Abstract
Let f be an arithmetic function and S= {x1, …, xn} be a set of n distinct positive integers. By (f(xi, xj)) (resp. (f[xi, xj])) we denote the n × n matrix having f evaluated at the greatest common divisor (xi, xj) (resp. the least common multiple [xi, xj]) of x, and xj as its (i, j)-entry, respectively. The set S is said to be gcd closed if (xi, xj) ∈ S for 1 ≤ i, j ≤ n. In this paper, we give formulas for the determinants of the matrices (f(xi, xj)) and (f[xi, xj]) if S consists of multiple coprime gcd-closed sets (i.e., S equals the union of S1, …, Sk with k ≥ 1 being an integer and S1, …, Sk being gcd-closed sets such that (lcm(Si), lcm(Sj)) = 1 for all 1 ≤ i ≠ j ≤ k). This extends the Bourque-Ligh, Hong’s and the Hong-Loewy formulas obtained in 1993, 2002 and 2011, respectively. It also generalizes the famous Smith’s determinant.
1 Introduction and statements of main results
Let n be a positive integer and f be an arithmetic function. Let S = {x1, …, xn} be a set of n distinct positive integers. We denote by (f(S)) = (f(xi, xj)) and (f[S]) = (f[xi , xj]) the n × n matrices having f evaluated at the greatest common divisor (xi, xj) and the least common multiple [xi, xj] of xi and xj as their (i, j)-entries, respectively. In 1875, Smith 25 published his famous result stating that
In 1989, Beslin and Ligh 3 extended Smith’s result by showing that
In 1993, Bourque and Ligh 5 generalized Smith’s determinant and Beslin and Ligh’s result 3 by proving that if S is gcd closed, then
In 2002, Hong 13 extended the Bourque-Ligh result by showing that
We say that S consists of multiple coprime gcd-closed sets if there is a positive integer h and h distinct gcd-closed sets S1,…, Sh with (lcm(Si), lcm(Sj)) = 1 for all integers i and j with 1 ≤ i ≠ j ≤ h such that S can be partitioned as the union of S1, …, Sh (see, for instance, 15). Clearly, if S consists of multiple coprime gcd-closed sets, then either we have 1 ∈ S or 1 ∉ S. For the former case 1 ∈ S, S is gcd closed and the formulas for determinants of the matrices (f(S)) and (f[S]) were given by Bourque and Ligh 5 and Hong 13, respectively. For the latter case 1 ∉ S, the formulas for determinants of the matrices (f(S)) and (f[S]) are unknown. This problem is still kept open so far.
In this paper, our main goal is to introduce a new method to investigate the above problem. Actually, we first give the formula for the determinant of (f(S)) on any positive integers set S. Then we present formulas for the determinants of the matrices (f(S)) and (f[S]) on the multiple coprime gcd-closed sets S. Evidently, any rearrangement of the elements of S yields matrices similar to the matrices (f(S)) and (f[S]). So we can rearrange the elements of S in any case of necessity. To give the main result, we need two concepts as follows.
Definition 1.1. Let f be an arithmetic function and T be a set of distinct positive integers. Then associated to f, we define the functions
Definition 1.2. Let S consist of h coprime gcd-closed sets S1, …, Sh. Then we define the set of minimal elements of S, denoted by M(S), to be
For example, if S ={2, 5, 6, 8, 11, 35, 143}, then S consists of three coprime gcd-closed sets and the set M(S) of minimal elements of S is equal to {2, 5, 11}.
Now we can state the main result of this paper.
Theorem 1.3. Let f be an arithmetic function. Let S consist of multiple coprime gcd-closed sets such that 1 ∉ S and M(S) denote the set of minimal elements of S. Then
Furthermore , if f is a multiplicative function and f(x) ≠ 0 for all x ∈ S, then
If letting S be a gcd-closed set, then Theorem 1.3 reduces to the Bourque-Ligh theorem 5 and Hong’s theorem 13. If S consists of coprime divisor chains, then Theorem 1.3 becomes the main result of 18. From Theorem 1.3, one can easily deduce the following interesting consequence.
Corollary 1.4. Let S consist of multiple coprime gcd-closed sets and M(S) denote the set of minimal elements of S. Then
and
where I is the arithmetic function defined by I (n) := n.
Obviously, picking S to be a gcd-closed set in Corollary 1.4 gives us the Beslin-Ligh result 3 and the Bourque-Ligh result 4. If S = M(S), then Corollary 1.4 is Lemma 2.1 of 17.
