You are currently browsing the category archive for the ‘math.CO’ category.

Peter Denton, Stephen Parke, Xining Zhang, and I have just uploaded to the arXiv a completely rewritten version of our previous paper, now titled “Eigenvectors from Eigenvalues: a survey of a basic identity in linear algebra“. This paper is now a survey of the various literature surrounding the following basic identity in linear algebra, which we propose to call the *eigenvector-eigenvalue identity*:

Theorem 1 (Eigenvector-eigenvalue identity)Let be an Hermitian matrix, with eigenvalues . Let be a unit eigenvector corresponding to the eigenvalue , and let be the component of . Thenwhere is the Hermitian matrix formed by deleting the row and column from .

When we posted the first version of this paper, we were unaware of previous appearances of this identity in the literature; a related identity had been used by Erdos-Schlein-Yau and by myself and Van Vu for applications to random matrix theory, but to our knowledge this specific identity appeared to be new. Even two months after our preprint first appeared on the arXiv in August, we had only learned of one other place in the literature where the identity showed up (by Forrester and Zhang, who also cite an earlier paper of Baryshnikov).

The situation changed rather dramatically with the publication of a popular science article in Quanta on this identity in November, which gave this result significantly more exposure. Within a few weeks we became informed (through private communication, online discussion, and exploration of the citation tree around the references we were alerted to) of over three dozen places where the identity, or some other closely related identity, had previously appeared in the literature, in such areas as numerical linear algebra, various aspects of graph theory (graph reconstruction, chemical graph theory, and walks on graphs), inverse eigenvalue problems, random matrix theory, and neutrino physics. As a consequence, we have decided to completely rewrite our article in order to collate this crowdsourced information, and survey the history of this identity, all the known proofs (we collect seven distinct ways to prove the identity (or generalisations thereof)), and all the applications of it that we are currently aware of. The citation graph of the literature that this *ad hoc* crowdsourcing effort produced is only very weakly connected, which we found surprising:

The earliest explicit appearance of the eigenvector-eigenvalue identity we are now aware of is in a 1966 paper of Thompson, although this paper is only cited (directly or indirectly) by a fraction of the known literature, and also there is a precursor identity of Löwner from 1934 that can be shown to imply the identity as a limiting case. At the end of the paper we speculate on some possible reasons why this identity only achieved a modest amount of recognition and dissemination prior to the November 2019 Quanta article.

Earlier this month, Hao Huang (who, incidentally, was a graduate student here at UCLA) gave a remarkably short proof of a long-standing problem in theoretical computer science known as the sensitivity conjecture. See for instance this blog post of Gil Kalai for further discussion and links to many other online discussions of this result. One formulation of the theorem proved is as follows. Define the -dimensional hypercube graph to be the graph with vertex set , and with every vertex joined to the vertices , where is the standard basis of .

Theorem 1 (Lower bound on maximum degree of induced subgraphs of hypercube)Let be a set of at least vertices in . Then there is a vertex in that is adjacent (in ) to at least other vertices in .

The bound (or more precisely, ) is completely sharp, as shown by Chung, Furedi, Graham, and Seymour; we describe this example below the fold. When combined with earlier reductions of Gotsman-Linial and Nisan-Szegedy; we give these below the fold also.

Let be the adjacency matrix of (where we index the rows and columns directly by the vertices in , rather than selecting some enumeration ), thus when for some , and otherwise. The above theorem then asserts that if is a set of at least vertices, then the minor of has a row (or column) that contains at least non-zero entries.

The key step to prove this theorem is the construction of rather curious variant of the adjacency matrix :

Proposition 2There exists a matrix which is entrywise dominated by in the sense that

Assuming this proposition, the proof of Theorem 1 can now be quickly concluded. If we view as a linear operator on the -dimensional space of functions of , then by hypothesis this space has a -dimensional subspace on which acts by multiplication by . If is a set of at least vertices in , then the space of functions on has codimension at most in , and hence intersects non-trivially. Thus the minor of also has as an eigenvalue (this can also be derived from the Cauchy interlacing inequalities), and in particular this minor has operator norm at least . By Schur’s test, this implies that one of the rows or columns of this matrix has absolute values summing to at least , giving the claim.

Remark 3The argument actually gives a strengthening of Theorem 1: there exists a vertex of with the property that for every natural number , there are at least paths of length in the restriction of to that start from . Indeed, if we let be an eigenfunction of on , and let be a vertex in that maximises the value of , then for any we have that the component of is equal to ; on the other hand, by the triangle inequality, this component is at most times the number of length paths in starting from , giving the claim.

This argument can be viewed as an instance of a more general “interlacing method” to try to control the behaviour of a graph on all large subsets by first generating a matrix on with very good spectral properties, which are then partially inherited by the minor of by interlacing inequalities. In previous literature using this method (see e.g., this survey of Haemers, or this paper of Wilson), either the original adjacency matrix , or some non-negatively weighted version of that matrix, was used as the controlling matrix ; the novelty here is the use of signed controlling matrices. It will be interesting to see what further variants and applications of this method emerge in the near future. (Thanks to Anurag Bishoi in the comments for these references.)

The “magic” step in the above argument is constructing . In Huang’s paper, is constructed recursively in the dimension in a rather simple but mysterious fashion. Very recently, Roman Karasev gave an interpretation of this matrix in terms of the exterior algebra on . In this post I would like to give an alternate interpretation in terms of the operation of *twisted convolution*, which originated in the theory of the Heisenberg group in quantum mechanics.

Firstly note that the original adjacency matrix , when viewed as a linear operator on , is a convolution operator

where

is the counting measure on the standard basis , and denotes the ordinary convolution operation

As is well known, this operation is commutative and associative. Thus for instance the square of the adjacency operator is also a convolution operator

where the convolution kernel is moderately complicated:

The factor in this expansion comes from combining the two terms and , which both evaluate to .

