Painlevé Transcendents and the Hankel Determinants Generated by a Discontinuous Gaussian Weight
|
|
- Rodger Casey
- 5 years ago
- Views:
Transcription
1 Painlevé Transcendents and the Hankel Determinants Generated by a Discontinuous Gaussian Weight arxiv: v [math-ph] May 018 Chao Min and Yang Chen November 9, 018 Abstract This paper studies the Hankel determinants generated by a discontinuous Gaussian weight withoneandtwojumps. Itis anextension of ChenandPruessner[8], inwhich theystudiedthe discontinuous Gaussian weight with a single jump. By using the ladder operator approach, we obtain a series of difference and differential equations to describe the Hankel determinant for the single jump case. These equations include the Chazy II equation, continuous and discrete Painlevé IV. In addition, we consider the large n behavior of the corresponding orthogonal polynomials and prove that they satisfy the biconfluent Heun equation. We also consider the jump at the edge under a double scaling, from which a Painlevé XXXIV appeared. Furthermore, we study the Gaussian weight with two jumps, and show that a quantity related to the Hankel determinant satisfies a two variables generalization of the Jimbo-Miwa-Okamoto σ form of the Painlevé IV. Keywords: Hankel determinants; Random matrices; Orthogonal polynomials; Ladder operators; Painlevé transcendents. Mathematics Subject Classification 010: 15B5, 4C05, 33E17 School of Mathematical Sciences, Huaqiao University, Quanzhou 3601, China; chaomin@hqu.edu.cn Correspondence to: Yang Chen, Department of Mathematics, Faculty of Science and Technology, University of Macau, Macau, China; yangbrookchen@yahoo.co.uk 1
2 1 Introduction In the theory of random matrix ensembles, the joint probability density function for the eigenvalues {x j } n j=1 of n n Hermitian matrices from an unitary ensemble is given by [] n Px 1,x,...,x n ) dx j = j=1 1 1 D n [w 0 ] n! n x k x j ) w 0 x j )dx j, 1.1) 1 j<k n j=1 where w 0 x) is a weight function supported on [a,b] and the moments of w 0 x), namely, µ j := b exist. Here D n [w 0 ] is the normalization constant In this paper, we take D n [w 0 ] = 1 n! a x j w 0 x)dx, j = 0,1,,... [a,b] n 1 j<k n w 0 x) = e x, x R, n x k x j ) w 0 x j )dx j. which corresponds to the Gaussian unitary ensemble. In this case, D n [w 0 ] has the closed-form expression [], where Gz) is the Barnes G-function, defined by j=1 D n [w 0 ] = π) n n Gn+1), Gz +1) = Γz)Gz), G1) = 1 and Γz) is the gamma function, Γz) := 0 t z 1 e t dt, Rez) > 0. Noting that, [] also gives the closed-form expression D n [w 0 ] for other weight functions, such as the Laguerre weight, w 0 x) = x α e x, α > 1, x R + and the Jacobi weight, w 0 x) = 1 x) α 1+x) β, α > 1, β > 1, x [ 1,1]. Linear statistics is a ubiquitous statistical characteristic in random matrix theory [1, 6, 7, 19, 3, 4]. A linear statistic is a linear sum of a certain function g of the random variable x j : n j=1 gx j). In our paper, {x j, j = 1,,...,n} are the eigenvalues of Hermitian matrices. Whenever one encounters a linear statistic, it is useful to consider its exponential generating function, which is defined by the expectation of e λ n j=1 gx j) over the joint probability distribution 1.1), where λ is a
