A quotient of the ring of symmetric functions generalizing quantum cohomology
Darij Grinberg
28 January 2019
Drexel University, Philadelphia, PA
slides: http:
//www.cip.ifi.lmu.de/~grinberg/algebra/drexel2019.pdf paper: http:
//www.cip.ifi.lmu.de/~grinberg/algebra/basisquot.pdf overview: http:
//www.cip.ifi.lmu.de/~grinberg/algebra/fpsac19.pdf
1 / 31
What is this about?
From a modern point of view, Schubert calculus (a.k.a.
classical enumerative geometry, or Hilbert’s 15th problem) is about two cohomology rings:
H∗
Gr (k,n)
| {z }
Grassmannian
and H∗
Fl (n)
| {z }
flag variety
(both varieties over C).
In this talk, we are concerned with the first.
Classical result: as rings, H∗(Gr (k,n))
∼= (symmetric polynomials inx1,x2, . . . ,xk overZ) (hn−k+1,hn−k+2, . . . ,hn)ideal,
where the hi are complete homogeneous symmetric polynomials (to be defined soon).
2 / 31
What is this about?
From a modern point of view, Schubert calculus (a.k.a.
classical enumerative geometry, or Hilbert’s 15th problem) is about two cohomology rings:
H∗
Gr (k,n)
| {z }
Grassmannian
and H∗
Fl (n)
| {z }
flag variety
(both varieties over C).
In this talk, we are concerned with the first.
Classical result: as rings, H∗(Gr (k,n))
∼= (symmetric polynomials inx1,x2, . . . ,xk overZ) (hn−k+1,hn−k+2, . . . ,hn)ideal,
where the hi are complete homogeneous symmetric polynomials (to be defined soon).
2 / 31
What is this about?
From a modern point of view, Schubert calculus (a.k.a.
classical enumerative geometry, or Hilbert’s 15th problem) is about two cohomology rings:
H∗
Gr (k,n)
| {z }
Grassmannian
and H∗
Fl (n)
| {z }
flag variety
(both varieties over C).
In this talk, we are concerned with the first.
Classical result: as rings, H∗(Gr (k,n))
∼= (symmetric polynomials inx1,x2, . . . ,xk overZ) (hn−k+1,hn−k+2, . . . ,hn)ideal,
where the hi are complete homogeneous symmetric polynomials (to be defined soon).
2 / 31
Quantum cohomology of Gr(k,n)
(Small)Quantum cohomology is a deformation of
cohomology from the 1980–90s. For the Grassmannian, it is QH∗(Gr (k,n))
∼= (symmetric polynomials in x1,x2, . . . ,xk overZ[q])
hn−k+1,hn−k+2, . . . ,hn−1,hn+ (−1)kq
ideal. Many properties of classical cohomology still hold here.
In particular: QH∗(Gr (k,n)) has a Z[q]-module basis (sλ)λ∈P
k,n of (projected) Schur polynomials (to be defined soon), with λranging over all partitions with ≤k parts and each part≤n−k. The structure constants are the
Gromov–Witten invariants. References:
Aaron Bertram, Ionut Ciocan-Fontanine, William Fulton, Quantum multiplication of Schur polynomials, 1999.
Alexander Postnikov, Affine approach to quantum Schubert calculus, 2005.
3 / 31
Quantum cohomology of Gr(k,n)
(Small)Quantum cohomology is a deformation of
cohomology from the 1980–90s. For the Grassmannian, it is QH∗(Gr (k,n))
∼= (symmetric polynomials in x1,x2, . . . ,xk overZ[q])
hn−k+1,hn−k+2, . . . ,hn−1,hn+ (−1)kq
ideal. Many properties of classical cohomology still hold here.
In particular: QH∗(Gr (k,n)) has a Z[q]-module basis (sλ)λ∈P
k,n of (projected) Schur polynomials (to be defined soon), with λranging over all partitions with ≤k parts and each part≤n−k. The structure constants are the
Gromov–Witten invariants. References:
Aaron Bertram, Ionut Ciocan-Fontanine, William Fulton, Quantum multiplication of Schur polynomials, 1999.