More generally, given any bilinear form , one can define the *twisted convolution*

of two functions . This operation is no longer commutative (unless is symmetric). However, it remains associative; indeed, one can easily compute that

In particular, if we define the twisted convolution operator

then the square is also a twisted convolution operator

and the twisted convolution kernel can be computed as

For general bilinear forms , this twisted convolution is just as messy as is. But if we take the specific bilinear form

then for and for , and the above twisted convolution simplifies to

and now is very simple:

Thus the only eigenvalues of are and . The matrix is entrywise dominated by in the sense of (1), and in particular has trace zero; thus the and eigenvalues must occur with equal multiplicity, so in particular the eigenvalue occurs with multiplicity since the matrix has dimensions . This establishes Proposition 2.

Remark 4Twisted convolution is actually just a component of ordinary convolution, but not on the original group ; instead it relates to convolution on a Heisenberg group extension of this group. More specifically, define the Heisenberg group to be the set of pairs with group lawand inverse operation

(one can dispense with the negative signs here if desired, since we are in characteristic two). Convolution on is defined in the usual manner: one has

for any . Now if is a function on the original group , we can define the lift by the formula

and then by chasing all the definitions one soon verifies that

for any , thus relating twisted convolution to Heisenberg group convolution .

Remark 5With the twisting by the specific bilinear form given by (2), convolution by and now anticommute rather than commute. This makes the twisted convolution algebra isomorphic to a Clifford algebra (the real or complex algebra generated by formal generators subject to the relations for ) rather than the commutative algebra more familiar to abelian Fourier analysis. This connection to Clifford algebra (also observed independently by Tom Mrowka and by Daniel Matthews) may be linked to the exterior algebra interpretation of the argument in the recent preprint of Karasev mentioned above.

Remark 6One could replace the form (2) in this argument by any other bilinear form that obeyed the relations and for . However, this additional level of generality does not add much; any such will differ from by an antisymmetric form (so that for all , which in characteristic two implied that for all ), and such forms can always be decomposed as , where . As such, the matrices and are conjugate, with the conjugation operator being the diagonal matrix with entries at each vertex .

Remark 7(Added later) This remark combines the two previous remarks. One can view any of the matrices in Remark 6 as components of a single canonical matrix that is still of dimensions , but takes values in the Clifford algebra from Remark 5; with this “universal algebra” perspective, one no longer needs to make any arbitrary choices of form . More precisely, let denote the vector space of functions from the hypercube to the Clifford algebra; as a real vector space, this is a dimensional space, isomorphic to the direct sum of copies of , as the Clifford algebra is itself dimensional. One can then define a canonical Clifford adjacency operator on this space bywhere are the generators of . This operator can either be identified with a Clifford-valued matrix or as a real-valued matrix. In either case one still has the key algebraic relations and , ensuring that when viewed as a real matrix, half of the eigenvalues are equal to and half equal to . One can then use this matrix in place of any of the to establish Theorem 1 (noting that Schur’s test continues to work for Clifford-valued matrices because of the norm structure on ).

To relate to the real matrices , first observe that each point in the hypercube can be associated with a one-dimensional real subspace (i.e., a line) in the Clifford algebra by the formula

for any (note that this definition is well-defined even if the are out of order or contain repetitions). This can be viewed as a discrete line bundle over the hypercube. Since for any , we see that the -dimensional real linear subspace of of sections of this bundle, that is to say the space of functions such that for all , is an invariant subspace of . (Indeed, using the left-action of the Clifford algebra on , which commutes with , one can naturally identify with , with the left action of acting purely on the first factor and acting purely on the second factor.) Any trivialisation of this line bundle lets us interpret the restriction of to as a real matrix. In particular, given one of the bilinear forms from Remark 6, we can identify with by identifying any real function with the lift defined by

whenever . A somewhat tedious computation using the properties of then eventually gives the intertwining identity

and so is conjugate to .

Let be some domain (such as the real numbers). For any natural number , let denote the space of symmetric real-valued functions on variables , thus

for any permutation . For instance, for any natural numbers , the elementary symmetric polynomials

will be an element of . With the pointwise product operation, becomes a commutative real algebra. We include the case , in which case consists solely of the real constants.

Given two natural numbers , one can “lift” a symmetric function of variables to a symmetric function of variables by the formula

where ranges over all injections from to (the latter formula making it clearer that is symmetric). Thus for instance

and

Also we have

With these conventions, we see that vanishes for , and is equal to if . We also have the transitivity

if .

The lifting map is a linear map from to , but it is not a ring homomorphism. For instance, when , one has

In general, one has the identity

for all natural numbers and , , where range over all injections , with . Combinatorially, the identity (2) follows from the fact that given any injections and with total image of cardinality , one has , and furthermore there exist precisely triples of injections , , such that and .

Example 1When , one haswhich is just a restatement of the identity

Note that the coefficients appearing in (2) do not depend on the final number of variables . We may therefore abstract the role of from the law (2) by introducing the real algebra of formal sums

where for each , is an element of (with only finitely many of the being non-zero), and with the formal symbol being formally linear, thus

and

for and scalars , and with multiplication given by the analogue

of (2). Thus for instance, in this algebra we have

and

Informally, is an abstraction (or “inverse limit”) of the concept of a symmetric function of an unspecified number of variables, which are formed by summing terms that each involve only a bounded number of these variables at a time. One can check (somewhat tediously) that is indeed a commutative real algebra, with a unit . (I do not know if this algebra has previously been studied in the literature; it is somewhat analogous to the abstract algebra of finite linear combinations of Schur polynomials, with multiplication given by a Littlewood-Richardson rule. )

For natural numbers , there is an obvious specialisation map from to , defined by the formula

Thus, for instance, maps to and to . From (2) and (3) we see that this map is an algebra homomorphism, even though the maps and are not homomorphisms. By inspecting the component of we see that the homomorphism is in fact surjective.

