Given a module M over a multigraded ring S or a product of toric varieties X, this method finds the minimal elements of the multigraded Castelnuovo-Mumford regularity of M as defined in Definition 1.1 of [MS04] (see arXiv:math/0305214). If the input is an ideal, multigraded regularity of $S^1/I$ is computed.
This is done by calling the cohomologyHashTable method from TateOnProducts and checking for the multidegrees where Hilbert polynomial and Hilbert function match and where the higher sheaf cohomology vanishes.
Note that the module or ideal is assumed to be saturated by the irrelevant ideal of the Cox ring.
As an example, here we compute the minimal elements of the multigraded regularity for Example 1.4 of [BES20] (see arXiv:1703.07631). We consider the example of a hyperelliptic curve of genus 4 in $\PP^1\times\PP^2$.
i1 : X = toricProjectiveSpace(1)**toricProjectiveSpace(2) o1 = X o1 : NormalToricVariety |
i2 : S = ring X; B = ideal X; o3 : Ideal of S |
i4 : I = ideal(x_0^2*x_2^2+x_1^2*x_3^2+x_0*x_1*x_4^2, x_0^3*x_4+x_1^3*(x_2+x_3)) 2 2 2 2 2 3 3 3 o4 = ideal (x x + x x + x x x , x x + x x + x x ) 0 2 1 3 0 1 4 1 2 1 3 0 4 o4 : Ideal of S |
After saturating the defining ideal by the irrelevant ideal we may compute its multigraded regularity.
i5 : J = saturate(I,B); o5 : Ideal of S |
i6 : debugLevel = 1 o6 = 1 |
i7 : L = multigradedRegularity(X, J) -- Pic X = ZZ^2, dim X = 3, reg M = 7, HP(M) = 2*i_0+8*i_1-3 -- Computing cohomologyHashTable from {0, 0} to {7, 7} using temporary file name /tmp/M2-9192-0/0 -- Beginning search in Picard group -- Calculating minimal generators o7 = {{1, 5}, {2, 2}, {4, 1}} o7 : List |
This method also accepts the ring provided by productOfProjectiveSpaces from the TateOnProducts package.
The input is assumed to be saturated. Moreover, if the input is a module generated in non-positive degrees, then the output may be incorrect. In that case, adding the optional argument
LowerLimit => apply(n, i -> min(degrees M / (deg -> deg_i))) - dim X
where M is the module and X is the toric variety, may be a sufficient solution.
The object multigradedRegularity is a method function with options.