Alexander Postnikov, Affine approach to quantum Schubert calculus, 2005.
3 / 31
Where are we going?
Goal: Deform H∗(Gr (k,n)) using k parameters instead of one, generalizing QH∗(Gr (k,n)).
The new ring has no geometric interpretation known so far, but various properties suggesting such an interpretation likely exists.
I will now start from scratch and define standard notations around symmetric polynomials, then introduce the deformed cohomology ring algebraically.
4 / 31
Where are we going?
Goal: Deform H∗(Gr (k,n)) using k parameters instead of one, generalizing QH∗(Gr (k,n)).
The new ring has no geometric interpretation known so far, but various properties suggesting such an interpretation likely exists.
I will now start from scratch and define standard notations around symmetric polynomials, then introduce the deformed cohomology ring algebraically.
4 / 31
Where are we going?
Goal: Deform H∗(Gr (k,n)) using k parameters instead of one, generalizing QH∗(Gr (k,n)).
The new ring has no geometric interpretation known so far, but various properties suggesting such an interpretation likely exists.
I will now start from scratch and define standard notations around symmetric polynomials, then introduce the deformed cohomology ring algebraically.
4 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tuple α∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, letxα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial. Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”). Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32
.
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, letxα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial. Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”). Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32
.
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, let xα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial.
Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”). Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32
.
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, let xα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial.
Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”). Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32
.
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, let xα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial.
Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”).
Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32 .
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, let xα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial.
Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”).
Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32 .
5 / 31
A more general setting: P and S Let kbe a commutative ring.
Let N={0,1,2, . . .}. Let k∈N.
Let P =k[x1,x2, . . . ,xk] be the polynomial ring ink indeterminates overk.
For eachk-tupleα ∈Nk and each i ∈ {1,2, . . . ,k}, let αi be thei-th entry ofα. Same for infinite sequences.
For eachα∈Nk, let xα be the monomial x1α1x2α2· · ·xkαk, and let |α|be the degreeα1+α2+· · ·+αk of this monomial.
Let S denote the ring of symmetricpolynomials inP. These are the polynomials f ∈ P satisfying
f (x1,x2, . . . ,xk) =f xσ(1),xσ(2), . . . ,xσ(k) for all permutationsσ of{1,2, . . . ,k}.
Theorem (Artin ≤1944): TheS-moduleP is free with basis (xα)α∈Nk;αi<i for eachi (or, informally: “
x1<1x2<2· · ·xk<k
”).
Example: Fork = 3, this basis is 1,x3,x32,x2,x2x3,x2x32 .
5 / 31
Symmetric polynomials
The ring S of symmetric polynomials inP =k[x1,x2, . . . ,xk] has several bases, usually indexed by certain sets of (integer) partitions.
First, let us recall what partitions are:
6 / 31
k-partitions: definition
A partitionmeans aweakly decreasing sequence of nonnegative integers that has only finitely many nonzero entries.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A partitionmeans aweakly decreasing sequence of nonnegative integers that has only finitely many nonzero entries.
Examples: (4,2,2,0,0,0, . . .) and (3,2,0,0,0,0, . . .) and (5,0,0,0,0,0, . . .) are three partitions.
(2,3,2,0,0,0, . . .) and (2,1,1,1, . . .) are not.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A k-partitionmeans aweakly decreasingk-tuple (λ1, λ2, . . . , λk)∈Nk.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A k-partitionmeans aweakly decreasingk-tuple (λ1, λ2, . . . , λk)∈Nk.
Examples: (4,2,2) and (3,2,0) and (5,0,0) are three 3-partitions.
(2,3,2) is not.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A k-partitionmeans aweakly decreasingk-tuple (λ1, λ2, . . . , λk)∈Nk.
Examples: (4,2,2) and (3,2,0) and (5,0,0) are three 3-partitions.
(2,3,2) is not.
Thus there is a bijection
{k-partitions} → {partitions with at mostk nonzero entries}, λ7→(λ1, λ2, . . . , λk,0,0,0, . . .).
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ) is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A k-partitionmeans aweakly decreasingk-tuple (λ1, λ2, . . . , λk)∈Nk.
