newPackage(
"RandomIdeals",
Version => "2.0",
Date => "May 9, 2016",
Authors => {
{Name => "Katie Ansaldi",
Email => "kansaldi@gmail.com"},
{Name => "David Eisenbud",
Email => "de@msri.org",
HomePage => "http://www.msri.org/~de"},
{Name => "Robert Krone",
Email => "rckrone@gmail.com",
HomePage => "http://rckr.one"},
{Name => "Jay Yang",
Email => "jkelleyy@gmail.com"}
},
HomePage => "http://www.msri.org/~de",
Headline => "creating random ideals of various sorts",
AuxiliaryFiles => false, -- set to true if package comes with auxiliary files,
PackageExports =>{"EdgeIdeals", "BinomialEdgeIdeals"},
DebuggingMode => false -- set to true only during development
)
export {
"randomIdeal",
"randomMonomialIdeal",
"randomSquareFreeMonomialIdeal",
"randomSquareFreeStep",
"randomBinomialIdeal",
"randomPureBinomialIdeal",
"randomSparseIdeal",
"randomEdgeIdeal",
"randomBinomialEdgeIdeal",
"randomToricEdgeIdeal",
"randomElementsFromIdeal",
"randomMonomial",
"squareFree",
"regSeq",
"AlexanderProbability",
"randomAddition",
"randomShelling",
"idealFromShelling",
"idealChainFromShelling",
"isShelling",
"randomShellableIdeal",
"randomShellableIdealChain"
}
randomMonomial = method(TypicalValue => RingElement)
randomMonomial(ZZ,Ring) := RingElement => (d,S) -> (
m := basis(d,S);
m_(0,random rank source m))
randomMonomialIdeal=method(TypicalValue => Ideal)
randomMonomialIdeal(Sequence, Ring) := Ideal => (L,S)->
randomMonomialIdeal(toList L, S)
randomMonomialIdeal(List, Ring) := Ideal => (L,S)->(
--produces an ideal minimally generated
--by random monomials of degrees given in L,
--unless the low degree terms generate all monomials
--in one of the higher degrees requested,
--in which case the ideal fulfills as many of
--the conditions as possible (from lowest degree),
--and produces a message.
--
--first produce a list in ascending order of degree,
--specifying how many elements are required from that degree
Lu := sort unique L;
Ls := apply(Lu, t -> #positions(L, s->(s==t)));
--handle the initial degree
M := flatten entries basis(Lu#0, S^1);
if Ls#0 < #M then I := ideal((random M)_{0..Ls#0-1})
else (
I = ideal(M);
<<"***** there are only "
<< binomial(Lu#0+numgens S-1,Lu#0)
<<" monomials of degree "
<< Lu#0
< Ideal)
randomSquareFreeMonomialIdeal(Sequence, Ring) := Ideal => (L,S)->
randomSquareFreeMonomialIdeal(toList L, S)
randomSquareFreeMonomialIdeal(List, Ring) := Ideal => (L,S)->(
--L is a list of degrees desired, The routine
--produces an ideal minimally generated by random
--square-free monomials of exactly the numbers and degrees given in L,
--unless the low degree terms generate
--all monomials in one of the higher degrees requested,
--in which case the ideal fulfills as
--many of the conditions as possible (from lowest degree).
--
--first produce a list in ascending order of degree, specifying how many elements are required from that degree
Lu := sort unique L;
Ls := apply(Lu, t -> #positions(L, s->(s==t)));
M := flatten entries gens squareFree(Lu#0, S);
if Ls#0 < #M then I := ideal((random M)_{0..Ls#0-1})
else (I = ideal(M);
<<"***** there are only " <{AlexanderProbability => .05})
soc = I -> (
--I should be a MonomialIdeal
p:=product gens ring I;
apply(flatten entries gens dual I, m->p//m))
prepare= I->(
--I should be a MonomialIdeal
Igens := flatten entries gens I;
ISocgens := soc I;
{I,Igens,ISocgens})
randomSquareFreeStep(Ideal) := o -> (I) ->(
if isMonomialIdeal I then randomSquareFreeStep(monomialIdeal I, o)
else error "ideal not generated by monomials")
randomSquareFreeStep(MonomialIdeal) := o -> (I) ->(
if isSquareFree I then randomSquareFreeStep (prepare I, o)
else error "ideal not square-free")
randomSquareFreeStep(List) := o -> (L) ->(
I := L_0;
S := ring I;
mm := monomialIdeal vars S;
mm2 := monomialIdeal apply (flatten entries vars S, p->p^2);
--With probability given by the option AlexanderProbability (default .05),
--the routine simply returns the Alexander dual of I.
