular as its commutative algebra engine, and Singular advertises this limitation. As an analogy, it's not a bug if a computer algebra system that cautions users that it works only modulo a prime p gives you 1-2=p-1 instead of -1. That doesn't mean we can't address it. Hilbert series & polynomials work with int, so maybe a workaround would be to check output from Singular, and see if 2^32 divides it; if so, cancel the corresponding term. I don't know if we can guarantee that this is always correct. Alternately, we can ask upstream. john perry -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Bug in hilbert numerator of a big ideal?
Dear John, Ok, I'll have a look at frobby! Just to check, I installed Macaulay2 and did the same computation and it gave me the right answer... So I could interface my code, but this is not the optimal way. If it is a bug, it should be looked at and repaired... Best, Jean-Philippe Le jeudi 21 mai 2015 23:38:02 UTC+3, john_perry_usm a écrit : > > Actually, there seems to be a convenient frobby.hilbert() function which > does what you want, though I don't know if it's happy with larger > coefficients. > > john perry > > On Thursday, May 21, 2015 at 3:35:23 PM UTC-5, john_perry_usm wrote: >> >> Hi >> >> I factored this polynomial and found out that the number: >>> >>> 4294967296 (which should ring a bell) >>> >>> is a factor of the polynomial! >>> >> >> It didn't ring a bell for me, but factor(4294967296) enlightens me. :-) >> >> >>> It seems that the hilbert numerator uses singular. Can that be that >>> singular can not deal with very big integers?? Or with too many variables? >>> >> >> I *think* the first sentence on this page is related to your question: >> >> http://www.singular.uni-kl.de/Manual/latest/sing_90.htm#SEC129 >> >> >> In principle you could get around it using another Hilbert function. You >> could try frobby, which is actually in Sage (in mine, anyway), but I've >> never used it, so I can't be more helpful, sorry. >> >> john perry >> >> -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Bug in hilbert numerator of a big ideal?
Actually, there seems to be a convenient frobby.hilbert() function which does what you want, though I don't know if it's happy with larger coefficients. john perry On Thursday, May 21, 2015 at 3:35:23 PM UTC-5, john_perry_usm wrote: > > Hi > > I factored this polynomial and found out that the number: >> >> 4294967296 (which should ring a bell) >> >> is a factor of the polynomial! >> > > It didn't ring a bell for me, but factor(4294967296) enlightens me. :-) > > >> It seems that the hilbert numerator uses singular. Can that be that >> singular can not deal with very big integers?? Or with too many variables? >> > > I *think* the first sentence on this page is related to your question: > > http://www.singular.uni-kl.de/Manual/latest/sing_90.htm#SEC129 > > > In principle you could get around it using another Hilbert function. You > could try frobby, which is actually in Sage (in mine, anyway), but I've > never used it, so I can't be more helpful, sorry. > > john perry > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Bug in hilbert numerator of a big ideal?
Hi I factored this polynomial and found out that the number: > > 4294967296 (which should ring a bell) > > is a factor of the polynomial! > It didn't ring a bell for me, but factor(4294967296) enlightens me. :-) > It seems that the hilbert numerator uses singular. Can that be that > singular can not deal with very big integers?? Or with too many variables? > I *think* the first sentence on this page is related to your question: http://www.singular.uni-kl.de/Manual/latest/sing_90.htm#SEC129 In principle you could get around it using another Hilbert function. You could try frobby, which is actually in Sage (in mine, anyway), but I've never used it, so I can't be more helpful, sorry. john perry -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Bug in hilbert numerator of a big ideal?
Hi Jean Philippe, On 2015-05-21, jplab wrote: > It seems that the hilbert numerator uses singular. Can that be that > singular can not deal with very big integers?? Or with too many variables? I wouldn't be surprised, as I have problems with computing Hilbert series with Singular in a different context, too. Best regards, Simon -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