Now suppose that we have a measure on the space , which then induces a product measure on every product space . To avoid degeneracies we will assume that the integral is strictly positive. Assuming suitable measurability and integrability hypotheses, a function can then be integrated against this product measure to produce a number

In the event that arises as a lift of another function , then from Fubini’s theorem we obtain the formula

is an element of the formal algebra , then

Note that by hypothesis, only finitely many terms on the right-hand side are non-zero.

Now for a key observation: whereas the left-hand side of (6) only makes sense when is a natural number, the right-hand side is meaningful when takes a fractional value (or even when it takes negative or complex values!), interpreting the binomial coefficient as a polynomial in . As such, this suggests a way to introduce a “virtual” concept of a symmetric function on a fractional power space for such values of , and even to integrate such functions against product measures , even if the fractional power does not exist in the usual set-theoretic sense (and similarly does not exist in the usual measure-theoretic sense). More precisely, for arbitrary real or complex , we now *define* to be the space of abstract objects

with and (and now interpreted as formal symbols, with the structure of a commutative real algebra inherited from , thus

In particular, the multiplication law (2) continues to hold for such values of , thanks to (3). Given any measure on , we formally define a measure on with regards to which we can integrate elements of by the formula (6) (providing one has sufficient measurability and integrability to make sense of this formula), thus providing a sort of “fractional dimensional integral” for symmetric functions. Thus, for instance, with this formalism the identities (4), (5) now hold for fractional values of , even though the formal space no longer makes sense as a set, and the formal measure no longer makes sense as a measure. (The formalism here is somewhat reminiscent of the technique of dimensional regularisation employed in the physical literature in order to assign values to otherwise divergent integrals. See also this post for an unrelated abstraction of the integration concept involving integration over supercommutative variables (and in particular over fermionic variables).)

Example 2Suppose is a probability measure on , and is a random variable; on any power , we let be the usual independent copies of on , thus for . Then for any real or complex , the formal integralcan be evaluated by first using the identity

(cf. (1)) and then using (6) and the probability measure hypothesis to conclude that

For a natural number, this identity has the probabilistic interpretation

whenever are jointly independent copies of , which reflects the well known fact that the sum has expectation and variance . One can thus view (7) as an abstract generalisation of (8) to the case when is fractional, negative, or even complex, despite the fact that there is no sensible way in this case to talk about independent copies of in the standard framework of probability theory.

In this particular case, the quantity (7) is non-negative for every nonnegative , which looks plausible given the form of the left-hand side. Unfortunately, this sort of non-negativity does not always hold; for instance, if has mean zero, one can check that

and the right-hand side can become negative for . This is a shame, because otherwise one could hope to start endowing with some sort of commutative von Neumann algebra type structure (or the abstract probability structure discussed in this previous post) and then interpret it as a genuine measure space rather than as a virtual one. (This failure of positivity is related to the fact that the characteristic function of a random variable, when raised to the power, need not be a characteristic function of any random variable once is no longer a natural number: “fractional convolution” does not preserve positivity!) However, one vestige of positivity remains: if is non-negative, then so is

One can wonder what the point is to all of this abstract formalism and how it relates to the rest of mathematics. For me, this formalism originated implicitly in an old paper I wrote with Jon Bennett and Tony Carbery on the multilinear restriction and Kakeya conjectures, though we did not have a good language for working with it at the time, instead working first with the case of natural number exponents and appealing to a general extrapolation theorem to then obtain various identities in the fractional case. The connection between these fractional dimensional integrals and more traditional integrals ultimately arises from the simple identity

(where the right-hand side should be viewed as the fractional dimensional integral of the unit against ). As such, one can manipulate powers of ordinary integrals using the machinery of fractional dimensional integrals. A key lemma in this regard is

Lemma 3 (Differentiation formula)Suppose that a positive measure on depends on some parameter and varies by the formula

for some function . Let be any real or complex number. Then, assuming sufficient smoothness and integrability of all quantities involved, we have

for all that are independent of . If we allow to now depend on also, then we have the more general total derivative formula

again assuming sufficient amounts of smoothness and regularity.

*Proof:* We just prove (10), as (11) then follows by same argument used to prove the usual product rule. By linearity it suffices to verify this identity in the case for some symmetric function for a natural number . By (6), the left-hand side of (10) is then

Differentiating under the integral sign using (9) we have

and similarly

where are the standard copies of on :

By the product rule, we can thus expand (12) as

where we have suppressed the dependence on for brevity. Since , we can write this expression using (6) as

where is the symmetric function

But from (2) one has

and the claim follows.

Remark 4It is also instructive to prove this lemma in the special case when is a natural number, in which case the fractional dimensional integral can be interpreted as a classical integral. In this case, the identity (10) is immediate from applying the product rule to (9) to conclude thatOne could in fact derive (10) for arbitrary real or complex from the case when is a natural number by an extrapolation argument; see the appendix of my paper with Bennett and Carbery for details.

Let us give a simple PDE application of this lemma as illustration:

Proposition 5 (Heat flow monotonicity)Let be a solution to the heat equation with initial data a rapidly decreasing finite non-negative Radon measure, or more explicitlyfor al . Then for any , the quantity

is monotone non-decreasing in for , constant for , and monotone non-increasing for .

*Proof:* By a limiting argument we may assume that is absolutely continuous, with Radon-Nikodym derivative a test function; this is more than enough regularity to justify the arguments below.

For any , let denote the Radon measure

Then the quantity can be written as a fractional dimensional integral

Observe that

and thus by Lemma 3 and the product rule

where we use for the variable of integration in the factor space of .