if random RR j!=p);
p' := I-p; -- returns the ideal generated by all the gens but p
sqFreeMults := (mm*p)-mm2;
-- sqFreeMults := substitute(monomialIdeal(gens(mm*ideal(Igens_p))%mm2), S);
-- J = trim (ideal(Igens_p')+sqFreeMults)
J = p'+sqFreeMults
);
--Decide whether to accept the step or not
Jgens := flatten entries gens J;
nJ := #Jgens;
JSocgens := soc J;
nSocJ := #JSocgens;
--accept with probability P (if P<1) or unconditionally (P>=1):
P := (nI+nSocI)/(nJ+nSocJ);
if random RR < P then {J,Jgens,JSocgens} else {I,Igens,ISocgens}
)
squareFree = method(TypicalValue => Ideal)
squareFree(ZZ,Ring) := Ideal => (d,S) ->
--returns the ideal generated by all square free monomials of degree d
ideal compress(basis(d,S^1) % ideal apply (flatten entries vars S, p -> p^2))
randomPureBinomialIdeal = method(TypicalValue => Ideal)
randomPureBinomialIdeal(Sequence, Ring) := Ideal => (L,S)->randomPureBinomialIdeal(toList L, S)
randomPureBinomialIdeal(List, Ring) := Ideal => (L,S)->(
--L=list of degrees of the generators
trim ideal apply(L, d->randomMonomial(d,S)-randomMonomial(d,S))
)
randomBinomialIdeal = method(TypicalValue => Ideal)
randomBinomialIdeal(Sequence, Ring) := Ideal => (L,S)->randomPureBinomialIdeal(toList L, S)
randomBinomialIdeal(List, Ring) := Ideal => (L,S)->(
--L=list of degrees of the generators
kk:=ultimate (coefficientRing, S);
trim ideal apply(L, d->randomMonomial(d,S)-random(kk)*randomMonomial(d,S))
)
///
///
randomSparseIdeal = method(TypicalValue => Ideal)
randomSparseIdeal(Matrix, ZZ, ZZ) := Ideal => (B,s,n) -> (
-- B is a 1xt matrix of monomials
-- s is the size of each poly
-- n is the number of polys
S := ring B;
t := rank source B;
BB := (first entries B);
kk := ultimate(coefficientRing, S);
trim ideal apply(n, j ->
sum apply(s, i -> random kk * BB#(random t)))
)
randomIdeal = method(TypicalValue => Ideal)
randomIdeal(Sequence,Matrix) := Ideal => (L,B) -> randomIdeal(toList L, B)
randomIdeal(List,Matrix) := Ideal => (L,B) -> (
-- B is a 1 x n matrix of homogeneous polynomials
-- L is a list of degrees
trim ideal(B * random(source B, (ring B)^(-L)))
)
regSeq = method(TypicalValue => Ideal)
regSeq (Sequence, Ring) := Ideal => (L,S)-> regSeq(toList L, S)
regSeq (List, Ring) := Ideal => (L,S)->(
--forms an ideal generated by powers of the variables.
--L=list of NN. uses the initial subsequence of L as powers
ideal for m to min(#L,rank source vars S)-1 list S_m^(L_m))
randomElementsFromIdeal = method(TypicalValue => Ideal)
randomElementsFromIdeal(List, Ideal) := Ideal => (L,I)->(
trim ideal((gens I)*random(source gens I, (ring I)^(-L))))
----------From Shelling
testNewSimplex = method()
testNewSimplex(List, List) := (P, D) ->(
--given a pure, d-dimensional simplicial complex (sc) as a list of ordered lists of d+1 vertices in [n], and
--a simplex D as such a list, tests whether the intersection of D with P is a union of facets of D.
d := #D-1; --dimension
ints := apply(P, D' -> intersectLists(D',D));
facets := apply(unique select(ints, E -> #E==d),set);
antiFacets := apply(facets,F -> (D-F)#0);
if facets == {} then return false;
smalls := unique select(ints, E -> #E any(antiFacets, v -> not member(v,e)))
)
intersectLists = (D',D) -> D - set(D-set D')
randomSubset = (n,m) -> (
L := new MutableList from toList (0..m-1);
for i from m to n-1 do (
j := random(i+1);
if j < m then L#j = i;
);
sort toList L
)
randomAddition = method()
randomAddition(ZZ,ZZ,List) := (n,m,P) ->(
if #P == 0 then return {randomSubset(n,m+1)};
Plarge := select(P, D-> #D >= m+1); -- the facets big enough to be glued to
if #Plarge == 0 then error "m is too large";
t := false;
D' := {null};
D := Plarge#(random(#Plarge)); -- a random facet from Plarge
compD := toList(0..n-1) - set D;
count := 0;
while not t and count < 20 do (
i := random (#compD);
J := randomSubset(#D,#D-m);
D' = sort(D - set apply(J, j->D_j) | {compD_i});
t = (testNewSimplex(P,D'));
count = count+1);
if count == 20 then return P;
unique (P|{D'})
)
listsToMonomials = (P,R) -> apply(P, D -> product apply(D,d->R_d))
squareFreeMonomialsToLists = (L,R)->(
varToIndex := i->position(gens R, j->j==i);
apply (L/support, j->j/varToIndex))
///
S = ZZ/101[a,b,c]
L = flatten entries matrix"ab,a2,abc2"
varToIndex b
viewHelp position
///
randomAddition(Ring,ZZ,List) := (R,m,L) -> (
P := squareFreeMonomialsToLists(L,R);
listsToMonomials(randomAddition(numgens R,m,P),R)
)
idealFromShelling = method()
idealFromShelling (Ring,List) := (S,P) -> (
Delta := toList (0..numgens S - 1);
V := vars S;
monomialIdeal intersect apply(P, D -> monomialIdeal {V_(Delta - set D)})
)
idealFromShelling List := P -> (
n := (max flatten P)+1;
x := symbol x;
S := QQ[x_0..x_(n-1)];
idealFromShelling(S,P)
)
///
R = ZZ/101[x_0..x_4];
I = randomShellableIdeal(R,2,6)
S
///
idealChainFromShelling = method()
idealChainFromShelling(Ring,List) := (S,P) -> toList apply(#P,i->idealFromShelling(S,take(P,i+1)))
idealChainFromShelling List := P -> toList apply(#P,i->idealFromShelling(take(P,i+1)))
isShelling = method()
isShelling(List) := P -> all(#P, i-> i==0 or testNewSimplex(take(P,i),P#i))
randomShelling = method()
-- random chain of shellable complexes on n vertices, with pure dim m, up to the complete m skeleton
randomShelling(ZZ,ZZ) := (n,m) -> randomShelling(n,m,binomial(n,m+1))
-- random chain of shellable complexes on n vertices, with pure dim m, and k facets
--Should we change the following to start with {{0..m}, {0..m-1,m} to diminish autos?
