Welsh s problem on the number of bases of matroids

Transcription

1 Welsh s problem on the number of bases of matroids Edward S. T. Fan 1 and Tony W. H. Wong 2 1 Department of Mathematics, California Institute of Technology 2 Department of Mathematics, Kutztown University of Pennsylvania February 24, 2014 Abstract In this paper, we study a problem raised by Dominic Welsh on the existence of matroids with prescribed size, ran, and number of bases. In case of coran being at most, by an explicit construction of matrices, we are able to show the existence of such matroids, except the counterexample found by Mayhew and Royle. This idea also extends to give asymptotic results on the existence of matroids with any fixed coran. 1 Bacground Let (n, r, b be a triple of integers such that 0 < r n and 1 b ( n r. Given a triple (n, r, b, Welsh [?] ased if there exists a matroid with n elements, ran r, and exactly b bases. Such a matroid would be called an (n, r, b-matroid. It was conjectured that an (n, r, b-matroid exists for every such triple, until Mayhew and Royle [?] found the lone counterexample to date, namely (n, r, b = (6,, 11. However, they proposed the following conjecture. Conjecture 1.1. An (n, r, b-matroid exists for all 0 < r n and 1 b ( n r except (n, r, b = (6,, 11. In this paper, we verify this conjecture for a large family of triples (n, r, b. 2 Introduction A matroid (X, I is a combinatorial structure defined on a finite ground set X of n elements, together with a family I of subsets of X called independent sets, satisfying the following three properties: sfan@caltech.edu wong@utztown.edu 1

2 1. I, or I is nonempty. 2. If I I and J I, then J I. This is sometimes nown as hereditary property.. If I, J I and J < I, then there exists x I\J such that J {x} I. This is nown as augmentation property or exchange property. A maximal independent set is called a basis of the matroid. A direct consequence of the augmentation property is that all bases have the same cardinality. This cardinality is defined as the ran of M, and the difference between the size of X and the ran of M is defined as the coran of M. One of the most important examples of matroids is a linear matroid. A linear matroid is defined from a matrix A over a field F, where X is the set of columns of A, and an independent set I I is a collection of columns which is linearly independent over F. For a linear matroid, the ran is exactly the ran of the matrix A, and the coran is the difference between the number of columns and the ran of A. Since the elements of a linear matroid are the columns of a matrix, a linear matroid is also called a column matroid. In this paper, we are going to show that a linear (n, r, b-matroid exists for a large family of triples (n, r, b. Throughout this paper, A is an r n matrix over Q with ran r and coran = n r, and the number of r r invertible submatrices of A is b. Note that b is an invariant if we perform row operations on A or permute the columns of A, so we can always assume that A = (I r M, where M is an r matrix. We start with the following observation that there is a natural duality between the ran and coran of a linear matroid. Proposition 2.1. The number of invertible square submatrices of M is b, and the number of invertible submatrices of (I M is also b. (Here, we adopt the convention that an empty matrix is invertible. Proof. Let S be the set of all invertible r r submatrices of A, where r 0, and let T be the set of all invertible square submatrices of M. Let S be a matrix in S with columns i 1, i 2,..., i r, where i 1,..., i j r and i j+1,..., i r > r. Then there is a bijection between S and T which sends S to the square submatrix of M with rows {1, 2,..., r}\{i 1, i 2,..., i j } and columns i j+1 r, i j+2 r,..., i r r. Hence, the number of invertible square submatrices of M is b. It is then obvious that the number of invertible square submatrices of M is b, and the above bijective argument in turn implies that the number of invertible submatrices of (I M is b. In view of this proposition, to determine the existence of a linear (n, r, b-matroid, we would try to construct an r matrix M with exactly b invertible square submatrices, where = n r. Such a matrix M will be called an (r,, b-matrix. Note that we only need to consider the case when r for our purpose of studying this Welsh s problem, since if > r, we can instead consider the existence of an (, r, b-matrix. Also note that the existence of a linear (n, r, b-matroid is equivalent to the existence of an (r,, b-matrix. Next, we observe that the naive extension by zero construction exhibits a useful relation between the existences of various (r,, b-matrices. 2