To simplify this expression we will take advantage of integration by parts in the variable. Specifically, in any direction , we have

and hence by Lemma 3

Multiplying by and integrating by parts, we see that

where we use the Einstein summation convention in . Similarly, if is any reasonable function depending only on , we have

and hence on integration by parts

We conclude that

and thus by (13)

The choice of that then achieves the most cancellation turns out to be (this cancels the terms that are linear or quadratic in the ), so that . Repeating the calculations establishing (7), one has

and

where is the random variable drawn from with the normalised probability measure . Since , one thus has

This expression is clearly non-negative for , equal to zero for , and positive for , giving the claim. (One could simplify here as if desired, though it is not strictly necessary to do so for the proof.)

Remark 6As with Remark 4, one can also establish the identity (14) first for natural numbers by direct computation avoiding the theory of fractional dimensional integrals, and then extrapolate to the case of more general values of . This particular identity is also simple enough that it can be directly established by integration by parts without much difficulty, even for fractional values of .

A more complicated version of this argument establishes the non-endpoint multilinear Kakeya inequality (without any logarithmic loss in a scale parameter ); this was established in my previous paper with Jon Bennett and Tony Carbery, but using the “natural number first” approach rather than using the current formalism of fractional dimensional integration. However, the arguments can be translated into this formalism without much difficulty; we do so below the fold. (To simplify the exposition slightly we will not address issues of establishing enough regularity and integrability to justify all the manipulations, though in practice this can be done by standard limiting arguments.)

Joni Teräväinen and I have just uploaded to the arXiv our paper “Value patterns of multiplicative functions and related sequences“, submitted to Forum of Mathematics, Sigma. This paper explores how to use recent technology on correlations of multiplicative (or nearly multiplicative functions), such as the “entropy decrement method”, in conjunction with techniques from additive combinatorics, to establish new results on the sign patterns of functions such as the Liouville function . For instance, with regards to length 5 sign patterns

of the Liouville function, we can now show that at least of the possible sign patterns in occur with positive upper density. (Conjecturally, all of them do so, and this is known for all shorter sign patterns, but unfortunately seems to be the limitation of our methods.)

The Liouville function can be written as , where is the number of prime factors of (counting multiplicity). One can also consider the variant , which is a completely multiplicative function taking values in the cube roots of unity . Here we are able to show that all sign patterns in occur with positive lower density as sign patterns of this function. The analogous result for was already known (see this paper of Matomäki, Radziwiłł, and myself), and in that case it is even known that all sign patterns occur with equal logarithmic density (from this paper of myself and Teräväinen), but these techniques barely fail to handle the case by itself (largely because the “parity” arguments used in the case of the Liouville function no longer control three-point correlations in the case) and an additional additive combinatorial tool is needed. After applying existing technology (such as entropy decrement methods), the problem roughly speaking reduces to locating patterns for a certain partition of a compact abelian group (think for instance of the unit circle , although the general case is a bit more complicated, in particular if is disconnected then there is a certain “coprimality” constraint on , also we can allow the to be replaced by any with divisible by ), with each of the having measure . An inequality of Kneser just barely fails to guarantee the existence of such patterns, but by using an inverse theorem for Kneser’s inequality in this previous paper of mine we are able to identify precisely the obstruction for this method to work, and rule it out by an *ad hoc* method.

The same techniques turn out to also make progress on some conjectures of Erdös-Pomerance and Hildebrand regarding patterns of the largest prime factor of a natural number . For instance, we improve results of Erdös-Pomerance and of Balog demonstrating that the inequalities

and

each hold for infinitely many , by demonstrating the stronger claims that the inequalities

and

each hold for a set of of positive lower density. As a variant, we also show that we can find a positive density set of for which

for any fixed (this improves on a previous result of Hildebrand with replaced by . A number of other results of this type are also obtained in this paper.

In order to obtain these sorts of results, one needs to extend the entropy decrement technology from the setting of multiplicative functions to that of what we call “weakly stable sets” – sets which have some multiplicative structure, in the sense that (roughly speaking) there is a set such that for all small primes , the statements and are roughly equivalent to each other. For instance, if is a level set , one would take ; if instead is a set of the form , then one can take . When one has such a situation, then very roughly speaking, the entropy decrement argument then allows one to estimate a one-parameter correlation such as

with a two-parameter correlation such as

(where we will be deliberately vague as to how we are averaging over and ), and then the use of the “linear equations in primes” technology of Ben Green, Tamar Ziegler, and myself then allows one to replace this average in turn by something like

where is constrained to be not divisible by small primes but is otherwise quite arbitrary. This latter average can then be attacked by tools from additive combinatorics, such as translation to a continuous group model (using for instance the Furstenberg correspondence principle) followed by tools such as Kneser’s inequality (or inverse theorems to that inequality).

(This post is mostly intended for my own reference, as I found myself repeatedly looking up several conversions between polynomial bases on various occasions.)

Let denote the vector space of polynomials of one variable with real coefficients of degree at most . This is a vector space of dimension , and the sequence of these spaces form a filtration:

A standard basis for these vector spaces are given by the monomials : every polynomial in can be expressed uniquely as a linear combination of the first monomials . More generally, if one has any sequence of polynomials, with each of degree exactly , then an easy induction shows that forms a basis for .

In particular, if we have *two* such sequences and of polynomials, with each of degree and each of degree , then must be expressible uniquely as a linear combination of the polynomials , thus we have an identity of the form

for some *change of basis coefficients* . These coefficients describe how to convert a polynomial expressed in the basis into a polynomial expressed in the basis.

Many standard combinatorial quantities involving two natural numbers can be interpreted as such change of basis coefficients. The most familiar example are the binomial coefficients , which measures the conversion from the shifted monomial basis to the monomial basis , thanks to (a special case of) the binomial formula:

thus for instance

More generally, for any shift , the conversion from to is measured by the coefficients , thanks to the general case of the binomial formula.