randomShelling(ZZ,ZZ,ZZ) := (n,m,k) -> (
if k > binomial(n,m+1) then error "k is too large";
P := {};
while #P < k do P = randomAddition(n,m,P);
P
)
randomShellableIdeal=method()
randomShellableIdeal(Ring,ZZ,ZZ) := (R,dimProj,deg) -> (
idealFromShelling(R,randomShelling(numgens R ,dimProj, deg))
)
randomShellableIdealChain=method()
randomShellableIdealChain(Ring,ZZ,ZZ) := (R,dimProj,deg)->(
idealChainFromShelling(R,randomShelling(numgens R,dimProj,deg))
)
randomShellableIdealChain(Ring,ZZ) := (R,dimProj)->(
idealChainFromShelling(R,randomShelling(numgens R,dimProj))
)
///
S = ZZ/101[x_0..x_5]
I = randomShellableIdeal(S,2,5)
dim I == 3
degree I == 5
S = ZZ/101[x_0..x_4]
I = randomShellableIdeal(S,2,6)
///
randomShelling(Ring,ZZ,ZZ) := (R,m,k) -> listsToMonomials(randomShelling(numgens R,m,k),R)
randomShelling(Ring,ZZ) := (R,m) -> listsToMonomials(randomShelling(numgens R,m),R)
randomEdgeIdeal = method()
randomEdgeIdeal(ZZ,ZZ) := (n,t) -> (
needsPackage "EdgeIdeals";
needsPackage "BinomialEdgeIdeals";
x:=symbol x;
G:=randomGraph(QQ[x_1..x_n],t);
(G, edgeIdeal(G))
)
TEST///
randomEdgeIdeal(8, 5)
randomBinomialEdgeIdeal(7, 4)
randomToricEdgeIdeal(6,10)
///
--Random binomial edge ideal
--n is number of vertices, t is number of edges of the graph
randomBinomialEdgeIdeal = method();
randomBinomialEdgeIdeal(ZZ,ZZ) := (n, t) -> (
needsPackage "EdgeIdeals";
needsPackage "BinomialEdgeIdeals";
x := symbol x;
G := randomGraph(QQ[x_1..x_n], t);
E := apply(edges G, i -> apply(i, j -> index j+1));
return (binomialEdgeIdeal(E), G)
)
--Random toric edge ideal
--n is number of variables, t is number of edges of the graph
randomToricEdgeIdeal = method();
randomToricEdgeIdeal(ZZ,ZZ) := (n, t) -> (
needsPackage "BinomialEdgeIdeals";
needsPackage "EdgeIdeals";
e := local e;
x := local x;
R := QQ[x_1..x_n];
S := QQ[e_1..e_t];
G := randomGraph(R,t);
E := apply(edges G, product);
(ker map(R,S,E), G)
)
------------------------------------------------------------
-- DOCUMENTATION RandomIdeals -- documentation
------------------------------------------------------------
beginDocumentation()
doc ///
Key
RandomIdeals
Headline
A package to construct various sorts of random ideals
Description
Text
This package can be used to make experiments, trying many ideals, perhaps
over small fields. For example...what would you expect the regularities of
"typical" monomial ideals with 10 generators of degree 3 in 6 variables to be?
Try a bunch of examples -- it's fast.
Here we do only 500 -- this takes about a second on a fast machine --
but with a little patience, thousands can be done conveniently.
Example
setRandomSeed(currentTime())
kk=ZZ/101;
S=kk[vars(0..5)];
time tally for n from 1 to 500 list regularity randomMonomialIdeal(10:3,S)
Text
How does this compare with the case of binomial ideals? or pure binomial ideals?
We invite the reader to experiment, replacing "randomMonomialIdeal" above with
"randomBinomialIdeal" or "randomPureBinomialIdeal", or taking larger numbers
of examples. Click the link "Finding Extreme Examples" below to see
some other, more elaborate ways to search.
SeeAlso
"Finding Extreme Examples"
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomSquareFreeStep
randomBinomialIdeal
randomPureBinomialIdeal
randomSparseIdeal
randomElementsFromIdeal
randomMonomial
randomShellableIdeal
randomShellableIdealChain
randomShelling
///
------------------------------------------------------------
-- DOCUMENTATION randomShelling
------------------------------------------------------------
doc ///
Key
randomShelling
(randomShelling,ZZ,ZZ)
(randomShelling,ZZ,ZZ,ZZ)
(randomShelling,Ring,ZZ)
(randomShelling,Ring,ZZ,ZZ)
Headline
produces a random chain of shellable complexes
Usage
P=randomShelling(n,m)
P=randomShelling(n,m,k)
P=randomShelling(R,m)
P=randomShelling(R,m,k)
Inputs
n:ZZ
the number of vertices
R:Ring
a polynomial ring with a variable for each vertex
m:ZZ
the dimension of the facets
k:ZZ
the number of facets (if omitted, the number will be {\tt n} choose {\tt m+1})
Outputs
P:List
A list of lists of integers. Each list of integers is a facet of the complex and the order is a shelling. If called with a Ring {\tt R} instead of an integer {\tt n}, each facet is represented by a square-free monomial instead of a list.
Description
Text
The function produces a list of facets of a random shellable simplicial complex.
The order of the facets is a shelling.
Text
The alogorithm works by choosing one of the previous facets at random, and replacing one of its vertices with a new vertex chosen at random.
If the choice meets the criteria of a shelling, that facet is added to list, otherwise it is discarded and the algorithm tries again.
The first facet is chosen uniformly at random.
The call randomShelling(n,m) produces a *complete* chain -- that is, a shelling
of the m-skeleton of the (n-1)-simplex, with the simplices listed in order,
so that any initial subsequence of length d gives a (random) shellable simplicial
complex with d facets.
The probability distribution for this random selection
is presumably not the uniform one; it would be nice to write a reversible
markov chain that could be used with the Metropolis algorithm to produce
the uniform distribution, as is done in randomSquareFreeStep, and the
randomSquareFreeMonomialIdeal codes
Example
P = randomShelling(6,3,10)
Q = randomShelling(6,3)
Caveat
No claim is made on the distribution of the random chain.