Examples: (4,2,2) and (3,2,0) and (5,0,0) are three 3-partitions.
(2,3,2) is not.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
k-partitions: definition
A k-partitionmeans aweakly decreasingk-tuple (λ1, λ2, . . . , λk)∈Nk.
Examples: (4,2,2) and (3,2,0) and (5,0,0) are three 3-partitions.
(2,3,2) is not.
Ifλ∈Nk is a k-partition, then its Young diagram Y(λ)is defined as a table made out of k left-aligned rows, where the i-th row hasλi boxes.
Example: Ifk = 6 andλ= (5,5,3,2,0,0), then
Y (λ) = .
(Empty rows are invisible.)
The same convention applies to partitions.
7 / 31
Symmetric polynomials: the e-basis
For eachm∈Z, we letem denote them-thelementary symmetric polynomial:
em= X
1≤i1<i2<···<im≤k
xi1xi2· · ·xim = X
α∈{0,1}k;
|α|=m
xα ∈ S.
(Thus, e0 = 1, and em= 0 when m<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
eν =eν1eν2· · ·eν` ∈ S.
Theorem (Gauss): The commutativek-algebraS is freely generated by the elementary symmetric polynomials
e1,e2, . . . ,ek. (That is, it is generated by them, and they are algebraically independent.)
Equivalent restatement: (eλ)λis a partition whose entries are≤k
is a basis of the k-module S. Note thatem= 0 when m>k.
8 / 31
Symmetric polynomials: the e-basis
For eachm∈Z, we letem denote them-thelementary symmetric polynomial:
em= X
1≤i1<i2<···<im≤k
xi1xi2· · ·xim = X
α∈{0,1}k;
|α|=m
xα ∈ S.
(Thus, e0 = 1, and em= 0 when m<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
eν =eν1eν2· · ·eν` ∈ S.
Theorem (Gauss): The commutativek-algebraS is freely generated by the elementary symmetric polynomials
e1,e2, . . . ,ek. (That is, it is generated by them, and they are algebraically independent.)
Equivalent restatement: (eλ)λis a partition whose entries are≤k
is a basis of the k-module S. Note thatem= 0 when m>k.
8 / 31
Symmetric polynomials: the e-basis
For eachm∈Z, we letem denote them-thelementary symmetric polynomial:
em= X
1≤i1<i2<···<im≤k
xi1xi2· · ·xim = X
α∈{0,1}k;
|α|=m
xα ∈ S.
(Thus, e0 = 1, and em= 0 when m<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
eν =eν1eν2· · ·eν` ∈ S.
Theorem (Gauss): The commutativek-algebraS is freely generated by the elementary symmetric polynomials
e1,e2, . . . ,ek. (That is, it is generated by them, and they are algebraically independent.)
Equivalent restatement: (eλ)λis a partition whose entries are≤k
is a basis of the k-moduleS.
Note thatem= 0 when m>k.
8 / 31
Symmetric polynomials: the e-basis
For eachm∈Z, we letem denote them-thelementary symmetric polynomial:
em= X
1≤i1<i2<···<im≤k
xi1xi2· · ·xim = X
α∈{0,1}k;
|α|=m
xα ∈ S.
(Thus, e0 = 1, and em= 0 when m<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
eν =eν1eν2· · ·eν` ∈ S.
Theorem (Gauss): The commutativek-algebraS is freely generated by the elementary symmetric polynomials
e1,e2, . . . ,ek. (That is, it is generated by them, and they are algebraically independent.)
Equivalent restatement: (eλ)λis a partition whose entries are≤k
is a basis of the k-moduleS. Note thatem= 0 when m>k.
8 / 31
Symmetric polynomials: the e-basis
For eachm∈Z, we letem denote them-thelementary symmetric polynomial:
em= X
1≤i1<i2<···<im≤k
xi1xi2· · ·xim = X
α∈{0,1}k;
|α|=m
xα ∈ S.
(Thus, e0 = 1, and em= 0 when m<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
eν =eν1eν2· · ·eν` ∈ S.
Theorem (Gauss): The commutativek-algebraS is freely generated by the elementary symmetric polynomials
e1,e2, . . . ,ek. (That is, it is generated by them, and they are algebraically independent.)