3 Lemma 2.2. If there exists an (r 0, 0, b-matrix, then for all r r 0, 0, (r,, b-matrix exists. Proof. If M 0 is an (r 0, 0, b-matrix, by building an r matrix M such that M 0 is a submatrix and all the extra entries are 0 s, M is an (r,, b-matrix. Combining the results from Sections?? and??, we show that Conjecture?? holds for all 1 b ( r+ r, except that (n, r, b = (6,, 11. In Section??, we prove Conjecture?? for all large r and large b. We strengthen this result in Section?? such that Conjecture?? holds for all large enough r, regardless the values of n and b. This gives the closest to complete answer to the Welsh s problem and provides a plausible reason of the existence of counterexample for small values of r. Existence of linear matroids with coran at most 2 In this section, we will study the existence of (r,, b-matrices for 2. The cases = 0 and = 1 are very straight forward. For the first nontrivial case = 2, the following lemma in number theory will help us to show the existence of matrices satisfying the parameters (r, 2, b. Lemma.1. Let s 5 be a positive integer, and let be a nonnegative integer such that s2 5s. Then there exist nonnegative integers a 4 1, a 2,..., a s such that a 1 +a 2 + +a s = s and a a a 2 s = s + 2. Proof. For 5 s 2, we verified the lemma by Mathematica; for s, we will use strong induction on s. Suppose the statement is true for all integers u such that 5 u < s for some s, i.e., for all nonnegative integers u2 5u, there exist nonnegative integers a 4 1,..., a u such that a a u = u and a a 2 u = u + 2. Let t and be integers such that 0 < t s 5 and 0 t2 t (s t2 5(s t. Then 2 4 u := s t falls in the range 5 u < s, and := t2 t u2 5u. By the induction 2 4 hypothesis, there are nonnegative integers a 1,..., a s t such that a a s t = s t and a a 2 s t = s t+2 ( t2 t 2 = s+2 t 2. If we set a s t+1 = t and a s t+2 = = a s = 0, then a a s = s and a a 2 s = s + 2, implying that the statement holds true for satisfying 0 t2 t (s t2 5(s t, or equivalently, t2 t t2 2st+t+s 2 5s [ It now suffices to show that the union of the intervals I(t := [ ] 0 < t s 5 covers 0, s2 5s 4 t 2 t, t2 2st+t+s 2 5s 2 4 when s. Let α(t = t2 t 2 and β(t = t2 2st+t+s 2 5s 4. ] for Claim 1. s2 5s β(t if and only if t 2 s 1, which is attainable for some t in the range 4 0 < t s 5 if s 12. Proof of claim 1. This inequality holds if and only if t 2 2st + t 0, which is equivalent to t 2 s 1 since t is positive. We finish by noticing that when s 12, s 5 2 s 1.

4 Claim 2. α(t 1 α(t β(t. Proof of claim 2. The first inequality holds since α(t is an increasing function for t 1, since α (t = 2t 1 > 0 when t 1, considering α as a continuous function on R. The second 2 inequality holds if and only if 5(s t (s t 2, which is always true since t s 5. Claim. α(t β(t 1 if and only if t 2s+1 16s+1 2. Proof of claim. This inequality holds if and only if t 2 (2s + 1t + s 2 s 0, which occurs if and only if t 2s+1 16s+1 or t 2s+1+ 16s+1. However, t 2s+1+ 16s+1 is rejected since t < s. By claims 2 and, if t 2s+1 16s+1, then I(1 I(t 1 I(t forms one closed 2 interval. If [ ] 2 2 s 1 2s+1 16s+1 s 1, then claim 1 implies that 0, s2 5s I(t. 2 4 To obtain 2 s 1 2s+1 16s+1, we loo for integers s satisfying 2s 2s+1 16s+1, or 2 2 equivalently, 16s + 1 2s +. This inequality holds if s s 2, or s. Theorem.2. If 2, then for all integers b such that 1 b ( r+ r, there exists an (r,, b-matrix M. Proof. It is trivial for = 0. If = 1, then M is a column vector with the first b 1 entries 1 s and the rest 0 s. If = 2, let the first column of M have the first s entries 1 s, the second column have the first s entries nonzero, and the rest be all 0 s. Furthermore, assume that there are a i i s in the second column, 1 i s, where a 1 + a a s = s. Then the number of invertible square submatrices of M is 1 + 2s + i<j a ia j = 1 + 2s + 1( 2 i a i 2 1( 2 i a2 i = 1 + 2s + s2 1( 2 i a2 i, and we would lie to set it to be b, which gives s + 2( ( s+2 2 b = i a2 i. By Lemma??, if 0 ( s+2 2 b s 2 5s, or equivalently s2 +11s+4 b s2 +s+2, there is a solution for a i s. It is easy to chec that the intervals [ s2 +11s+4, s2 +s+2] cover all integers 4 2 b 9, and the only missing integers are in [1, 20] [22, 26] [29, 2] [7, 8]. Here, we finish the proof by constructing M explicitly for each of these b s. In the following table, 0 represents a column vector of all 0 s (possibly of length 0, which fills up the column so that M has r rows. b = t=