But there are other bases of interest too. For instance if one uses the falling factorial basis

then the conversion from falling factorials to monomials is given by the Stirling numbers of the first kind :

thus for instance

and the conversion back is given by the Stirling numbers of the second kind :

thus for instance

If one uses the binomial functions as a basis instead of the falling factorials, one of course can rewrite these conversions as

and

thus for instance

and

As a slight variant, if one instead uses rising factorials

then the conversion to monomials yields the unsigned Stirling numbers of the first kind:

thus for instance

One final basis comes from the polylogarithm functions

For instance one has

and more generally one has

for all natural numbers and some polynomial of degree (the *Eulerian polynomials*), which when converted to the monomial basis yields the (shifted) Eulerian numbers

For instance

These particular coefficients also have useful combinatorial interpretations. For instance:

- The binomial coefficient is of course the number of -element subsets of .
- The unsigned Stirling numbers of the first kind are the number of permutations of with exactly cycles. The signed Stirling numbers are then given by the formula .
- The Stirling numbers of the second kind are the number of ways to partition into non-empty subsets.
- The Eulerian numbers are the number of permutations of with exactly ascents.

These coefficients behave similarly to each other in several ways. For instance, the binomial coefficients obey the well known Pascal identity

(with the convention that vanishes outside of the range ). In a similar spirit, the unsigned Stirling numbers of the first kind obey the identity

and the signed counterparts obey the identity

The Stirling numbers of the second kind obey the identity

and the Eulerian numbers obey the identity

Let , be additive groups (i.e., groups with an abelian addition group law). A map is a homomorphism if one has

for all . A map is an *affine* homomorphism if one has

for all *additive quadruples* in , by which we mean that and . The two notions are closely related; it is easy to verify that is an affine homomorphism if and only if is the sum of a homomorphism and a constant.

Now suppose that also has a translation-invariant metric . A map is said to be a quasimorphism if one has

for all , where denotes a quantity at a bounded distance from the origin. Similarly, is an *affine quasimorphism* if

for all additive quadruples in . Again, one can check that is an affine quasimorphism if and only if it is the sum of a quasimorphism and a constant (with the implied constant of the quasimorphism controlled by the implied constant of the affine quasimorphism). (Since every constant is itself a quasimorphism, it is in fact the case that affine quasimorphisms are quasimorphisms, but now the implied constant in the latter is not controlled by the implied constant of the former.)

“Trivial” examples of quasimorphisms include the sum of a homomorphism and a bounded function. Are there others? In some cases, the answer is no. For instance, suppose we have a quasimorphism . Iterating (2), we see that for any integer and natural number , which we can rewrite as for non-zero . Also, is Lipschitz. Sending , we can verify that is a Cauchy sequence as and thus tends to some limit ; we have for , hence for positive , and then one can use (2) one last time to obtain for all . Thus is the sum of the homomorphism and a bounded sequence.

In general, one can phrase this problem in the language of group cohomology (discussed in this previous post). Call a map a *-cocycle*. A *-cocycle* is a map obeying the identity

for all . Given a -cocycle , one can form its *derivative* by the formula

Such functions are called *-coboundaries*. It is easy to see that the abelian group of -coboundaries is a subgroup of the abelian group of -cocycles. The quotient of these two groups is the first group cohomology of with coefficients in , and is denoted .

If a -cocycle is bounded then its derivative is a bounded -coboundary. The quotient of the group of bounded -cocycles by the derivatives of bounded -cocycles is called the *bounded first group cohomology* of with coefficients in , and is denoted . There is an obvious homomorphism from to , formed by taking a coset of the space of derivatives of bounded -cocycles, and enlarging it to a coset of the space of -coboundaries. By chasing all the definitions, we see that all quasimorphism from to are the sum of a homomorphism and a bounded function if and only if this homomorphism is injective; in fact the quotient of the space of quasimorphisms by the sum of homomorphisms and bounded functions is isomorphic to the kernel of .

In additive combinatorics, one is often working with functions which only have additive structure a fraction of the time, thus for instance (1) or (3) might only hold “ of the time”. This makes it somewhat difficult to directly interpret the situation in terms of group cohomology. However, thanks to tools such as the Balog-Szemerédi-Gowers lemma, one can upgrade this sort of -structure to -structure – at the cost of restricting the domain to a smaller set. Here I record one such instance of this phenomenon, thus giving a tentative link between additive combinatorics and group cohomology. (I thank Yuval Wigderson for suggesting the problem of locating such a link.)

Theorem 1Let , be additive groups with , let be a subset of , let , and let be a function such thatfor additive quadruples in . Then there exists a subset of containing with , a subset of with , and a function such that

for all (thus, the derivative takes values in on ), and such that for each , one has

Presumably the constants and can be improved further, but we have not attempted to optimise these constants. We chose as the domain on which one has a bounded derivative, as one can use the Bogulybov lemma (see e.g, Proposition 4.39 of my book with Van Vu) to find a large Bohr set inside . In applications, the set need not have bounded size, or even bounded doubling; for instance, in the inverse theory over a small finite fields , one would be interested in the situation where is the group of matrices with coefficients in (for some large , and being the subset consisting of those matrices of rank bounded by some bound .

*Proof:* By hypothesis, there are triples such that and

Thus, there is a set with such that for all , one has (6) for pairs with ; in particular, there exists such that (6) holds for values of . Setting , we conclude that for each , one has

Consider the bipartite graph whose vertex sets are two copies of , and and connected by a (directed) edge if and (7) holds. Then this graph has edges. Applying (a slight modification of) the Balog-Szemerédi-Gowers theorem (for instance by modifying the proof of Corollary 5.19 of my book with Van Vu), we can then find a subset of with with the property that for any , there exist triples such that the edges all lie in this bipartite graph. This implies that, for all , there exist septuples obeying the constraints

and for . These constraints imply in particular that

Also observe that

Thus, if and are such that , we see that

for octuples in the hyperplane

By the pigeonhole principle, this implies that for any fixed , there can be at most sets of the form with , that are pairwise disjoint. Using a greedy algorithm, we conclude that there is a set of cardinality , such that each set with , intersects for some , or in other words that

This implies that there exists a subset of with , and an element for each , such that

for all . Note we may assume without loss of generality that and .