Equivalent restatement: (eλ)λis a partition whose entries are≤k
is a basis of the k-moduleS. Note thatem= 0 when m>k.
8 / 31
Symmetric polynomials: the h-bases
For eachm∈Z, we lethm denote them-thcomplete homogeneous symmetric polynomial:
hm = X
1≤i1≤i2≤···≤im≤k
xi1xi2· · ·xim= X
α∈Nk;
|α|=m
xα∈ S.
(Thus, h0 = 1, and hm = 0 whenm<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
hν =hν1hν2· · ·hν` ∈ S.
Theorem: The commutativek-algebra S is freely generated by the complete homogeneous symmetric polynomials h1,h2, . . . ,hk.
Equivalent restatement: (hλ)λis a partition whose entries are≤k
is a basis of the k-module S.
Theorem: (hλ)λis ak-partition is a basis of the k-moduleS. (Another basis!)
9 / 31
Symmetric polynomials: the h-bases
For eachm∈Z, we lethm denote them-thcomplete homogeneous symmetric polynomial:
hm = X
1≤i1≤i2≤···≤im≤k
xi1xi2· · ·xim= X
α∈Nk;
|α|=m
xα∈ S.
(Thus, h0 = 1, and hm = 0 whenm<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
hν =hν1hν2· · ·hν` ∈ S.
Theorem: The commutativek-algebra S is freely generated by the complete homogeneous symmetric polynomials h1,h2, . . . ,hk.
Equivalent restatement: (hλ)λis a partition whose entries are≤k
is a basis of the k-module S.
Theorem: (hλ)λis ak-partition is a basis of the k-moduleS. (Another basis!)
9 / 31
Symmetric polynomials: the h-bases
For eachm∈Z, we lethm denote them-thcomplete homogeneous symmetric polynomial:
hm = X
1≤i1≤i2≤···≤im≤k
xi1xi2· · ·xim= X
α∈Nk;
|α|=m
xα∈ S.
(Thus, h0 = 1, and hm = 0 whenm<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
hν =hν1hν2· · ·hν` ∈ S.
Theorem: The commutativek-algebra S is freely generated by the complete homogeneous symmetric polynomials h1,h2, . . . ,hk.
Equivalent restatement: (hλ)λis a partition whose entries are≤k
is a basis of the k-moduleS.
Theorem: (hλ)λis ak-partition is a basis of the k-moduleS. (Another basis!)
9 / 31
Symmetric polynomials: the h-bases
For eachm∈Z, we lethm denote them-thcomplete homogeneous symmetric polynomial:
hm = X
1≤i1≤i2≤···≤im≤k
xi1xi2· · ·xim= X
α∈Nk;
|α|=m
xα∈ S.
(Thus, h0 = 1, and hm = 0 whenm<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
hν =hν1hν2· · ·hν` ∈ S.
Theorem: The commutativek-algebra S is freely generated by the complete homogeneous symmetric polynomials h1,h2, . . . ,hk.
Equivalent restatement: (hλ)λis a partition whose entries are≤k
is a basis of the k-moduleS.
Theorem: (hλ)λis ak-partition is a basis of the k-moduleS.
(Another basis!)
9 / 31
Symmetric polynomials: the h-bases
For eachm∈Z, we lethm denote them-thcomplete homogeneous symmetric polynomial:
hm = X
1≤i1≤i2≤···≤im≤k
xi1xi2· · ·xim= X
α∈Nk;
|α|=m
xα∈ S.
(Thus, h0 = 1, and hm = 0 whenm<0.)
For eachν = (ν1, ν2, . . . , ν`)∈Z` (e.g., ak-partition when
`=k), set
hν =hν1hν2· · ·hν` ∈ S.
Theorem: The commutativek-algebra S is freely generated by the complete homogeneous symmetric polynomials h1,h2, . . . ,hk.
Equivalent restatement: (hλ)λis a partition whose entries are≤k
is a basis of the k-moduleS.
Theorem: (hλ)λis ak-partition is a basis of the k-moduleS.
(Another basis!)