SeeAlso
randomAddition
idealChainFromShelling
idealFromShelling
randomShellableIdeal
randomSquareFreeStep
randomSquareFreeMonomialIdeal
///
------------------------------------------------------------
-- DOCUMENTATION isShelling
------------------------------------------------------------
doc ///
Key
isShelling
(isShelling,List)
Headline
determines whether a list represents a shelling of a simplicial complex.
Usage
b = isShelling(P)
Inputs
P:List
A list of lists of integers. Each list of integers is a facet of the complex and the order is a possible shelling.
Outputs
b:Boolean
true if and only if P is a shelling.
Description
Text
An ordering $F_1,..F_d$ of the facets of a simplicial complex $P$ is shellable
if $(F_1 \cup .. \cup F_{k-1}) \cap F_k$ is pure of dim$F_k -1$ for all $k = 2,..,d$.
Determines if a list of faces is a shelling order of the simplicial complex.
Example
P = {{1, 2, 3}, {1, 2, 5}};
isShelling(P)
Q = {{1,2,3},{3,4,5},{2,3,4}};
isShelling(Q)
///
------------------------------------------------------------
-- DOCUMENTATION randomAddition
------------------------------------------------------------
doc ///
Key
randomAddition
(randomAddition,ZZ,ZZ,List)
(randomAddition,Ring,ZZ,List)
Headline
Adds a random facet to a shellable complex
Usage
p=randomAddition(n,m,P)
p=randomAddition(R,m,P)
Inputs
n:ZZ
the number of vertices (if a ring is specified, {\tt n} is the number of variables.
m:ZZ
the dimension of the new facet
P:List
A list of lists of integers. Each list of integers is a facet of the complex and the order is a shelling.
R:Ring
A polynomial ring.
Outputs
p:List
A list of lists of integers. Each list of integers is a facet of the complex and the order is a shelling.
Description
Text
This function randomly chooses a facet of size {\tt m+1} and checks whether the facet can be shellably added to the shelling.
If it can be shellably added to the shelling, it is added to the shelling and the new shelling is returned.
Otherwise, the process repeats up to 20 times.
Text
This function can be used to randomly construct non-pure shellable complexes. A new {\tt m}-simplex can only be
glued to previous simplices of dimension at least {\tt m}. If all previous simplices are smaller, then the addition will fail.
Example
P={{1,2,3}}
P=randomAddition(6,2,P)
P=randomAddition(6,1,P)
Caveat
If the input is not a shellable simplicial complex, the new complex will not be shellable. The function does not check whether the input is shellable.
SeeAlso
randomShelling
idealChainFromShelling
idealFromShelling
///
------------------------------------------------------------
-- DOCUMENTATION idealFromShelling
------------------------------------------------------------
doc ///
Key
idealFromShelling
(idealFromShelling,List)
(idealFromShelling,Ring,List)
Headline
Produces an ideal from a shelling
Usage
I = idealFromShelling(P)
I = idealFromShelling(S,P)
Inputs
S:Ring
(If omitted, it will use {\tt S=QQ[x_0..x_{n-1}]} where {\tt n} is the maximum integer in the lists of {\tt P}.
P:List
A list of lists of integers. Each list of integers is a facet of the complex and the order is a shelling.
S:Ring
(If omitted, it will use {\tt S=QQ[x_0..x_{n-1}]} where {\tt n} is the maximum integer in the lists of {\tt P}.
Outputs
I:Ideal
generated by the monomials representing the minimal nonfaces of {\tt P}
Description
Text
This gives the Stanley-Reisner ideal for the simplicial complex, that is the ideal generated by the monomials representing the minimal nonfaces of {\tt P}.
Example
S = QQ[x_0,x_1,x_2,x_3,x_4]
P = {{1, 2, 4}, {0, 1, 4}, {0, 2, 4}, {0, 3, 4}};
idealFromShelling(S,P)
SeeAlso
idealChainFromShelling
randomShellableIdeal
///
------------------------------------------------------------
-- DOCUMENTATION idealChainFromShelling
------------------------------------------------------------
doc ///
Key
idealChainFromShelling
(idealChainFromShelling,List)
(idealChainFromShelling,Ring,List)
Headline
Produces chains of ideals from a shelling.
Usage
L = idealChainFromShelling(P)
L = idealChainFromShelling(R,P)
Inputs
R:Ring
Polynomial ring
P:List
A (possibly impure) shelled simplicial complex, represented by a
list of lists of integers.
Each list of integers is a facet of the
complex and the order is a shelling. If the ring R is specified, the output
is a list of ideals in R; else it is a list of ideals in
QQ[x_0..x_{n-1}], where n is the maximum number of elements in one of the lists
of integers
Outputs
L:List
a list of ideals
Description
Text
Outputs the Stanley-Reisner ideal for each successive simplicial complex formed by truncating the shelling.
Example
P = {{1, 2, 4}, {0, 1, 4}, {0, 2, 4}, {0, 3, 4}};
idealChainFromShelling(P)
SeeAlso
idealFromShelling
randomShellableIdealChain
///
------------------------------------------------------------
-- DOCUMENTATION randomShellableIdeal
------------------------------------------------------------
doc ///
Key
randomShellableIdeal
(randomShellableIdeal,Ring,ZZ,ZZ)
Headline
Produces a ideal from a random shellable simplicial complex
Usage
I = randomShellableIdeal(R,m,k)
Inputs
R:Ring
a polynomial ring
m:ZZ
dimension of facets in shellable complex
k:ZZ
the degree of the shellable complex
Outputs
I:MonomialIdeal
the Stanley-Reisner ideal of a random shellable complex
Description
Text
The Stanley-Reisner ideal of a shellable simplicial complex is always
Cohen-Macaulay; the converse is not true, although, to paraphrase Arnol'd,
square-free monomial ideals that have a serious reason to be Cohen-Macaulay
generally do come from shellable complexes.