By construction of , and permuting labels, we can find 16-tuples such that

and

for . We sum this to obtain

and hence by (8)

where . Since

we see that there are only possible values of . By the pigeonhole principle, we conclude that at most of the sets can be disjoint. Arguing as before, we conclude that there exists a set of cardinality such that

whenever (10) holds.

For any , write arbitrarily as for some (with if , and if ) and then set

Then from (11) we have (4). For we have , and (5) then follows from (9).

I have just uploaded to the arXiv the paper “An inverse theorem for an inequality of Kneser“, submitted to a special issue of the Proceedings of the Steklov Institute of Mathematics in honour of Sergei Konyagin. It concerns an inequality of Kneser discussed previously in this blog, namely that

whenever are compact non-empty subsets of a compact connected additive group with probability Haar measure . (A later result of Kemperman extended this inequality to the nonabelian case.) This inequality is non-trivial in the regime

The connectedness of is essential, otherwise one could form counterexamples involving proper subgroups of of positive measure. In the blog post, I indicated how this inequality (together with a more “robust” strengthening of it) could be deduced from submodularity inequalities such as

which in turn easily follows from the identity and the inclusion , combined with the inclusion-exclusion formula.

In the non-trivial regime (2), equality can be attained in (1), for instance by taking to be the unit circle and to be arcs in that circle (obeying (2)). A bit more generally, if is an arbitrary connected compact abelian group and is a non-trivial character (i.e., a continuous homomorphism), then must be surjective (as has no non-trivial connected subgroups), and one can take and for some arcs in that circle (again choosing the measures of these arcs to obey (2)). The main result of this paper is an inverse theorem that asserts that this is the only way in which equality can occur in (1) (assuming (2)); furthermore, if (1) is close to being satisfied with equality and (2) holds, then must be close (in measure) to an example of the above form . Actually, for technical reasons (and for the applications we have in mind), it is important to establish an inverse theorem not just for (1), but for the more robust version mentioned earlier (in which the sumset is replaced by the partial sumset consisting of “popular” sums).

Roughly speaking, the idea is as follows. Let us informally call a *critical pair* if (2) holds and the inequality (1) (or more precisely, a robust version of this inequality) is almost obeyed with equality. The notion of a critical pair obeys some useful closure properties. Firstly, it is symmetric in , and invariant with respect to translation of either or . Furthermore, from the submodularity inequality (3), one can show that if and are critical pairs (with and positive), then and are also critical pairs. (Note that this is consistent with the claim that critical pairs only occur when come from arcs of a circle.) Similarly, from associativity , one can show that if and are critical pairs, then so are and .

One can combine these closure properties to obtain further ones. For instance, suppose is such that . Then (cheating a little bit), one can show that is also a critical pair, basically because is the union of the , , the are all critical pairs, and the all intersect each other. This argument doesn’t quite work as stated because one has to apply the closure property under union an uncountable number of times, but it turns out that if one works with the robust version of sumsets and uses a random sampling argument to approximate by the union of finitely many of the , then the argument can be made to work.

Using all of these closure properties, it turns out that one can start with an arbitrary critical pair and end up with a small set such that and are also critical pairs for all (say), where is the -fold sumset of . (Intuitively, if are thought of as secretly coming from the pullback of arcs by some character , then should be the pullback of a much shorter arc by the same character.) In particular, exhibits linear growth, in that for all . One can now use standard technology from inverse sumset theory to show first that has a very large Fourier coefficient (and thus is biased with respect to some character ), and secondly that is in fact almost of the form for some arc , from which it is not difficult to conclude similar statements for and and thus finish the proof of the inverse theorem.

In order to make the above argument rigorous, one has to be more precise about what the modifier “almost” means in the definition of a critical pair. I chose to do this in the language of “cheap” nonstandard analysis (aka asymptotic analysis), as discussed in this previous blog post; one could also have used the full-strength version of nonstandard analysis, but this does not seem to convey any substantial advantages. (One can also work in a more traditional “non-asymptotic” framework, but this requires one to keep much more careful account of various small error terms and leads to a messier argument.)

*[Update, Nov 15: Corrected the attribution of the inequality (1) to Kneser instead of Kemperman. Thanks to John Griesmer for pointing out the error.]*

Let be the Liouville function, thus is defined to equal when is the product of an even number of primes, and when is the product of an odd number of primes. The Chowla conjecture asserts that has the statistics of a random sign pattern, in the sense that

for all and all distinct natural numbers , where we use the averaging notation

For , this conjecture is equivalent to the prime number theorem (as discussed in this previous blog post), but the conjecture remains open for any .

In recent years, it has been realised that one can make more progress on this conjecture if one works instead with the logarithmically averaged version

of the conjecture, where we use the logarithmic averaging notation

Using the summation by parts (or telescoping series) identity

it is not difficult to show that the Chowla conjecture (1) for a given implies the logarithmically averaged conjecture (2). However, the converse implication is not at all clear. For instance, for , we have already mentioned that the Chowla conjecture

is equivalent to the prime number theorem; but the logarithmically averaged analogue

is significantly easier to show (a proof with the Liouville function replaced by the closely related Möbius function is given in this previous blog post). And indeed, significantly more is now known for the logarithmically averaged Chowla conjecture; in this paper of mine I had proven (2) for , and in this recent paper with Joni Teravainen, we proved the conjecture for all odd (with a different proof also given here).

In view of this emerging consensus that the logarithmically averaged Chowla conjecture was easier than the ordinary Chowla conjecture, it was thus somewhat of a surprise for me to read a recent paper of Gomilko, Kwietniak, and Lemanczyk who (among other things) established the following statement:

Theorem 1Assume that the logarithmically averaged Chowla conjecture (2) is true for all . Then there exists a sequence going to infinity such that the Chowla conjecture (1) is true for all along that sequence, that is to sayfor all and all distinct .