9 / 31
Symmetric polynomials: the s-basis (Schur polynomials) For eachk-partitionλ, we letsλ be theλ-th Schur polynomial:
sλ= det
xiλj+k−j
1≤i≤k,1≤j≤k
det
xik−j
1≤i≤k,1≤j≤k
(alternant formula)
= det
(hλi−i+j)1≤i≤k,1≤j≤k
(Jacobi-Trudi). Theorem: The equality above holds, and sλ is a symmetric polynomial with nonnegative coefficients.
Theorem: (sλ)λis ak-partition is a basis of the k-moduleS.
10 / 31
Symmetric polynomials: the s-basis (Schur polynomials) For eachk-partitionλ, we letsλ be theλ-th Schur polynomial:
sλ= det
xiλj+k−j
1≤i≤k,1≤j≤k
det
xik−j
1≤i≤k,1≤j≤k
(alternant formula)
= det
(hλi−i+j)1≤i≤k,1≤j≤k
(Jacobi-Trudi). Theorem: The equality above holds, and sλ is a symmetric polynomial with nonnegative coefficients. Explicitly,
sλ = X
Tis a semistandardλ-tableau with entries 1,2,...,k
k
Y
i=1
xi(number ofi’s inT),
where a semistandardλ-tableau with entries1,2, . . . ,k is a way of putting an integer i ∈ {1,2, . . . ,k}into each box of Y (λ) such that the entriesweakly increase along rows and strictly increase along columns.
Theorem: (sλ)λis ak-partition is a basis of the k-module S. 10 / 31
Symmetric polynomials: the s-basis (Schur polynomials) For eachk-partitionλ, we letsλ be theλ-th Schur polynomial:
sλ= det
xiλj+k−j
1≤i≤k,1≤j≤k
det
xik−j
1≤i≤k,1≤j≤k
(alternant formula)
= det
(hλi−i+j)1≤i≤k,1≤j≤k
(Jacobi-Trudi). Theorem: The equality above holds, and sλ is a symmetric polynomial with nonnegative coefficients.
Theorem: (sλ)λis ak-partition is a basis of the k-module S.
10 / 31
Symmetric polynomials: Littlewood-Richardson coefficients Ifλandµ are twok-partitions, then the product sλsµ can be again written as a k-linear combination of Schur polynomials (since these form a basis):
sλsµ= X
νis ak-partition
cλ,µν sν,
where the cλ,µν lie ink. Thesecλ,µν are called the Littlewood-Richardson coefficients.
Theorem: These Littlewood-Richardson coefficients cλ,µν are nonnegative integers (and count something).
11 / 31
Symmetric polynomials: Littlewood-Richardson coefficients Ifλandµ are twok-partitions, then the product sλsµ can be again written as a k-linear combination of Schur polynomials (since these form a basis):
sλsµ= X
νis ak-partition
cλ,µν sν,
where the cλ,µν lie ink. Thesecλ,µν are called the Littlewood-Richardson coefficients.
Theorem: These Littlewood-Richardson coefficients cλ,µν are nonnegative integers (and count something).
11 / 31
Symmetric polynomials: Schur polynomials for non-partitions We have defined
sλ = det
(hλi−i+j)1≤i≤k,1≤j≤k for k-partitions λ.
Apply the same definition to arbitrary λ∈Zk.
Proposition: Ifα∈Zk, then sα is either 0 or equals ±sλ for somek-partitionλ.
More precisely: Let
β = (α1+ (k−1), α2+ (k−2), . . . , αk+ (k−k)). Ifβ has a negative entry, thensα= 0.
Ifβ has two equal entries, thensα = 0.
Otherwise, letγ be the k-tuple obtained by sorting β in decreasing order, and letσ be the permutation of the indices that causes this sorting. Let λbe the k-partition (γ1−(k−1), γ2−(k−2), . . . , γk−(k−k)). Then, sα = (−1)σsλ.
Also, the alternant formula still holds if all λi + (k−i) are
≥0.
12 / 31
Symmetric polynomials: Schur polynomials for non-partitions We have defined
sλ = det
(hλi−i+j)1≤i≤k,1≤j≤k for k-partitions λ.