We organize the paper as follows. In Section 2, we present some lemmas which are needed in the proof of Theorem 1.3. In Section 3, we prove Theorem 1.3 and Corollary 1.4.
2 Several lemmas
In this section, we present some useful lemmas that are needed in the next section. The first two lemmas are well known.
Lemma 2.1 (13). Let f be any arithmetic function and n be a positive integer. Then
Lemma 2.2 (24). Let m , n be any positive integers and f be a multiplicative function. Then f(m)f(n) = f((m, n))f([m, n]).
Lemma 2.3. Let g be any arithmetic function and S be gcd closed. Then for any x ∈ S, we have
Proof. Clearly, the terms in the sum of the right-hand side of (2) are non-repetitive. Now we show that the terms in the sum of the left-hand side of (2) are non-repetitive. For this purpose, for any y ∈ S with y | x, we let D(y) = {d ∈Z+ : d | y, d ∤ z, z < y, z ∈ S}. Claim that D(y1) ∩ D(y2) = ϕ for any distinct elements y1 and y2 in the set S satisfying y1| x and y2| x. Otherwise, we may let d ∈ D(y1) ∩ D(y2) . Then d | y1 and d | y2. So d |(y1, y2). But the assumption that S being gcd closed tells us that (y1, y2) = y3 for some y3 ∈ S. Hence d | y3. On the other hand, we have y3 < y1 and y3 < y2 since y1 ≠ y2. It then follows from d ∈ D(y1) that d ∤ y3. We arrive at a contradiction. The claim is proved. By the claim we know immediately that the terms in the sum of the left-hand side of (2) are non-repetitive.
For any term g(d) in the sum of the left-hand side of (2), one has d | y, y| x and y ∈ S. Thus d | x. This implies that g(d) is a term in the sum of the right-hand side of (2). To show that the converse is true, for any given positive integer d and x ∈ S with d | x, we let I(d, x) = {u : d| u, u| x, u ∈ S}. Then I(d, x) ≠ ϕ since x ∈ I(d, x) and I(d, x) is finite. Let υ = min(I(d, x)). Then υ | x, υ ∈ S and d |υ and d ∤ z for any z ∈ S with z < υ. It infers that the term g(d) in the sum of the right-hand side of (2) is also a term in the sum of the left-hand side of (2). So (2) is proved.
This ends the proof of Lemma 2.3. □
Lemma 2.4. Let S be gcd closed. Then for any x ∈ S,
Note that a special case of Lemma 2.3 is due to Beslin and Ligh 3 and a more general form is given in (3.4) of 10.
Proof. Letting g = f * μ in Lemma 2.3 gives us that
Then the desired result follows from the definition of αS, f (d) and Lemma 2.1. This completes the proof of Lemma 2.4. □
We need the following definition to state Lemma 2.6 below.
Definition 2.5. Let S = {x1, …, xn} be a set of positive integers and S = {y1, …, ym} be the minimal gcd-closed set containing S. Then we define the n × m matrix E(S) = (eij) by
For 1 ≤ l ≤ m , we define El (S) to be the n × (m – 1) matrix obtained from E (S) by deleting its lth column.
We can now use the gcd-closed set to describe the structure of the matrix (f(S)) on any set S of positive integers.
Lemma 2.6. Let f be an arithmetic function and S = {xi, …., xn} be a set of distinct positive integers and S = {y1, …. , ym} be the minimal gcd-closed set containing S. Then (f(S)) = E(S) · diag(αS, f (y1), …, αS, f (ym)) · E(S)T.
Proof. Let S = {x1, …, xn} and ∆ = diag(αS, f(y1), …, αS, f (ym)). Then for any integers i and j with 1 ≤ i, j ≤ n, we have
Since S is the minimal gcd-closed set containing S, one has (xi, xj) ∈ S. Then there exists one element yh ∈ S such that yh = (xi, xj). It follows that
But Lemma 2.4 together with the fact that S being gcd closed implies that
Thus by (3) and (4), one has (E(S)∆E(S)T)ij = (f(S))ij as desired. This completes the proof of Lemma 2.6. □
Li 21, Hong 12 and Mattila and Haukkanen 22 made use of the Cauchy-Binet formula to the Smith’s matrices. Now we use this renowned formula to show the following lemma.