5 b = b = b = Theorem.. Conjecture?? holds for all 1 b ( r+2 r. Proof. By Theorem?? and Lemma??, for all 1 b ( r+2 r, for all r and such that b, (r,, b-matrices and hence linear (n, r, b-matroids exist. ( r+ r 4 Existence of linear matroids with ample amount of bases In Section??, we used an induction argument together with computational exhaustion to show the existence of linear (n, r, b-matroids with 1 b ( r+2 r. In this section, we will establish an asymptotic existence result by studying a special class of matrices obtained from a generic hyperplane arrangement. A generic hyperplane arrangement is a finite collection H = {Γ 1, Γ 2,..., Γ l } of ( 1-dimensional linear subspaces of Q satisfying 5

6 1. The hyperplanes Γ j s do not contain any coordinate directions, i.e. e i / Γ j for all i = 1, 2,..., and j = 1, 2,..., l, where e i = (0,..., 1,..., 0 is the i-th coordinate vector. 2. No hyperplanes are of the form {(x 1, x 2,..., x : x i = x j } for some i j.. For all j = 1, 2,..., min(, l, the intersection of any j hyperplanes from H is a ( j- dimensional subspace. Given a generic hyperplane arrangement H together with integers n 0, n 1,..., n l, a set of generic (n 0, n 1,..., n l -vectors from H is a collection of vectors, where n j of them come from Γ j and n 0 of them from the complement of the union of the hyperplanes, such that 1. No vector lies in more than one hyperplane. 2. Any 1 vectors are linearly independent.. Any vectors are linearly dependent only when they are all chosen from the same hyperplane. It is well nown that such generic hyperplane arrangements and sets of generic vectors of any size always exist. The former can be shown by choosing normal vectors in general positions and the latter follows easily from an inductive construction. Now, given nonnegative integers a, a +1,..., a r with r s= a s s r, we first pic a generic hyperplane arrangement H of size l = r a s. Then we choose a set of generic (n 0, n 1,..., n l - s= vectors from H, where n 0 = r r a s s, and among n 1, n 2,..., n l, there are exactly a s copies s= of s for each s =, + 1,..., r. If we put these vectors together as row vectors of a matrix, we get an r matrix M. This construction builds an (r,, b-matrix M with b = ( n r ( r b, where b = a s s denotes the number of singular r r submatrices in A, or the number of singular square submatrices in M. In the following proposition, we determine the range of b that is achievable by our construction. Proposition 4.1. For each, there exists r 0 N such that for all integers r r 0, and all 0 b ( r+ 1 1, there exist nonnegative integers a, a +1,..., a r such that r ( (i a s s = b, and (ii s= r a s s r. s= s= 6

7 Proof. Let κ be a real number greater than ( Then κ > ( > 2. Let b 0 = b. For each integer i 0, if b i ( κ, choose si N such that ( s i bi < ( s i +1. Note that 1 b i / ( s i < si +1 / si which is less than 2 since si +1 2 si if and only if s i 2 1. Let a si = 1, and b i+1 = b i ( s i. If bi < ( κ for some i 0, then let a = b i and all other undetermined a s be 0. r From this definition, condition (i is clearly satisfied. For condition (ii, a s s = s 0 + s= s s m + b m+1 for some m 0. To give an upper bound to this sum, let f (x = ( x and f 1 (x = ( x 1. Note that they are both strictly increasing functions when x, so f 1 f 1 is also a strictly increasing function. For all i 0, b i+1 = b i ( s i < si +1 si = si 1. Hence, s0 (f 1 f 1(r + 1 and s i+1 f 1 (b i+1 < (f 1 f 1(s i. When x κ > 2, f 1 f ( 1(x < f 1 x 1 = f 1 ( 1/ x ( 1/ < 1 1 x + 1, ( 1! which is less than ( 1 +1x 1 when x > ( 1 1. Since 2 1 = (1+1 1 > 1+ 1 > 1, we have ( 1 < 2, implying 2 > ( 1 1. Therefore, for all x 2. In particular, and for i 1, s i (f 1 f 1 f 1(x < ( x! s 0 < ( (r + 1 1, (1 f 1 i+1 (r + 1 < ( ( ( 1 ( (r i+1. i (r i+1 As (f 1 f 1(s m < κ, we would lie to find the minimum integer m such that This is equivalent to or (m + 2 log ( 1 m ( (r m +2 κ. log log ( ( log log(r + 1 log log ( κ ( 1/ +1 log log(r + 1, / κ log ( ( 1/ +1 Combining (?? and (??, we have s 0 + s 1 + s m + b m+1 ms 0 + ( κ log log r for fixed, which is less than r when r is large. O(r (2 m s 0 + ( κ = This gives the following immediate asymptotic result on the existence of linear (n, r, b- matroids. 7