Apply the same definition to arbitrary λ∈Zk.
Proposition: Ifα∈Zk, then sα is either 0 or equals ±sλ for somek-partitionλ.
(So we get nothing really new.) More precisely: Let
β = (α1+ (k−1), α2+ (k−2), . . . , αk+ (k−k)).
Ifβ has a negative entry, thensα= 0. Ifβ has two equal entries, thensα = 0.
Otherwise, letγ be the k-tuple obtained by sorting β in decreasing order, and letσ be the permutation of the indices that causes this sorting. Let λbe the k-partition (γ1−(k−1), γ2−(k−2), . . . , γk−(k−k)). Then, sα = (−1)σsλ.
Also, the alternant formula still holds if all λi + (k−i) are
≥0.
12 / 31
Symmetric polynomials: Schur polynomials for non-partitions We have defined
sλ = det
(hλi−i+j)1≤i≤k,1≤j≤k for k-partitions λ.
Apply the same definition to arbitrary λ∈Zk.
Proposition: Ifα∈Zk, then sα is either 0 or equals ±sλ for somek-partitionλ.
More precisely: Let
β = (α1+ (k−1), α2+ (k−2), . . . , αk+ (k−k)).
Ifβ has a negative entry, thensα= 0.
Ifβ has two equal entries, thensα = 0.
Otherwise, letγ be the k-tuple obtained by sorting β in decreasing order, and letσ be the permutation of the indices that causes this sorting. Let λbe the k-partition (γ1−(k−1), γ2−(k−2), . . . , γk−(k−k)). Then, sα = (−1)σsλ.
Also, the alternant formula still holds if all λi + (k−i) are
≥0.
12 / 31
Symmetric polynomials: Schur polynomials for non-partitions We have defined
sλ = det
(hλi−i+j)1≤i≤k,1≤j≤k for k-partitions λ.
Apply the same definition to arbitrary λ∈Zk.
Proposition: Ifα∈Zk, then sα is either 0 or equals ±sλ for somek-partitionλ.
More precisely: Let
β = (α1+ (k−1), α2+ (k−2), . . . , αk+ (k−k)).
Ifβ has a negative entry, thensα= 0.
Ifβ has two equal entries, thensα = 0.
Otherwise, letγ be the k-tuple obtained by sorting β in decreasing order, and letσ be the permutation of the indices that causes this sorting. Let λbe the k-partition (γ1−(k−1), γ2−(k−2), . . . , γk−(k−k)). Then, sα = (−1)σsλ.
Also, the alternant formula still holds if all λi + (k−i) are
≥0. 12 / 31
Symmetric polynomials: Schur polynomials for non-partitions We have defined
sλ = det
(hλi−i+j)1≤i≤k,1≤j≤k for k-partitions λ.
Apply the same definition to arbitrary λ∈Zk.
Proposition: Ifα∈Zk, then sα is either 0 or equals ±sλ for somek-partitionλ.
More precisely: Let
β = (α1+ (k−1), α2+ (k−2), . . . , αk+ (k−k)).
Ifβ has a negative entry, thensα= 0.
Ifβ has two equal entries, thensα = 0.
Otherwise, letγ be the k-tuple obtained by sorting β in decreasing order, and letσ be the permutation of the indices that causes this sorting. Let λbe the k-partition (γ1−(k−1), γ2−(k−2), . . . , γk−(k−k)). Then, sα = (−1)σsλ.
Also, the alternant formula still holds if all λi + (k−i) are
≥0. 12 / 31
A more general setting: a1,a2, . . . ,ak and J
Pick any integer n≥k.
Let a1,a2, . . . ,ak ∈ P such that degai <n−k+i for all i. (For example, this holds if ai ∈k.)
Let J be the ideal of P generated by the k differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
Theorem (G.): The k-module PJ is free with basis (xα)α∈
Nk;αi<n−k+ifor eachi
informally: “
x1<n−k+1x2<n−k+2· · ·xn<n
”
where the overline means “projection” onto whatever quotient we need (here: from P ontoPJ).
(This basis has n(n−1)· · ·(n−k+ 1) elements.)