Lemma 2.7. Let f be an arithmetic function and S = {x1, …., xn} be a set of n distinct positive integers , and S = {y1, …, ym} be the minimal gcd-closed set containing S. Then
with E(S)(k1, …, kn)being then × n matrix whose columns are the k1th, … , knth columns of E(S).
Proof. Let A = E(S) ·
where A(k1, …, kn) is the n × n matrix whose columns are the k1th, …, knth column of A. One can easily check that
It then follows that
So the desired formula (5) follows immediately. This finishes the proof of Lemma 2.7. □
In what follows, we write
Let S: = S ∪{1} = {x11,…, x1n1,…, xh1,…, xhnh, 1}. Clearly S is the minimal gcd-closed set containing S.
Lemma 2.8. Let S be as in (6) and t be a given integer such that 1 ≤ t ≤ h. Let lt = n1 + … + nt. Let nt ≥ 2. Then each of the following is true.
(i) Ifxt , nt –1 does not divide xtnt, then det(Elt (S)) = det(Elt –1(S \ {xt,nt –1})).
(ii) Ifxt nt –1dividesxtnt, then
Proof. Since S is as in (6), by the definition of E(S) we have
where for 1 ≤ l ≤ h, one has
with
But Elt(S) is the lh × lh matrix obtained from E(S) by deleting its ltth column. So one has
where
(i). xt, nt–1. ∤ xtnt. Then one has that
with
On the other hand, by the definition of S, one can easily deduce that S \ {xt, nt –1} consists of multiple coprime gcd-closed sets and S \ {xt, nt –1} is the minimal gcd-closed set containing the set S \ {xt, nt –1}. Hence by the definition of Elt – 1(S \ {xt, nt –1}), one knows that the right-hand side of (8) is equal to det(Elt – 1(S \ {xt, nt –1})). So the desired result follows. Part (i) is proved.
(ii). xt, nt–1. | xtnt. Thus
With
Clearly S \ {xt, nt} consists of multiple coprime gcd-closed sets and S \ {xt, nt} is the minimal gcd-closed set containing S \ {xt, nt}. Thus by the definition of Elt–1(S \ {xt, nt}), we know that the right-hand side of (9) is equal to det(Elt–1(S \ {xt,nt–1})) – det(Elt–1(S \ {xt, nt})) So part (ii) is true.
This concludes the proof of Lemma 2.8. □
In ending this section, we show the following relation between
Lemma 2.9. Let f be an arithmetic function and T be a set of distinct positive integers. If f(x) ≠ 0 for any x ∈ T and f(x) = 1 then one has that
Proof. Since f(x) ≠ 0 for any x ∈ T and f(x) = 1, it follows that
as desired. So Lemma 2.9 is proved. □
3 Proofs of Theorem 1.3 and Corollary 1.4
In this section, we prove Theorem 1.3 and Corollary 1.4. We begin with the proof of Theorem 1.3.
Proof of Theorem 1.3. Since S consists of multiple coprime gcd-closed sets such that 1 ∉ S, one may write S as in the form of (6). For 1 ≤ i ≤ h , let li = n1 + … + ni. Then the lh × (lh + 1) matrix E(S) is of the form (7).
Let’s first deal with det(f(S)). Define x1 = x11, …, xl1 = x1n1, …, xlh – 1 + 1 = xh1, …, xlh = xhnh, xlh + 1 = 1 and l0 = 0. Lemma 2.7 tells us that
For any {k1,…, klh} with 1 ≤k1 < … < kh < lh+ 1, write {k1,…, klh} := {1,2,…, lh + 1 } \ {k}. Then det(E(S)(k1,…, kth)) = det(Ek (S)).
We claim that (det(E(S)(k1 ,…, kt)))2 = 1 if k = li + 1 for some i ∈ {0, 1, …, h}, and det(E(S)(k1,…,klh)) = 0 if
First, let k = li + 1 for some integer i with 0 ≤ i ≤ h. If 0 ≤ i ≤ h– 1, then the li + 1 row of Eli+ 1(S) is (0,…, 0,1). So the Laplace expansion applied to det(Eli+ 1(S)) gives us a lower triangular determinant with all the diagonal elements being 1. It follows that
If i = h, then Elh + 1(S) is a lower triangular matrix with all the diagonal elements being 1. Hence det (E(S)(k1),…,klh))= det(Elh + 1(S)) = 1. Therefore the first part of the claim is true.