The program makes a (Cohen-Macaulay) square-free monomial ideal
from the Stanley-Reisner ideal of a random shellable simplicial complex.
simplicial complex relies on the code for producing random shellable simplicial
complexes; see randomShelling for a description.
Example
R = ZZ/101[x_0..x_4];
I = randomShellableIdeal(R,2,6)
Caveat
No claim is made on the distribution of the ideal.
SeeAlso
randomShelling
idealFromShelling
idealChainFromShelling
randomShellableIdealChain
///
------------------------------------------------------------
-- DOCUMENTATION randomShellableIdealChain
------------------------------------------------------------
doc ///
Key
randomShellableIdealChain
(randomShellableIdealChain,Ring,ZZ,ZZ)
(randomShellableIdealChain,Ring,ZZ)
Headline
Produces a chain of ideals from a random shelling
Usage
L = randomShellableIdealChain(R,m,k)
L = randomShellableIdealChain(R,m)
Inputs
R:Ring
a polynomial ring
m:ZZ
dimension of the facets in the shellable complex
k:ZZ
the degree of the smallest ideal
Outputs
L:List
list of Stanley-Riesner ideals of the simplicial comlexes of the truncations of the shelling.
Description
Text
Example
R = ZZ/101[x_0..x_3];
L = randomShellableIdealChain(R,1)
Caveat
No claim is made on the distribution of the ideal.
SeeAlso
randomShelling
idealFromShelling
idealChainFromShelling
randomShellableIdeal
///
TEST///
setRandomSeed 0
S = ZZ/101[a,b,c,d,e]
I = randomShellableIdeal(S,2,3)
I == monomialIdeal (a, c*d*e)
///
TEST///
assert(#randomShelling(5,2,6)==6)
assert(#randomShelling(5,2)==binomial(5,3))
R=QQ[x1,x2,x3,x4,x5];
assert(#randomShelling(R,2,6)==6)
///
TEST///
assert(isShelling({}))
assert(isShelling({{1,2,3}}))
assert(isShelling({{1,2,3},{2,3,4}}))
assert(isShelling(randomShelling(5,3,5)))
--non pure shellings
assert(isShelling({{1,2,3},{2,4}}))
assert(isShelling({{1},{2}}))
assert(not isShelling({{1,3},{2,4}}))
assert(isShelling({{1,2},{3}}))
assert(not isShelling({{3},{1,2}}))
///
TEST///
setRandomSeed(0);
assert(#randomAddition(6,2,{{1,2,3}})==2)
assert(#randomAddition(6,3,{{1,2,3,4}})==2)
///
TEST///
needsPackage "SimplicialComplexes"
needsPackage "SimplicialDecomposability"
R=QQ[x1,x2,x3,x4,x5];
assert(isShellable simplicialComplex randomShelling(R,2,6))
///
beginDocumentation()
doc ///
Key
randomMonomial
(randomMonomial, ZZ, Ring)
Headline
Choose a random monomial of given degree in a given ring
Usage
m = randomMonomial(d,S)
Inputs
d: ZZ
non-negative
S: Ring
polynomial ring
Outputs
m: RingElement
monomial of S
Description
Text
Chooses a random monomial.
Example
setRandomSeed(currentTime())
kk=ZZ/101
S=kk[a,b,c]
randomMonomial(3,S)
SeeAlso
randomMonomialIdeal
randomSquareFreeMonomialIdeal
///
doc ///
Key
randomSquareFreeMonomialIdeal
(randomSquareFreeMonomialIdeal, List, Ring)
(randomSquareFreeMonomialIdeal, Sequence, Ring)
Headline
random square-free monomial ideal with given degree generators
Usage
I = randomSquareFreeMonomialIdeal(L,S)
Inputs
L:List
or sequence of non-negative integers
S:Ring
Polynomial ring
Outputs
I:Ideal
square-free monomial ideal with generators of specified degrees
Description
Text
Choose a random square-free monomial
ideal whose generators have the degrees
specified by the list or squence L.
Example
setRandomSeed(currentTime())
kk=ZZ/101
S=kk[a..e]
L={3,5,7}
randomSquareFreeMonomialIdeal(L, S)
randomSquareFreeMonomialIdeal(5:2, S)
Caveat
The ideal is constructed degree by degree, starting from the lowest degree
specified. If there are not enough monomials of the next specified degree that
are not already in the ideal, the function prints a warning and returns an ideal
containing all the generators of that degree.
SeeAlso
randomMonomial
randomMonomialIdeal
///
doc ///
Key
randomSquareFreeStep
(randomSquareFreeStep, MonomialIdeal)
(randomSquareFreeStep, Ideal)
(randomSquareFreeStep, List)
[randomSquareFreeStep,AlexanderProbability]
Headline
A step in a random walk with uniform distribution over all monomial ideals
Usage
M = randomSquareFreeStep(I)
M = randomSquareFreeStep(I, AlexanderProbability => p)
M = randomSquareFreeStep(L)
M = randomSquareFreeStep(L, AlexanderProbability => p)
Inputs
I:Ideal
square-free monomial Ideal or MonomialIdeal
L:List
{I,Igens,ISocgens} where I is a square-free MonomialIdeal,
Igens is a List of its minimal generators,
ISocgens is a List of the minimal generators of the socle mod I.
Outputs
M:List
{J,Jgens,JSocgens} where J is a square-free MonomialIdeal,
Jgens is a List of its minimal generators,
JSocgens is a List of the minimal generators of the socle mod J.
Description
Text
With probability p the routine takes the Alexander dual of I;
the default value of p is .05, and it can be set with the option
AlexanderProbility.
Otherwise uses the Metropolis algorithm to produce a random walk on the space
of square-free ideals. Note that there are a LOT of square-free ideals;
these are the Dedekind numbers, and the sequence (with 1,2,3,4,5,6,7,8 variables)
begins
3,6,20,168,7581, 7828354, 2414682040998, 56130437228687557907788.