13 / 31
A more general setting: a1,a2, . . . ,ak and J
Pick any integer n≥k.
Let a1,a2, . . . ,ak ∈ P such that degai <n−k+i for all i. (For example, this holds if ai ∈k.)
Let J be the ideal ofP generated by thek differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
Theorem (G.): The k-module PJ is free with basis (xα)α∈
Nk;αi<n−k+ifor eachi
informally: “
x1<n−k+1x2<n−k+2· · ·xn<n
”
where the overline means “projection” onto whatever quotient we need (here: from P ontoPJ).
(This basis has n(n−1)· · ·(n−k+ 1) elements.)
13 / 31
A more general setting: a1,a2, . . . ,ak and J
Pick any integer n≥k.
Let a1,a2, . . . ,ak ∈ P such that degai <n−k+i for all i. (For example, this holds if ai ∈k.)
Let J be the ideal ofP generated by thek differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
Theorem (G.): The k-module PJ is free with basis (xα)α∈
Nk;αi<n−k+ifor eachi
informally: “
x1<n−k+1x2<n−k+2· · ·xn<n
”
where the overline means “projection” onto whatever quotient we need (here: from P ontoPJ).
(This basis has n(n−1)· · ·(n−k+ 1) elements.)
13 / 31
A more general setting: a1,a2, . . . ,ak and J
Pick any integer n≥k.
Let a1,a2, . . . ,ak ∈ P such that degai <n−k+i for all i. (For example, this holds if ai ∈k.)
Let J be the ideal ofP generated by thek differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
Theorem (G.): The k-module PJ is free with basis (xα)α∈
Nk;αi<n−k+ifor eachi
informally: “
x1<n−k+1x2<n−k+2· · ·xn<n
”
where the overline means “projection” onto whatever quotient we need (here: from P ontoPJ).
(This basis has n(n−1)· · ·(n−k+ 1) elements.)
13 / 31
A slightly less general setting: symmetric a1,a2, . . . ,ak and J
FROM NOW ON, assume that a1,a2, . . . ,ak ∈ S.
Let I be the ideal ofS generated by the k differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak. (Same differences as forJ, but we are generating an ideal of S now.)
Let ω= (n−k,n−k, . . . ,n−k)
| {z }
kentries
and
Pk,n ={λis a k-partition | λ1≤n−k}
={k-partitions λ⊆ω}.
Here, for two k-partitions α andβ, we say that α⊆β if and only if αi ≤βi for all i.
Theorem (G.): The k-module SI is free with basis (sλ)λ∈P
k,n.
14 / 31
A slightly less general setting: symmetric a1,a2, . . . ,ak and J
FROM NOW ON, assume that a1,a2, . . . ,ak ∈ S.
Let I be the ideal ofS generated by the k differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
(Same differences as forJ, but we are generating an ideal of S now.)
Let ω= (n−k,n−k, . . . ,n−k)
| {z }
kentries
and
Pk,n={λis a k-partition | λ1≤n−k}
={k-partitions λ⊆ω}.
Here, for two k-partitions α andβ, we say that α⊆β if and only if αi ≤βi for all i.
Theorem (G.): The k-module SI is free with basis (sλ)λ∈P
k,n.
14 / 31
A slightly less general setting: symmetric a1,a2, . . . ,ak and J
FROM NOW ON, assume that a1,a2, . . . ,ak ∈ S.
Let I be the ideal ofS generated by the k differences hn−k+1−a1, hn−k+2−a2, . . . , hn−ak.
(Same differences as forJ, but we are generating an ideal of S now.)
Let ω= (n−k,n−k, . . . ,n−k)
| {z }
kentries
and
Pk,n={λis a k-partition | λ1≤n−k}
={k-partitions λ⊆ω}.
Here, for two k-partitions α andβ, we say that α⊆β if and only if αi ≤βi for all i.
Theorem (G.): The k-module SI is free with basis (sλ)λ∈P
k,n.
14 / 31
An even less general setting: constant a1,a2, . . . ,ak
FROM NOW ON, assume that a1,a2, . . . ,ak ∈k.