Obviously, the second part of the claim is equivalent to the statement that det(Et (S)) = 0 for all integers t withli–1 + 2 ≤ t ≤ li and 1 ≤i ≤h, which will be proved in the following.
Given any integer i with 1 ≤i ≤h. We prove the claim by using induction on ni. If ni = 2, then t = li = li–1 + 2. Further,xi1xi2. Using Lemma 2.8, wehavedet(Eli(S)) = det(Eli + 1(S\{xi1}))–det(Eli –1(S)\{xi2})). Since ni = 2, we derive that Eii – 1(S\{xi1}) = Eii – 1(S\{xi2}). This implies immediately that det(Eli (S)) = 0. So the claim is true ifni = 2.
Let ni ≥ 3 and assume that the claim is true for the ni – 1 case. In what follows we consider the ni case. For li–1 + 2 ≤ t ≤ li – 1, noting that all elements of the (li – 1)-th column of Et (S) are zero except for its (li,li – 1)-entry is 1, applying the Laplace theorem to det(Et (S)) gives that det(Et (S)) = – det(Et (S \ {xi,ni})). But the inductive assumption implies that det(Et (S \ {xi,ni})) = 0. Thus det(Et (S)) = 0 as claimed. Now lett =li .If xi,ni – 1 ∤ xini, then Lemma 2.8 infers that det(Eli. (S)) = det(Eli– 1 (S \ {xi,ni – 1})). However, by the induction assumption we have det(Eli – 1(S\{xi,ni – 1})) = 0. Hence det(Eli. (S)) = 0 as required. If xini–1 |xini, by Lemma 2.8 we have
But the induction assumption tells us that
So det(Eli (S)) = 0 and the claim is proved.
Now, by (10) and the claim, one deduces that
From (1), we deduce thatαS, f(1) =f(1) andαS, f (xi1) =f(xi1) –f(1). It then follows from (11) that
as desired.
Finally, we turn our attention to det(f[S]). Letf be a multiplicative function andf(x) ≠ 0 for allx ∈S. Thenf(1) = 1 and from Lemma 2.2 we derive that
Therefore
as required. This finishes the proof of Theorem 1.3. □
We are now ready to prove Corollary 1.4 as the conclusion of this paper.
Proof of Corollary 1.4. Letf =I. Then Theorem 1.3 tells us that
By the definition of αS,I (x) and I * μ = φ, we have
Thus the desired result follows immediately. By Theorem 1.3, one has
Since
the desired result then follows immediately. This completes the proof of Corollary 1.4. □
Remark 3.1.If S consists of multiple coprime gcd-closed sets such that 1 ∉ S, then Theorem 1.3 gives us formulas for det(f(S)) and det(f[S]). One can easily see that Theorem 1.3 is not true if S consists of multiple gcd-closed sets which are not coprime. If S consists of multiple gcd-closed sets such that gcd(S) = 1 ∉ S and these gcd-closed sets are not coprime, then what are the formulas for det(f(S)) and det(f[S])? This interesting problem keeps open.
Acknowledgement
This research was supported partially by National Science Foundation of China Grant # 11371260. The authors thank the anonymous referees for helpful comments and suggestions.