(see the Online Encyclopedia of Integer Sequences for more information).
Given I in a polynomial ring S, we make a list
ISocgens of the square-free minimal monomial generators of the socle of S/(squares+I)
and a list of minimal generators Igens of I. A candidate "next" ideal J is formed as follows:
We choose randomly from the
union of these lists; if a socle element is chosen, it's added to I; if
a minimal generator is chosen, it's replaced by the square-free part of
the maximal ideal times it.
the chance of making the given move is then 1/(#ISocgens+#Igens), and
the chance of making the move back would be the similar quantity for J,
so we make the move or not depending on whether
random RR < (nJ+nSocJ)/(nI+nSocI) or not; here random RR is
a random number in [0,1].
Example
setRandomSeed(currentTime())
S=ZZ/2[vars(0..3)]
J = monomialIdeal"ab,ad, bcd"
randomSquareFreeStep J
Text
With 4 variables and 168 possible monomial ideals, a run of 5000
takes less than 6 seconds on a reasonably fast machine. With
10 variables a run of 1000 takes about 2 seconds.
Example
setRandomSeed(1)
rsfs = randomSquareFreeStep
J = monomialIdeal 0_S
time T=tally for t from 1 to 5000 list first (J=rsfs(J,AlexanderProbability => .01));
#T
T
J
///
doc ///
Key
AlexanderProbability
Headline
option to randomSquareFreeStep
Usage
M = randomSquareFreeStep(L, AlexanderProbability => p)
Inputs
p: RR
real number between 0 and 1.
Description
Text
Controls how often the Alexander dual is taken
SeeAlso
randomSquareFreeStep
///
doc ///
Key
squareFree
Headline
ideal of all square-free monomials of given degree
Usage
I = squareFree(d,S)
Inputs
d:ZZ
positive
S:Ring
Polynomial ring
Outputs
I:Ideal
all square-free monomials of degree d
Description
Example
kk=ZZ/101
S=kk[a..e]
squareFree(3, S)
SeeAlso
randomSquareFreeMonomialIdeal
///
doc ///
Key
(squareFree, ZZ, Ring)
Headline
ideal of all square-free monomials of given degree
Usage
I = squareFree(d,S)
Inputs
d:ZZ
positive
S:Ring
Polynomial ring
Outputs
I:Ideal
all square-free monomials of degree d
Description
Example
kk=ZZ/101
S=kk[a..e]
squareFree(3, S)
SeeAlso
randomSquareFreeMonomialIdeal
///
doc ///
Key
regSeq
(regSeq, List, Ring)
(regSeq, Sequence, Ring)
Headline
regular sequence of powers of the variables, in given degrees
Usage
I = regSeq(L,S)
Inputs
L:List
or sequence of positive integers
S:Ring
Polynomial ring
Outputs
I:Ideal
generated by the given powers of the variables
Description
Example
kk=ZZ/101
S=kk[a..e]
regSeq((2,3,4),S)
Caveat
If the number of elements of L differs from the number of
variables in the ring, the length of the regular sequence
will be the minimum of the two.
///
doc ///
Key
randomIdeal
(randomIdeal, List, Matrix)
(randomIdeal, Sequence, Matrix)
Headline
randomIdeal made from a given set of monomials
Usage
I = randomIdeal(L,m)
Inputs
L:List
or sequence of positive integers
m: Matrix
1xn matrix of homogeneous polynomials in a ring S
Outputs
I:Ideal
generated by random linear combinations of degrees given by L of the given polynomials
Description
Text
This function composes m with a random map from a free module with degrees
specified by L to the source of m.
Example
kk=ZZ/101
S=kk[a..e]
L={3,3,4,6}
m = matrix{{a^3,b^4+c^4,d^5}}
I=randomIdeal(L,m)
SeeAlso
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomBinomialIdeal
randomPureBinomialIdeal
randomElementsFromIdeal
///
doc ///
Key
randomBinomialIdeal
(randomBinomialIdeal, List, Ring)
(randomBinomialIdeal, Sequence, Ring)
Headline
randomBinomialIdeal with binomials of given degrees
Usage
I = randomBinomialIdeal(L,S)
Inputs
L:List
or sequence of positive integers
S: Ring
Polynomial ring
Outputs
I:Ideal
generated by random binomials of the given degrees
Description
Example
kk=ZZ/101
S=kk[a..e]
L={3,3,4,6}
I=randomBinomialIdeal(L,S)
Caveat
The binomials are generated one at a time, and there is no checking to
see whether the ideal returned is minally generated by fewer elements,
so the number of minimal generators may not be what you expect.
SeeAlso
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomPureBinomialIdeal
randomElementsFromIdeal
///
doc ///
Key
randomPureBinomialIdeal
(randomPureBinomialIdeal, List, Ring)
(randomPureBinomialIdeal, Sequence, Ring)
Headline
randomPureBinomialIdeal with binomials of given degrees
Usage
I = randomPureBinomialIdeal(L,S)
Inputs
L:List
or sequence of positive integers
S: Ring
Polynomial ring
Outputs
I:Ideal
generated by random pure binomials (that is, differences of monomials without coefficients) of the given degrees
Description
Text
Example
kk=ZZ/101
S=kk[a..e]
L={3,3,4,6}
I=randomPureBinomialIdeal(L,S)
Caveat
The binomials are generated one at a time, and there is no checking to
see whether the ideal returned is minally generated by fewer elements,
so the number of minimal generators may not be what you expect.