8 Theorem 4.2. For each fixed, let r 0 N be as defined in Proposition??. Then for all integers r r 0, for all integers b such that b ( r 0 + 1, linear (n, r, b-matroids always exist, where n = r +. Proof. By Proposition??, for every integer r r 0, for all integers b such that ( r+ 1 = ( r+ r+ 1 1 b r+, we can construct an (r,, b-matrix following the procedures introduced [( at the beginning of Section??. Finally, we are done by noticing that the intervals r0 + 1+i, r0 ] ++i, i = 0, 1,..., r r0, cover all integers ( r b n r. In fact, we can remove the lower bound on the values of b in Theorem??. Theorem 4.. For each fixed, there exists R N such that for all r R, linear (n, r, b-matroids always exist, where n = r +. The proof of this theorem will be shown at the end of Section??. 5 Existence of linear matroids with coran at most In this section, we will show that when =, there always exists a linear (r +, r, b-matroid except (n, r, b = (6,, 11. For 4 r 10, the existence of an (r,, b-matrix will be verified through explicit constructions, which is described in Appendix??. Then we will base on the method employed in Proposition?? and modify it into an inductive argument. Proposition 5.1. For each r 49 and for all 0 b ( r+2 2, there exist nonnegative integers a, a 4,..., a r such that r ( (i a s s = b, and s= r (ii a s s r. s= Proof. Let the statement of the proposition be denoted by P(r for r 49. In Appendix??, we show that P(r holds for 49 r 20. Suppose that P(r is true for all 49 r r 0 for some r We would lie to chec the validity of P(r Let b be an integer such that 0 b ( r If b r0 +2 2, we can apply P(r0 to b to obtain nonnegative integers a, a 4,..., a r0 and a r0 +1 = 0 such that (i r 0+1 ( a s s = b, and s= (ii r 0+1 a s s r 0 r 0 + 1, s= so it suffices to consider ( r < b r Note that there exists a unique integer s0 such that ( s 0 b < s0 +1. Set b := b ( s 0 (. If b = 0, we are done. If b 1, we have 1 b < s0 +1 s0 = s0 2. Clearly, s 0 < r 0 + 2, so we can apply the induction hypothesis P(s 0 2 to b as long as s From this, we can find nonnegative integers a, a 4,..., a s0 2 with 8

9 ( a s s = b, and (i s 0 2 s= (ii s 0 2 s= a s s s 0 2. By setting a s0 1 = 0, a s0 = 1, and a s0 +1 = = a r0 +1 = 0, we have (i r 0+1 ( a s s = b, and s= (ii r 0+1 s= a s s 2s 0 2. It suffices to show that s 0 51 and 2s 0 2 r First, observe that for r 0 20, we have ( s0 +1 > b > r = 20910, or equivalently, (s 0 + 1s 0 (s 0 1 > A straight forward calculation shows that it holds if and only if s Next, for any integer x 1, ( x+ 2 x+ 2 = (x+(x+1(x 1 (x+(x = 1 (x + 48 (x2 24x 49 0 if and only if x 2 24x By solving the quadratic inequality, it is easy to see that it holds for all integers x 26. Since r 0 20 and ( x is a strictly increasing function for x, we have ( s0 r0 + r0 + b 2, 2 which implies r 0+ 2 s 0 or 2s 0 2 r To complete the case for =, we still need to consider r = and 11 r 48. When r =, Theorem?? implies that linear (6,, b-matroid exists for 1 b 10, and the following constructions produce the (,, b-matrices M, where 12 b 20. b = b =