3 parameter, that is, E e λ ) n j=1 gx j) := 1 1 D n [w 0 ] n! More generally, we can consider n ) 1 1 E fx j ) := D n [w 0 ] n! j=1 In this paper, we investigate the case, ) n 1 j<k n, ) n 1 j<k n n x k x j ) w 0 x j )e λgxj) dx j. j=1 n x k x j ) w 0 x j )fx j )dx j. 1.) j=1 fx) = A+B 1 θx t 1 )+B θx t ), t 1 < t, 1.3) where θx) is the Heaviside step function, i.e., θx) is 1 for x > 0 and 0 otherwise, and A,B 1,B are constants, A 0, A+B 1 0, A+B 1 +B 0. Let wx,t 1,t ) := w 0 x)a+b 1 θx t 1 )+B θx t )) 1.4) and D n [w] := 1 n!, ) n 1 j<k n n x k x j ) wx j,t 1,t )dx j. We will see that D n [w] is the Hankel determinant generated by the weight wx,t 1,t ). The motivation of this paper comes from Chen and Pruessner [8], in which they studied the Hankel determinant generated by the Gaussian weight with a single jump. In addition, Basor and Chen [], Chen and Zhang [9] investigated the Hankel determinants for the Laguerre weight and Jacobi weight with a single jump, respectively. We would like to consider the discontinuous Gaussian weight 1.4). The single jump case is recovered if B = 0, and we also study this case since we have some important results beyond [8]. So this paper is divided into two parts, the first part is the single jump case and the other is the general case with two jumps. We would like to point out that, there are three important special cases from 1.) and 1.3): the first one is A = 0, B 1 = 1, B = 0, and this leads us to compute the probability that the smallest eigenvalue is greater than t 1 ; the second one is A = 1, B 1 = 1, B = 0, and this allows us to compute the probability that the largest eigenvalue is less than t 1 ; the last one is A = 0, B 1 = 1, B = 1 and is related to the probability of the eigenvalues lying in the interval t 1,t ), which has been studied by Basor, Chen and Zhang [3]. Moreover, Tracy and Widom 3 j=1
4 [3] used the Fredholm theory of integral equations to study the probability distribution of the largest and smallest eigenvalue in the Gaussian unitary ensemble. They expressed it as a Fredholm determinant deti K), and proved that the logarithmic derivative of deti K) satisfies a third order differential equation. This equation can be integrated to the σ form of a Painlevé IV if the integration constant is zero see 5.14) in[3]). They also studied the largest and smallest eigenvalue distribution in the Laguerre and Jacobi unitary ensemble. Similarly, the Laguerre case is related to a third order differential equation and can be integrated to the σ form of a Painlevé V. But for the Jacobi case, they only obtained a third order differential equation. Later, Haine and Semengue [15] used the Virasoro approach to obtain another third order differential equation and showed that the difference of them can be reduced to the σ form of a Painlevé VI. The approach in this paper is based on the ladder operators of orthogonal polynomials and the associated compatibility conditions S 1 ), S ) and S ). An elementary method to compute D n[w] is to write it as a Hankel determinant. It is a well-known fact that Hankel determinants can be expressed as the product of the square of the L norms of orthogonal polynomials. Based on the ladder operators adapted to these orthogonal polynomials, and from the associated supplementary conditions, a series of difference and differential equations can be derived to ultimately give a description of D n [w]. This paper is organized as follows: In Sec., we consider the Gaussian weight with a single jump. From the ladder operators and supplementary conditions on the weight wx,t 1 ), we obtain in addition to what was found in Chen and Pruessner [8], and prove that the auxiliary quantities r n t 1 ), R n t 1 ) and σ n t 1 ) satisfy the second order difference and differential equations respectively. Moreover, we consider the large n behavior of the monic orthogonal polynomials P n z,t 1 ) and the double scaling of the auxiliary quantities. In Sec. 3, we consider the general case with two jumps. We also use the ladder operator approach to obtain a partial differential equation on σ n t 1,t ), which is a two variables generalization of the σ form of the Painlevé IV. The conclusion is given in Sec. 4. 4