SeeAlso
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomBinomialIdeal
randomElementsFromIdeal
///
doc ///
Key
randomSparseIdeal
(randomSparseIdeal, Matrix, ZZ, ZZ)
Headline
randomSparseIdeal made from a given set of monomials
Usage
I = randomSparseIdeal(B,s,n)
Inputs
B:Matrix
1xn matrix of monomials
s: ZZ
positive integer, the number of terms in the generators of I
n: ZZ
positive integer, the number of generators of I
Outputs
I:Ideal
generated by n polynomials, each a random linear combination of s monomials
Description
Text
Each generator of I is formed by randomly choosing s (the sparsity) entries
of the matrix B and taking a random linear combinations with coefficients in
the (ultimate) coefficient ring of S, the ring in which the monomials lie.
Example
kk=ZZ/101
S=kk[a..e]
L={3,3,4,6}
B = matrix{{a^3,b^4,d^5,a*b*c,e}}
I=randomSparseIdeal(B,3,2)
SeeAlso
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomBinomialIdeal
randomPureBinomialIdeal
randomElementsFromIdeal
///
doc ///
Key
randomElementsFromIdeal
(randomElementsFromIdeal,List, Ideal)
Headline
Chooses random elements of given degrees in a given ideal.
Usage
I = randomElementsFromIdeal(L,I)
Inputs
L:List
of integers
I:Ideal
that should be homogeneous
Outputs
I:Ideal
generated by (at most) n homogeneous polynomials that are random linear combination of the
generators of I, with degrees specified by the list L
Description
Example
kk=ZZ/101
S=kk[a..c]
L={3,3,4,6}
I = ideal(a^3,b^3, c^3)
J=randomElementsFromIdeal(L,I)
SeeAlso
"Finding Extreme Examples"
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomBinomialIdeal
randomPureBinomialIdeal
///
doc ///
Key
randomMonomialIdeal
(randomMonomialIdeal, List, Ring)
(randomMonomialIdeal, Sequence, Ring)
Headline
random monomial ideal with given degree generators
Usage
I = randomMonomialIdeal(L,S)
Inputs
L:List
or sequence of non-negative integers
S:Ring
Polynomial ring
Outputs
I:Ideal
monomial ideal with generators of specified degrees
Description
Text
Choose a random ideal whose generators have the degrees specified by the list or squence L.
Example
kk=ZZ/101
S=kk[a..e]
L={3,5,7}
randomMonomialIdeal(L, S)
randomMonomialIdeal(5:2, S)
Caveat
The ideal is constructed degree by degree, starting from the lowest degree
specified. If there are not enough monomials of the next specified degree that
are not already in the ideal, the function prints a warning and returns an ideal
containing all the generators of that degree.
SeeAlso
randomMonomial
randomSquareFreeMonomialIdeal
///
doc ///
Key
"Finding Extreme Examples"
Headline
Ways to use random ideals to search for (counter)-examples
Description
Text
A common use of Macaulay2 is to look for extreme or particularly
interesting examples. Here are some examples of how this may be done.
Supposing first that some space of examples is finite; for
example, we might be interested in monomial ideals with a certain
number of generators of a certain degree d. Suppose, to be concrete,
that we want to compare the
maximum degree of a first syzygy with the regularity of the ideal,
and also with the maximum degree of the last syzygy. (To make the
comparison interesting, it seems reasonable to subtract i from the maximum
degree of an i-th syzygy.)
Text
We may have no idea where to look for extreme examples, and it seems
that examples with small numbers of variables and generators may not
show the range of phenomena that actually occur. In large degree there
may be too many examples to search systematically; so instead we may
choose many examples at random, and hope to see a pattern.
Here is a simple example
First we tally the projective dimensions of 500
random square-free monomial ideals (what's the average?), then
looking how big the difference between the regularity of R/I and the
"relation degree"-2 can be. It turns out this the differences are rather
small, only 1 in a typical run of 5000. So one might look for ideals with
a difference of 2, as in the following (in a real run, one would make
the number of iterations much bigger; here we keep it small so
that Macaulay2 doesn't take too long to build it's documentation files.)
Example
kk=ZZ/101
S=kk[vars(0..5)]
L=for n from 1 to 100 list res randomSquareFreeMonomialIdeal(10:3,S);
tally apply(L, F -> length F)
tally apply(L, F -> regularity F - ((max flatten degrees F_2) - 2))
L=for n from 1 to 500 list res randomSquareFreeMonomialIdeal(10:3,S);
scan(L, F -> if 1<(regularity F - (max flatten degrees F_2) + 2) then print F.dd_1)
Text
A typical problem might be to find how high the regularity of R/I can
be when R has reasonably few variables, and the degrees of the generators of
I are reasonably small; despite the wild examples of Mayr-Mayer, we don't
know how to make examples with large regularity without letting the
number of variables become large. The following program computes
"rep" examples of random ideals with monomial and binomial generators,
and prints any whose regularity exceeds the number "bound"
looper = (rep,bound, L1, L2) -> (for i from 1 to rep do (
if i % 1000 == 0 then << "." << flush;
J := randomMonomialIdeal(L1,S) + randomBinomialIdeal(L2,S);
m := regularity coker gens J;
if m >= bound then << "reg " << m << " " << toString J << endl;))
For example:
kk=ZZ/2
S=kk[a,b,c,d]
looper(30000,10,{4},{4,4}) -- finds examples with on monomial of degree 4
and 2 binomials of degree 4. The largest largest regularity it has found
(and the largest I know for an ideal in 4 variables of degree 4) is 14.
Here is an example it found:
ideal(a*b^3,a^4+b^4,b*c^3+a*d^3)
Similarly:
looper(30000,10,{4,4},{4}) -- looks for examples with
2 monomials and 1 binomial of degree 4. Suggestively, the
largest regularity it found was also 14:
betti res ideal(c^4,b^4,a^3*c+b*d^3) -- reg 14
Text
A more sophisticated and difficult situation arises when the space
of examples is not necessarily finite (except over a finite field) but
is a unirational
variety (such as the space of ideals generated by (at most) a certain
number of forms of certain given degrees, or the space of smooth curves
of genus g for some g <= 14) one may be able to do a search for random
examples, taking a rational parametrization of the space of examples
and plugging in random inputs.