References
[1] Apostol T.M., Arithmetical properties of generalized Ramanujan sums, Pacific J. Math., 1972, 41, 281–29310.2140/pjm.1972.41.281Search in Google Scholar
[2] Bege A., Generalized LCM matrices, Publ. Math. Debrecen 2011, 79, 309–31510.5486/PMD.2011.5131Search in Google Scholar
[3] Beslin S., Ligh S., Another generalization of Smith’s determinant, Bull. Aust. Math. Soc., 1989, 40, 413–41510.1017/S0004972700017457Search in Google Scholar
[4] Bourque K., Ligh S., On GCD and LCM matrices, Linear Algebra Appl., 1992, 174, 65–7410.1016/0024-3795(92)90042-9Search in Google Scholar
[5] Bourque K., Ligh S., Matrices associated with classes of arithmetical functions, J. Number Theory, 1993, 45, 367–37610.1006/jnth.1993.1083Search in Google Scholar
[6] Bourque K., Ligh S., Matrices associated with arithmetical functions, Linear Multilinear Algebra, 1993, 34, 261–26710.1080/03081089308818225Search in Google Scholar
[7] Codecá P., Nair M., Calculating a determinant associated with multiplicative functions, Boll. Unione Mat. Ital. Sez. B Artic. Ric. Mat., 2002, 5, 545–555Search in Google Scholar
[8] Gantmacher F.R., Matrix theory, Vol. I, Chelsea, New York, 1960Search in Google Scholar
[9] Haukkanen P., Higher-dimensional GCD matrices, Linear Algebra Appl., 1992, 170, 53–6310.1016/0024-3795(92)90409-4Search in Google Scholar
[10] Haukkanen P., On meet matrices on posets, Linear Algebra Appl., 1996, 249, 111–12310.1016/0024-3795(95)00349-5Search in Google Scholar
[11] Hilberdink T., Determinants of multiplicative Toeplitz matrices, Acta Arith., 2006, 125, 265–28410.4064/aa125-3-4Search in Google Scholar
[12] Hong S., The lower bound of the determinants of matrices associated with a class of arithmetical functions, Chinese Ann. Math., Ser. A, 2000, 21, 377–382 (in Chinese)Search in Google Scholar
[13] Hong S., Gcd-closed sets and determinants of matrices associated with arithmetical functions, Acta Arith., 2002, 101, 321–33210.4064/aa101-4-2Search in Google Scholar
[14] Hong S., Lee K.S.E., Asymptotic behavior of eigenvalues of reciprocal power LCM matrices, Glasgow Math. J., 2008, 50, 163–17410.1017/S0017089507003953Search in Google Scholar
[15] Hong S., Li M., Wang B., Hyperdeterminants associated with multiple even functions, Ramanujan J., 2014, 34, 265–28110.1007/s11139-013-9548-1Search in Google Scholar
[16] Hong S., Loewy R., Asymptotic behavior of eigenvalues of greatest common divisor matrices, Glasgow Math. J., 2004, 46, 551–569.10.1017/S0017089504001995Search in Google Scholar
[17] Hong S., Loewy R., Asymptotic behavior of the smallest eigenvalue of matrices associated with completely even functions (modr), Int. J. Number Theory, 2011, 7, 1681–170410.1142/S179304211100437XSearch in Google Scholar
[18] Hu S., Hong S., Multiple divisor chains and determinants of matrices associated with completely even functions (mod r), Linear Multilinear Algebra, 2014, 62, 1240–1257.10.1080/03081087.2013.818993Search in Google Scholar
[19] Korkee I., Haukkanen P., On a general form of join matrices associated with incidence functions, Aequationes Math., 2008, 75, 29–4210.1007/s00010-007-2922-6Search in Google Scholar
[20] Li M., Tan Q., Divisibility of matrices associated with multiplicative functions, Discrete Math., 2011, 311, 2276–228210.1016/j.disc.2011.07.015Search in Google Scholar
[21] Li Z., The determinants of GCD matrices, Linear Algebra Appl., 1990, 134, 137–14310.1016/0024-3795(90)90012-2Search in Google Scholar
[22] Mattila M., Haukkanen P., Determinant and inverse of join matrices on two sets, Linear Algebra Appl., 2013, 438, 3891–390410.1016/j.laa.2011.12.014Search in Google Scholar
[23] McCarthy P.J., A generalization of Smith’s determinant, Canad. Math. Bull., 1986, 29, 109–11310.4153/CMB-1986-020-1Search in Google Scholar
[24] Nathanson M.B., Elementary methods in number theory, Grad. Texts in Math., 195, Springer, Berlin-New York-Heidelberg, 2000Search in Google Scholar
[25] Smith H.J.S., On the value of a certain arithmetical determinant, Proc. London Math. Soc., 1875–1876, 7, 208–21210.1112/plms/s1-7.1.208Search in Google Scholar
[26] Tan Q., Li M., Divisibility among power GCD matrices and among power LCM matrices on finitely many coprime divisor chains, Linear Algebra Appl., 2013, 438, 1454–146610.1016/j.laa.2012.08.036Search in Google Scholar
[27] Tan Q., Luo M., Lin Z., Determinants and divisibility of power GCD and power LCM matrices on finitely many coprime divisor chains, Appl. Math. Comput., 2013, 219, 8112–812010.1016/j.amc.2010.09.053Search in Google Scholar
[28] Yamasaki Y., Arithmetical properties of multiple Ramanujan sums, Ramanujan J., 2010, 21, 241–26110.1007/s11139-010-9223-8Search in Google Scholar
© 2016 Hong et al., published by De Gruyter Open
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.