5 Gaussian Weight with a Single Jump We set B = 0 and B 1 0, which corresponds to the Gaussian weight with a single jump wx,t 1 ) := e x A+B 1 θx t 1 ))..1) This case has been studied by Chen and Pruessner [8] with parameters A = 1 β, B 1 = β. They obtained a Painlevé IV for the diagonal recurrence coefficient α n t 1 ) of the monic orthogonal polynomials with respect to.1), which is actually the same with our result.41). Furthermore, Its [16] studied the weight e iβπ, x < t 1 ; wx,t 1 ) = e x e iβπ, x > t 1, which corresponds to A = e iβπ, B 1 = e iβπ e iβπ in our problem, and they also obtained a Painlevé IV for α n t 1 ) by using the Riemann-Hilbert approach. See also [4, 17] for more detailed discussion on this weight. In addition, Xu and Zhao [34] considered the Gaussian weight with a jump at the edge..1 Ladder Operators and Supplementary Conditions Let P n x,t 1 ) be the monic polynomials of degree n orthogonal with respect to the weight function wx,t 1 ), that is, P j x,t 1 )P k x,t 1 )wx,t 1 )dx = h j t 1 )δ jk, j,k = 0,1,,...,.) where wx,t 1 ) = w 0 x)a+b 1 θx t 1 )), w 0 x) := e v0x), v 0 x) = x. The monic polynomials P n x,t 1 ) have the monomial expansion P n x,t 1 ) = x n +pn,t 1 )x n 1 + +P n 0,t 1 ),.3) and we see later that pn,t 1 ), the coefficient of x n 1, will play a significant role. From the orthogonality condition.), we have the recurrence relation [31] xp n x,t 1 ) = P n+1 x,t 1 )+α n t 1 )P n x,t 1 )+β n t 1 )P n 1 x,t 1 ).4) 5
6 with the initial conditions An easy consequence of.4) and.3) gives P 0 x,t 1 ) = 1, β 0 t 1 )P 1 x,t 1 ) = 0. α n t 1 ) = pn,t 1 ) pn+1,t 1 ).5) and a telescopic sum followed by n 1 α j t 1 ) = pn,t 1 )..6) Moreover, it follows from.4) and.) that In this section, we also denote D n t 1 ) := D n [w] = 1 n! β n t 1 ) = h nt 1 ) h n 1 t 1 )..7), ) n 1 j<k n n x k x j ) wx j,t 1 )dx j. It is a well-known fact that D n t 1 ) can be expressed as a Hankel determinant generated by the weight wx,t 1 ) [31] ) n 1 D n t 1 ) = det x i+j wx,t 1 )dx i, n 1 = h j t 1 )..8) For convenience, we would not show the t 1 dependence in P n x), h n, α n and β n unless it is needed in the following discussions. From Lemma 1 in [] and noting that v 0 z) v 0 y) z y we have the following theorem. j=1 = in our problem, Theorem.1. The monic orthogonal polynomials with respect to the weight wx,t 1 ) satisfy the following differential recurrence relations: P n z) = β na n z)p n 1 z) B n z)p n z),.9) P n 1 z) = B nz)+v 0 z))p n 1z) A n 1 z)p n z),.10) where A n z) = + R nt 1 ) z t 1,.11) 6