This implication does not use any special properties of the Liouville function (other than that they are bounded), and in fact proceeds by ergodic theoretic methods, focusing in particular on the ergodic decomposition of invariant measures of a shift into ergodic measures. Ergodic methods have proven remarkably fruitful in understanding these sorts of number theoretic and combinatorial problems, as could already be seen by the ergodic theoretic proof of Szemerédi’s theorem by Furstenberg, and more recently by the work of Frantzikinakis and Host on Sarnak’s conjecture. (My first paper with Teravainen also uses ergodic theory tools.) Indeed, many other results in the subject were first discovered using ergodic theory methods.

On the other hand, many results in this subject that were first proven ergodic theoretically have since been reproven by more combinatorial means; my second paper with Teravainen is an instance of this. As it turns out, one can also prove Theorem 1 by a standard combinatorial (or probabilistic) technique known as the second moment method. In fact, one can prove slightly more:

Theorem 2Let be a natural number. Assume that the logarithmically averaged Chowla conjecture (2) is true for . Then there exists a set of natural numbers of logarithmic density (that is, ) such thatfor any distinct .

It is not difficult to deduce Theorem 1 from Theorem 2 using a diagonalisation argument. Unfortunately, the known cases of the logarithmically averaged Chowla conjecture ( and odd ) are currently insufficient to use Theorem 2 for any purpose other than to reprove what is already known to be true from the prime number theorem. (Indeed, the even cases of Chowla, in either logarithmically averaged or non-logarithmically averaged forms, seem to be far more powerful than the odd cases; see Remark 1.7 of this paper of myself and Teravainen for a related observation in this direction.)

We now sketch the proof of Theorem 2. For any distinct , we take a large number and consider the limiting the second moment

We can expand this as

If all the are distinct, the hypothesis (2) tells us that the inner averages goes to zero as . The remaining averages are , and there are of these averages. We conclude that

By Markov’s inequality (and (3)), we conclude that for any fixed , there exists a set of upper logarithmic density at least , thus

such that

By deleting at most finitely many elements, we may assume that consists only of elements of size at least (say).

For any , if we let be the union of for , then has logarithmic density . By a diagonalisation argument (using the fact that the set of tuples is countable), we can then find a set of natural numbers of logarithmic density , such that for every , every sufficiently large element of lies in . Thus for every sufficiently large in , one has

for some with . By Cauchy-Schwarz, this implies that

interchanging the sums and using and , this implies that

We conclude on taking to infinity that

as required.

Szemerédi’s theorem asserts that all subsets of the natural numbers of positive density contain arbitrarily long arithmetic progressions. *Roth’s theorem* is the special case when one considers arithmetic progressions of length three. Both theorems have many important proofs using tools from additive combinatorics, (higher order) Fourier analysis, (hyper) graph regularity theory, and ergodic theory. However, the original proof by Endre Szemerédi, while extremely intricate, was purely combinatorial (and in particular “elementary”) and almost entirely self-contained, except for an invocation of the van der Waerden theorem. It is also notable for introducing a prototype of what is now known as the Szemerédi regularity lemma.

Back in 2005, I rewrote Szemerédi’s original proof in order to understand it better, however my rewrite ended up being about the same length as the original argument and was probably only usable to myself. In 2012, after Szemerédi was awarded the Abel prize, I revisited this argument with the intention to try to write up a more readable version of the proof, but ended up just presenting some ingredients of the argument in a blog post, rather than try to rewrite the whole thing. In that post, I suspected that the cleanest way to write up the argument would be through the language of nonstandard analysis (perhaps in an iterated hyperextension that could handle various hierarchies of infinitesimals), but was unable to actually achieve any substantial simplifications by passing to the nonstandard world.

A few weeks ago, I participated in a week-long workshop at the American Institute of Mathematics on “Nonstandard methods in combinatorial number theory”, and spent some time in a working group with Shabnam Akhtari, Irfam Alam, Renling Jin, Steven Leth, Karl Mahlburg, Paul Potgieter, and Henry Towsner to try to obtain a manageable nonstandard version of Szemerédi’s original proof. We didn’t end up being able to do so – in fact there are now signs that perhaps nonstandard analysis is not the optimal framework in which to place this argument – but we did at least clarify the existing standard argument, to the point that I was able to go back to my original rewrite of the proof and present it in a more civilised form, which I am now uploading here as an unpublished preprint. There are now a number of simplifications to the proof. Firstly, one no longer needs the full strength of the regularity lemma; only the simpler “weak” regularity lemma of Frieze and Kannan is required. Secondly, the proof has been “factored” into a number of stand-alone propositions of independent interest, in particular involving just (families of) one-dimensional arithmetic progressions rather than the complicated-looking multidimensional arithmetic progressions that occur so frequently in the original argument of Szemerédi. Finally, the delicate manipulations of densities and epsilons via double counting arguments in Szemerédi’s original paper have been abstracted into a certain key property of families of arithmetic progressions that I call the “double counting property”.

The factoring mentioned above is particularly simple in the case of proving Roth’s theorem, which is now presented separately in the above writeup. Roth’s theorem seeks to locate a length three progression in which all three elements lie in a single set. This will be deduced from an easier variant of the theorem in which one locates (a family of) length three progressions in which just the first two elements of the progression lie in a good set (and some other properties of the family are also required). This is in turn derived from an even easier variant in which now just the first element of the progression is required to be in the good set.

More specifically, Roth’s theorem is now deduced from

Theorem 1.5. Let be a natural number, and let be a set of integers of upper density at least . Then, whenever is partitioned into finitely many colour classes, there exists a colour class and a family of 3-term arithmetic progressions with the following properties:

- For each , and lie in .
- For each , lie in .
- The for are in arithmetic progression.