This setting still is general enough to encompass ...
classical cohomology: Ifk=Zand
a1=a2=· · ·=ak = 0, thenSI becomes the cohomology ring H∗(Gr (k,n)); the basis (sλ)λ∈P
k,n
corresponds to the Schubert classes.
quantum cohomology: Ifk=Z[q] and
a1=a2=· · ·=ak−1 = 0 andak =−(−1)kq, then SI becomes the quantum cohomology ring
QH∗(Gr (k,n)).
The above theorem lets us work in these rings (and more generally) without relying on geometry.
15 / 31
An even less general setting: constant a1,a2, . . . ,ak
FROM NOW ON, assume that a1,a2, . . . ,ak ∈k.
This setting still is general enough to encompass ...
classical cohomology: Ifk=Zand
a1=a2=· · ·=ak = 0, thenSI becomes the cohomology ring H∗(Gr (k,n)); the basis (sλ)λ∈P
k,n
corresponds to the Schubert classes.
quantum cohomology: Ifk=Z[q] and
a1=a2=· · ·=ak−1 = 0 andak =−(−1)kq, then SI becomes the quantum cohomology ring
QH∗(Gr (k,n)).
The above theorem lets us work in these rings (and more generally) without relying on geometry.
15 / 31
An even less general setting: constant a1,a2, . . . ,ak
FROM NOW ON, assume that a1,a2, . . . ,ak ∈k.
This setting still is general enough to encompass ...
classical cohomology: Ifk=Zand
a1=a2=· · ·=ak = 0, thenSI becomes the cohomology ring H∗(Gr (k,n)); the basis (sλ)λ∈P
k,n
corresponds to the Schubert classes.
quantum cohomology: Ifk=Z[q] and
a1=a2=· · ·=ak−1 = 0 andak =−(−1)kq, then SI becomes the quantum cohomology ring
QH∗(Gr (k,n)).
The above theorem lets us work in these rings (and more generally) without relying on geometry.
15 / 31
S3-symmetry of the Gromov–Witten invariants Recall that (sλ)λ∈P
k,n is a basis of the k-moduleSI.
For each µ∈Pk,n, let coeffµ:SI →ksend each element to its sµ-coordinate wrt this basis.
For every k-partition ν= (ν1, ν2, . . . , νk)∈Pk,n, we define ν∨:= (n−k−νk,n−k−νk−1, . . . ,n−k−ν1)∈Pk,n. This k-partition ν∨ is called the complementof ν.
For any three k-partitionsα, β, γ ∈Pk,n, let gα,β,γ:= coeffγ∨(sαsβ)∈k.
These generalize the Littlewood–Richardson coefficients and (3-point) Gromov–Witten invariants.
Theorem (G.): For anyα, β, γ∈Pk,n, we have
gα,β,γ =gα,γ,β=gβ,α,γ =gβ,γ,α=gγ,α,β=gγ,β,α
= coeffω(sαsβsγ).
Equivalent restatement: Each ν∈Pk,n andf ∈ SI satisfy coeffω(sνf) = coeffν∨(f).
16 / 31
S3-symmetry of the Gromov–Witten invariants Recall that (sλ)λ∈P
k,n is a basis of the k-moduleSI.
For each µ∈Pk,n, let coeffµ:SI →ksend each element to its sµ-coordinate wrt this basis.
For every k-partition ν= (ν1, ν2, . . . , νk)∈Pk,n, we define ν∨:= (n−k−νk,n−k−νk−1, . . . ,n−k−ν1)∈Pk,n. Thisk-partitionν∨ is called the complementof ν.
For any three k-partitionsα, β, γ ∈Pk,n, let gα,β,γ:= coeffγ∨(sαsβ)∈k.
These generalize the Littlewood–Richardson coefficients and (3-point) Gromov–Witten invariants.
Theorem (G.): For anyα, β, γ∈Pk,n, we have
gα,β,γ =gα,γ,β=gβ,α,γ =gβ,γ,α=gγ,α,β=gγ,β,α
= coeffω(sαsβsγ).
Equivalent restatement: Each ν∈Pk,n andf ∈ SI satisfy coeffω(sνf) = coeffν∨(f).
16 / 31