7 and R n t 1 ) = B 1 P n t 1,t 1 )e t 1 h n t 1 ) Here P n t 1,t 1 ) := P n x,t 1 ) x=t1. B n z) = r nt 1 ) z t 1,.1), r n t 1 ) = B 1 P n t 1,t 1 )P n 1 t 1,t 1 )e t 1. h n 1 t 1 ) The following lemmas can be found in [, 8, 9]. See also [5, 1, 1, 5] for more information. Lemma.. The functions A n z) and B n z) satisfy the conditions: B n+1 z)+b n z) = z α n )A n z) v 0 z), S 1) 1+z α n )B n+1 z) B n z)) = β n+1 A n+1 z) β n A n 1 z). S ) The combination of S 1 ) and S ) produces the following identity which gives better insight into the β n term. Lemma.3. A n z), B n z) and n 1 A jz) satisfy the identity n 1 Bnz)+v 0z)B n z)+ A j z) = β n A n z)a n 1 z). S ) Remark. The three identities S 1 ), S ) and S ) are valid for z C { }. Lemma.4. P n z) satisfy the following second order differential equation: ) P n z) v 0 z)+ A n z) )P n A n z) z)+ B n z) B nz) A n z) n 1 A n z) + A j z) P n z) = 0..13) Substituting.11) and.1) into S 1 ), we obtain Letting z, we have r n+1 t 1 )+r n t 1 ) z t 1 = z α n)r n t 1 ) z t 1 α n..14) Equating the residues at the simple pole t from.14), we find R n t 1 ) = α n..15) r n+1 t 1 )+r n t 1 ) = t 1 α n )R n t 1 )..16) 7
8 Similarly, plugging.11) and.1) into S ), we obtain rn t 1) z t 1 ) + zr nt 1 )+ n 1 R jt 1 ) +n = β nr n t 1 )R n 1 t 1 ) + β nr n t 1 )+R n 1 t 1 )) +4β z t 1 z t 1 ) n. z t 1 Letting z, we have Multiplying both sides of.17) by z t 1 ) and letting z gives.17) r n t 1 ) = β n n..18) r nt 1 ) = β n R n t 1 )R n 1 t 1 )..19) Then.17) becomes zr n t 1 )+ n 1 R jt 1 ) z t 1 +n = β nr n t 1 )+R n 1 t 1 )) z t 1 +4β n. Equating the residues at the simple pole t 1 produces n 1 t 1 r n t 1 ) β n R n t 1 )+R n 1 t 1 ))+ R j t 1 ) = 0..0) Using.18) and.19) to eliminate β n and R n 1 t 1 ) from the above, we have t 1 r n t 1 ) n+r n t 1 ))R n t 1 ) r n t n 1 1) R n t 1 ) + R j t 1 ) = 0..1) Remark. The above equalities.15),.16),.18),.19) and.1) also appeared in [8], and are crucial for the following discussions.. Non-linear Difference Equations Satisfied by r n t 1 ), R n t 1 ) and σ n t 1 ) In this subsection, we would like to obtain the second order non-linear difference equations satisfied by r n t 1 ), R n t 1 ) and σ n t 1 ) respectively. These are the new results beyond [8]. Eliminating α n t 1 ) from.15) and.16), we have r n+1 t 1 )+r n t 1 ) = t 1 1 ) R nt 1 ) R n t 1 )..) Similarly, eliminating β n t 1 ) from.18) and.19), we find r n t 1) = 1 n+r nt 1 ))R n t 1 )R n 1 t 1 )..3) 8
9 Solving for R n t 1 ) from.), a quadratic, and substituting either solution into.3), we find after clearing the square root, a non-linear second order difference equation for r n := r n t 1 ), [ n+rn ) r n 1 +r n t 1)r n+1 +n+r n ) r n r n 1 +nr 3 n nt 1 n)r n n t 1 r n ] = t 1 n+r n) t 1 r n 1 r n )[nr n +n+r n )r n+1 ]..4) On the other hand, regarding.3) as a quadratic equation in r n t 1 ), and substituting either solution into.), we find after clearing the square root, the following second order non-linear difference equation for R n := R n t 1 ), [ R n 4t 1 R n +R n 1 R n 4n 4)R n+1 +Rn t 1R n n)r n 1 +R n R n t 1 ) ] = R n 1 R n+1 R n 1 R n +8n)R n R n+1 +8n+8)..5) Remark..5) may be related to the discrete Painlevé IV equation [13, 7, 8]. Now we introduce a quantity σ n t 1 ), defined by n 1 σ n t 1 ) := R j t 1 )..6) We will see in the next subsection that σ n t 1 ) is the logarithmic derivative of D n t 1 ), i.e. It is easy to see that Using.18),.6) and.7),.0) becomes Then, σ n t 1 ) = d dt 1 lnd n t 1 ). R n t 1 ) = σ n t 1 ) σ n+1 t 1 )..7) t 1 r n t 1 ) n+r n t 1 ))σ n 1 t 1 ) σ n+1 t 1 )) σ n t 1 ) = 0. r n t 1 ) = nσ n 1t 1 ) σ n+1 t 1 ))+σ n t 1 )..8) t 1 σ n 1 t 1 )+σ n+1 t 1 ) With the aid of.18) and.7), it follows from.19) that rnt 1 ) = n+r nt 1 ) σ n t 1 ) σ n+1 t 1 ))σ n 1 t 1 ) σ n t 1 ))..9) Substituting.8)into.9),weobtainasecondorderdifferenceequationsatisfiedbyσ n := σ n t 1 ), [σ n +nσ n 1 σ n+1 )] = σ n σ n+1 )σ n 1 σ n )σ n +nt 1 )σ n+1 σ n 1 +t 1 ), which is the discrete σ form of the Painlevé IV equation. 9