The situation in this theorem is depicted by the following diagram, in which elements of are in blue and elements of are in grey:

Theorem 1.5 is deduced in turn from the following easier variant:

Theorem 1.6. Let be a natural number, and let be a set of integers of upper density at least . Then, whenever is partitioned into finitely many colour classes, there exists a colour class and a family of 3-term arithmetic progressions with the following properties:

- For each , lie in .
- For each , and lie in .
- The for are in arithmetic progression.

The situation here is described by the figure below.

Theorem 1.6 is easy to prove. To derive Theorem 1.5 from Theorem 1.6, or to derive Roth’s theorem from Theorem 1.5, one uses double counting arguments, van der Waerden’s theorem, and the weak regularity lemma, largely as described in this previous blog post; see the writeup for the full details. (I would be interested in seeing a shorter proof of Theorem 1.5 though that did not go through these arguments, and did not use the more powerful theorems of Roth or Szemerédi.)

Fix a non-negative integer . Define an (weak) integer partition of length to be a tuple of non-increasing non-negative integers . (Here our partitions are “weak” in the sense that we allow some parts of the partition to be zero. Henceforth we will omit the modifier “weak”, as we will not need to consider the more usual notion of “strong” partitions.) To each such partition , one can associate a Young diagram consisting of left-justified rows of boxes, with the row containing boxes. A semi-standard Young tableau (or *Young tableau* for short) of shape is a filling of these boxes by integers in that is weakly increasing along rows (moving rightwards) and strictly increasing along columns (moving downwards). The collection of such tableaux will be denoted . The *weight* of a tableau is the tuple , where is the number of occurrences of the integer in the tableau. For instance, if and , an example of a Young tableau of shape would be

The weight here would be .

To each partition one can associate the Schur polynomial on variables , which we will define as

using the multinomial convention

Thus for instance the Young tableau given above would contribute a term to the Schur polynomial . In the case of partitions of the form , the Schur polynomial is just the complete homogeneous symmetric polynomial of degree on variables:

thus for instance

Schur polyomials are ubiquitous in the algebraic combinatorics of “type objects” such as the symmetric group , the general linear group , or the unitary group . For instance, one can view as the character of an irreducible polynomial representation of associated with the partition . However, we will not focus on these interpretations of Schur polynomials in this post.

This definition of Schur polynomials allows for a way to describe the polynomials recursively. If and is a Young tableau of shape , taking values in , one can form a sub-tableau of some shape by removing all the appearances of (which, among other things, necessarily deletes the row). For instance, with as in the previous example, the sub-tableau would be

and the reduced partition in this case is . As Young tableaux are required to be strictly increasing down columns, we can see that the reduced partition must *intersperse* the original partition in the sense that

for all ; we denote this interspersion relation as (though we caution that this is *not* intended to be a partial ordering). In the converse direction, if and is a Young tableau with shape with entries in , one can form a Young tableau with shape and entries in by appending to an entry of in all the boxes that appear in the shape but not the shape. This one-to-one correspondence leads to the recursion

where , , and the size of a partition is defined as .

One can use this recursion (2) to prove some further standard identities for Schur polynomials, such as the determinant identity

for , where denotes the Vandermonde determinant

with the convention that if is negative. Thus for instance

We review the (standard) derivation of these identities via (2) below the fold. Among other things, these identities show that the Schur polynomials are symmetric, which is not immediately obvious from their definition.

One can also iterate (2) to write

where the sum is over all tuples , where each is a partition of length that intersperses the next partition , with set equal to . We will call such a tuple an *integral Gelfand-Tsetlin pattern* based at .

One can generalise (6) by introducing the skew Schur functions

for , whenever is a partition of length and a partition of length for some , thus the Schur polynomial is also the skew Schur polynomial with . (One could relabel the variables here to be something like instead, but this labeling seems slightly more natural, particularly in view of identities such as (8) below.)

By construction, we have the decomposition

whenever , and are partitions of lengths respectively. This gives another recursive way to understand Schur polynomials and skew Schur polynomials. For instance, one can use it to establish the generalised Jacobi-Trudi identity

with the convention that for larger than the length of ; we do this below the fold.

The Schur polynomials (and skew Schur polynomials) are “discretised” (or “quantised”) in the sense that their parameters are required to be integer-valued, and their definition similarly involves summation over a discrete set. It turns out that there are “continuous” (or “classical”) analogues of these functions, in which the parameters now take real values rather than integers, and are defined via integration rather than summation. One can view these continuous analogues as a “semiclassical limit” of their discrete counterparts, in a manner that can be made precise using the machinery of geometric quantisation, but we will not do so here.

The continuous analogues can be defined as follows. Define a *real partition* of length to be a tuple where are now real numbers. We can define the relation of interspersion between a length real partition and a length real partition precisely as before, by requiring that the inequalities (1) hold for all . We can then define the continuous Schur functions for recursively by defining

for and of length , where and the integral is with respect to -dimensional Lebesgue measure, and as before. Thus for instance

and

More generally, we can define the continuous skew Schur functions for of length , of length , and recursively by defining

and

for . Thus for instance

and

By expanding out the recursion, one obtains the analogue

of (6), and more generally one has

We will call the tuples in the first integral *real Gelfand-Tsetlin patterns* based at . The analogue of (8) is then

where the integral is over all real partitions of length , with Lebesgue measure.

By approximating various integrals by their Riemann sums, one can relate the continuous Schur functions to their discrete counterparts by the limiting formula

as for any length real partition and any , where

and

More generally, one has

as for any length real partition , any length real partition with , and any .

As a consequence of these limiting formulae, one expects all of the discrete identities above to have continuous counterparts. This is indeed the case; below the fold we shall prove the discrete and continuous identities in parallel. These are not new results by any means, but I was not able to locate a good place in the literature where they are explicitly written down, so I thought I would try to do so here (primarily for my own internal reference, but perhaps the calculations will be worthwhile to some others also).

## Recent Comments