If the "interesting" examples live in
a subvariety whose codimension is small, then, working over a small field
(say 2,3, or 5 elements) one might hope to see elements of the subvariety
"not too rarely". This principle has been used to good effect for example
by (Caviglia and Decker-Schreyer, ****--Schreyer).
SeeAlso
randomIdeal
randomMonomialIdeal
randomSquareFreeMonomialIdeal
randomMonomial
randomBinomialIdeal
randomPureBinomialIdeal
///
--Documentation
doc ///
Key
randomEdgeIdeal
(randomEdgeIdeal, ZZ, ZZ)
Headline
Creates an edge ideal from a random graph with n vertices and t edges.
Usage
(I,G) = randomEdgeIdeal(n,t)
Inputs
n:ZZ
number of vertices
t:ZZ
number of edges
Outputs
I:Ideal
a random edge ideal
G:Graph
the graph underlying I
Description
Text
The edge ideal of a graph is the quadratic monomial ideal generated by x_v*x_w for all edges (v,w) in the graph.
This method returns the edge ideal {\tt I} of a random graph {\tt G} which has n vertices and t edges.
Example
randomEdgeIdeal(7, 4)
SeeAlso
randomGraph
edgeIdeal
///
doc ///
Key
randomBinomialEdgeIdeal
(randomBinomialEdgeIdeal, ZZ, ZZ)
Headline
Creates a binomial edge ideal from a random graph with n vertices and t edges.
Usage
(I,G) = randomBinomialEdgeIdeal(n,t)
Inputs
n:ZZ
number of vertices
t:ZZ
number of edges
Outputs
I:Ideal
a random binomial edge ideal
G:Graph
the graph underlying I
Description
Text
The binomial edge ideal associated to a graph G is the quadratic binomial ideal generated by the set containinig x_v*y_w-x_w-y_v for every edge (v,w) in G.
This method returns the binomial edge ideal {\tt I} of a random graph {\tt G} which has n vertices and t edges.
Example
randomBinomialEdgeIdeal(7, 4)
SeeAlso
randomGraph
binomialEdgeIdeal
randomToricEdgeIdeal
///
doc ///
Key
randomToricEdgeIdeal
(randomToricEdgeIdeal, ZZ, ZZ)
Headline
Creates a toric edge ideal from a random graph with n vertices and t edges.
Usage
(I, G) = randomToricEdgeIdeal(n,t)
Inputs
n:ZZ
number of vertices
t:ZZ
number of edges
Outputs
I:Ideal
a random toric edge ideal
G:Graph
the graph underlying I
Description
Text
The toric edge ideal of a graph G is the kernel of the map from the polynomial ring k[edges(G)] to the polynomial ring k[vertices G] taking an x_e to y_i*y_j, where e = (i,j).
This method returns the toric edge ideal {\tt I} of a random graph {\tt G} which has n vertices and t edges. {\tt I} is the kernel of the homomorphism from QQ[x_1..x_n] to QQ/101[e_1..e_t] which sends each vertex in the graph {\tt G} to the product of its endpoints.
Example
randomToricEdgeIdeal(4,5)
Text
Note that his is different than the randomBinomialEdgeIdeal!
Example
randomBinomialEdgeIdeal(4,5)
SeeAlso
randomGraph
randomBinomialEdgeIdeal
///
TEST ///
S=ZZ/101[a..e]
setRandomSeed 123456
assert (randomMonomial(7,S)==a*b^3*c^3)
setRandomSeed 123456
assert(randomMonomialIdeal({3,4,5}, S)==ideal(d*e^2,a*b*d^2,b*c^3*e))
setRandomSeed 123456
assert(randomSquareFreeMonomialIdeal({6,4,4},S)==ideal(a*b*c*e,a*c*d*e))
setRandomSeed 123456
assert(ideal(8*a^2+5*a*b+4*b^2+35*b*c+3*b*d+36*b*e,29*a^2+22*a*b+32*b^2-44*b*c-6*b*d+40*b*e) == randomIdeal({2,2},matrix{{a^2,b}}))
setRandomSeed 123456
assert(ideal(-29*b^2+b*e,-4*a^2+b*c,45*a*d) == randomBinomialIdeal({2,2,2},S))
setRandomSeed 123456
assert(ideal(b*e-d*e,b^2-b*c,-a^2+a*e)== randomPureBinomialIdeal({2,2,2}, S))
setRandomSeed 123456
assert(randomSparseIdeal(matrix"a2,ab,b2", 2,2)==ideal(8*a*b+5*b^2,4*a^2+35*a*b))
assert(ideal(a*b,a*c,a*d,a*e,b*c,b*d,b*e,c*d,c*e,d*e)==squareFree(2,S))
assert( regSeq((1,2,3,4,5,6), S)==ideal(a,b^2,c^3,d^4,e^5))
setRandomSeed 123456
assert(degrees randomElementsFromIdeal({2,3,6},ideal"a2,ab,c5") == {{2}, {3}, {6}})
S=ZZ/2[a,b]
setRandomSeed 1
--assert(prepare monomialIdeal(a^2, a*b)=={monomialIdeal (a^2 , a*b), {a^2 , a*b}, {0, 1}})
setRandomSeed 1
S=ZZ/2[vars(0..3)]
J = ideal"ab,ad, bcd"
assert( (randomSquareFreeStep J) === {monomialIdeal map((S)^1,(S)^{{-2},{-2}},{{a*b, a*d}}),{a*b,a*d},{b*c*d,a*c}} );
///
end--
restart
loadPackage ("RandomIdeals", Reload =>true)
load "RandomIdeals.m2"
uninstallPackage "RandomIdeals"
restart
installPackage "RandomIdeals"
check "RandomIdeals"
viewHelp RandomIdeals