10 As for 11 r 48, we have to slightly modify the generic hyperplane arrangement construction as described in Section?? to reduce the number of rows r needed to produce a matrix M with a prescribed number of singular submatrices b. For simplicity, we will only show the construction for =, but the corresponding one for general is similar. There are three types of modification: 1. Allow some of the 2-planes to contain exactly one coordinate direction, i.e. 2-planes with equations ax + by = 0, by + cz = 0, or ax + cz = 0, where a, b, c 0, and pic vectors from these 2-planes on the same line parallel but distinct to the axis. Each of these modified 2-planes increases b by changing a summand of ( s to s+1. In other words, we save one row every time we modify a 2-plane. 2. Tae one to three 2-planes to be those orthogonal to a coordinate axis, i.e. x = 0, y = 0, or z = 0. The resulting b will increase by changing a summand of ( s to s+2. In other words, we can save two rows for up to three times.. For each vector chosen from the complement of the union of hyperplanes, we duplicate ( them t times in the matrix M. Each such duplication will increase the resulting b by t + t 2 (r t +. Proposition 5.2. For each r such that 11 r 48, and for each 0 b ( r+2 2, there exists an ( r,, ( r+ r b -matrix M. Proof. For each r and b such that 11 r 48 and 0 b ( r+2 2, let b0 = b. For each integer i 0, let t i be largest integer such that ( t i ( + ti 2 (r ti + b, let c i = /(( ti bi + ti 2 (r ti + ((, and let b i+1 = b i c ti i + ti 2 (r ti +. We repeat the process until b i+1 < ( ( (r 2 + = r + 1. In this process, we are building an r matrix M using only rows from the Type modification. After using only rows from Type modification, we still need to pic appropriate rows to give an additional b i+1 singular submatrices if b i+1 > 0. Note that b i+1 r 48. The following is a list of partitions of integers from 1 to 48. Each summand ( s+2 corresponds to s rows from Type 2 modification, and summand ( s+1 corresponds to s rows from Type 1 modification. 10

11 1 = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( +2 4 = ( = ( = ( 12 = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( = ( 44 = ( = ( = ( = ( = ( = ( = ( = ( = ( From this table, we can deduce directly the number of additional rows from Type 1 and Type 2 modifications we need to obtain b singular square submatrices in M. Together with the rows from Type modification, it remains to verify that the total number of rows we have used is at most r. Once again, we employ Mathematica to finish the verification, and the program codes are provided in Appendix?? for reference. Theorem 5.. Linear (n, r, b-matroids exist for all 1 b ( r+ r, as long as n r b, except (n, r, b = (6,, 11. Proof. Propositions?? and?? imply that for all r 11, linear (n, r, b-matroids exist for all ( r+2 = r+ r+2 2 b r+. Hence, we are done by connecting with the explicit constructions for r 10. Now, we have all the tools for proving Theorem??. Proof of Theorem??. For each i = 0, 1, 2,...,, let r 0 ( i N be the constants obtained by applying Proposition?? on i. Note that r 0 ( r 0 ( 1 r 0 (. Let R 0 = r 0 (. Then there exist integers R 1, R 2,..., R such that 1. R 0 R 1 R 2 R, and 11

12 2. ( R i + i 1 i Ri+1 + (i+1 for all i = 0, 1, 2,..., 4. (i+1 Let R = R, and fix r R. By Theorem??, an (r,, b-matrix exists for all integers b such that ( R b r+ r. By Proposition??, for each i = 1, 2,...,, since R i r 0 ( i, an (R i, i, b-matrix exists for all integers b such that ( R i + i 1 i b ( Ri + i i. By Lemma??, for each i = 1, 2,...,, an (r,, b-matrix exists for all b such that ( R i + i 1 i b Ri + i i. Our definition of Ri s implies that [( R 0 + 1, r+ ] r [( Ri + i 1 i, Ri ] ( + i i covers all integers R+2 b r+ r. Therefore, an (r,, b-matrix i=1 exists for all b ( R+2. ( R+ Finally, we are done by Theorem??, which says an (r,, b-matrix exists for all b. 6 Conclusion and remars In our treatment, we always assume the base field is Q. In fact, the same argument wors for any infinite field. In particular, if we consider the algebraic closure F p for some prime p, the constructed (r,, b-matrix naturally descends to some finite extension of F p. However, we cannot ensure there is a fixed finite extension which captures all of them. In any case, the matroid structure arising from matrices will not be affected. All our results focus on the situation when n and r are very close together. Recall from Lemma??, we only need to consider r n 2r. Hence, our next goal is to investigate the case when n is close to 2r. In view of the counterexample of the non-existence of (6,, 11- matroids, this latter goal should be much harder. But the general direction towards a complete solution to Conjecture?? will be another asymptotic result for the existence of (r,, b-matrices for large enough which should isolate a finite number of cases for direct checing. A Construction of (r +,, b-matroids for 4 r 10 If we can construct a matrix A = (I r M, where M is an r matrix, such that A has exactly b invertible r r submatrices, then for all r r, the matrix A = (I r M, where M is obtained by appending r r rows of all zeros to M, has exactly b invertible r r submatrices. Hence, for r 5, we only need to construct matrices with b invertible r r submatrices, where ( r+2 < b r+. Theorem?? implies that when r = 4 and 1 b ( r+2 2 = 15, there exists a matrix A of size r (r + such that the number of invertible r r submatrices is b. When r = 4 and 15 < b ( r+ = 5, we give the construction of A = (Ir M as follows. 12

13 b = b = b = When r = 5 and 5 < b ( r+ = 56, we give the construction of A = (Ir M as follows. b = b = b =

14 When r = 6 and 56 < b ( r+ = 84, we give the construction of A = (Ir M as follows. b = b = b = b = When r = 7 and 84 < b ( r+ = 120, we give the construction of A = (Ir M as follows. 14

15 b = b = b = b =

16 b = b = When r = 8 and 120 < b ( r+ = 165, we give the construction of A = (Ir M as follows. b = b =

17 b = b = b = b =

18 b = When r = 9 and 165 < b ( r+ = 220, we give the construction of A = (Ir M as follows. b = b = b =

19 b = b = b = b =

20 b = When r = 10 and 220 < b ( r+ = 286, we give the construction of A = (Ir M as follows. b = b =

21 b = b = b =

22 b = b = b =

23 b = b = B Proof of Proposition?? for 49 r 20 We use Mathematica to chec the validity of Proposition?? for 49 r 20. IntDiv[n, d ] := Bloc[{}, (n - Mod[n, d]/d]; Do[bin = Table[Binomial[s, ], {s, r}]; a = Table[0, {s, r}]; Do[btemp = bbar; Do[a[[s]] = IntDiv[btemp, bin[[s]]]; btemp = Mod[btemp, bin[[s]]], {s, r,, -1}]; If[(Sum[a[[s]]*s, {s,, r}] > r, Print[r, " ", bbar, " ", Sum[a[[s]]*s, {s,, r}]]], {bbar, 0, Binomial[r + 2, 2]}], {r, 49, 20}] As there is no output after these lines finish running, we finish our verification. 2

24 C Proof of Proposition?? for 11 r 48 From the list in Proposition??, we use leng below to record the number of rows from Type 1 and Type 2 modifications to obtain additional singular submatrices. leng = {0, 1, 2,, 2,, 4, 6, 4, 5,, 4, 5, 7, 5, 6, 8, 10, 7, 9, 4, 5, 6, 8, 6, 7, 9, 11, 8, 10, 7, 8, 10, 12, 9, 5, 6, 7, 9, 7, 8, 10, 12, 9, 11, 8, 9, 11, 1}; Do[tbin = Table[Binomial[t, ] + Binomial[t, 2] (r - t +, {t, r}]; a = Table[0, {s, r}]; max = 0; Do[btemp = bbar; Do[a[[s]] = IntDiv[btemp, tbin[[s]]]; btemp = Mod[btemp, tbin[[s]]], {s, r, 2, -1}]; If[Sum[a[[s]]*s, {s, 2, r}] + leng[[btemp + 1]] > max, max = Sum[a[[s]]*s, {s, 2, r}] + leng[[btemp + 1]]], {bbar, Binomial[r + 2, 2]}]; If[r < max, Print[r, " ", max]], {r, 11, 48}] As there is no output after these lines finish running, we finish our verification. References [1] D. Mayhew and G. F. Royle, Matroids with nine elements, J. Combinatorial Theory Ser. B 98 (2008, [2] V. Sivaraman, On a conjecture of Welsh, preprint. [] D. Welsh, Combinatorial problems in matroid theory, Combinatorial Math. and its Applications (Proc. Conf. Oxford 1969, Academic Press, London,