10 .3 Painlevé IV, Chazy II, and the σ Form In this subsection, we derive the second order differential equations satisfied by R n t 1 ) and r n t 1 ) respectively, which are related to the Painlevé IV and Chazy II. The Painlevé IV satisfied by R n t 1 ) or α n t 1 ) was obtained by [8], but the Chazy equation satisfied by r n t 1 ) is a new one. We would like to say that Chazy equations appear in the random matrix theory regularly. For example, Witte, Forrester and Cosgrove [33] obtained the Chazy equations when they studied the gap probability of Gaussian and Jacobi unitary ensembles. Recently, Lyu, Chen and Fan [0] also obtained the Chazy equation when they studied the gap probability of Gaussian unitary ensembles by using the different method. In the end, we find that σ n t 1 ), the logarithmic derivative of D n t 1 ), satisfies the Jimbo-Miwa-Okamoto σ form of the Painlevé IV. We begin with taking a derivative with respect to t 1 in the following equation, P n x,t 1)wx,t 1 )dx = h n t 1 ), n = 0,1,,..., we obtain h n t 1) = B Pn t 1)e t 1. It follows that lnh n t 1 )) = R n t 1 ).30) and [lnβ n t 1 )] = lnh n t 1 )) lnh n 1 t 1 )) = R n 1 t 1 ) R n t 1 ), where we have made use of.7) in the first step. Hence, β n t 1) = β n R n 1 t 1 ) β n R n t 1 )..31) Moreover, from.8),.30) and.6) we obtain d lnd n t 1 ) = d n 1 n 1 ln h j t 1 ) = R j t 1 ) = σ n t 1 )..3) dt 1 dt 1 On the other hand, differentiating with respect to t 1 in the equation P n x,t 1 )P n 1 x,t 1 )wx,t 1 )dx = 0, n = 0,1,,..., 10
11 gives d dt 1 pn,t 1 ) = r n t 1 )..33) Now we have the Toda equations on α n t 1 ) and β n t 1 ). These are also obtained by [8]. Proposition.5. β n t 1) = β n α n 1 α n ).34) α nt 1 ) = β n β n+1 )+1.35) Proof. The combination of.15) and.31) results in.34)..35) comes from.5),.18) and.33). The following lemma is very important for the derivation of the second order differential equations satisfied by R n t 1 ), r n t 1 ) and σ n t 1 ). Lemma.6. r n t 1 ) and R n t 1 ) satisfy the following coupled Riccati equations: r nt 1 ) = r n t 1) R n t 1 ) n+r nt 1 ))R n t 1 ),.36) R n t 1) = R n t 1) t 1 R n t 1 )+4r n t 1 )..37) Proof. From.31) and.19), we have It follows from.18) that β nt 1 ) = r n t 1) R n t 1 ) β nr n t 1 )..38) β n = n+r nt 1 ),.39) Substituting.39) into.38), we obtain the Riccati equation satisfied by r n t 1 ), r n t 1) = r n t 1) R n t 1 ) n+r nt 1 ))R n t 1 ). Now we come to prove the second equation.37), the Riccati equation satisfied by R n t 1 ). From.15) and.5) we find R n t 1 ) = [pn,t 1 ) pn+1,t 1 )], 11
12 then with the aid of.33), [ R nt dpn,t1 ) 1 ) = dpn+1,t ] 1) dt 1 dt 1 = [r n t 1 ) r n+1 t 1 )]. Using.16) to decrease the index n+1 to n, we have R nt 1 ) = [r n t 1 ) t 1 α n )R n t 1 )] [ = r n t 1 ) t 1 1 ) ] R nt 1 ) R n t 1 ) = R n t 1) t 1 R n t 1 )+4r n t 1 ), where we have used.15) in the second equality. This establishes the lemma. Theorem.7. R n t 1 ) and r n t 1 ) satisfy the following second order differential equations, R n t 1) = R nt 1 )) R n t 1 ) + 3 R3 n t 1) 4t 1 R n t 1)+t 1 n 1)R nt 1 ).40) and [ r nt 1 )+1r nt 1 )+8nr n t 1 ) ] = 4t 1 [ r n t 1 )) +8r 3 nt 1 )+8nr nt 1 ) ] respectively. Moreover, letting yt 1 ) = R n t 1 ), then yt 1 ) satisfies the Painlevé IV equation [14], y t 1 ) = y t 1 )) yt 1 ) + 3 y3 t 1 )+4t 1 y t 1 )+t 1 α 1 )yt 1 )+ β 1 yt 1 ).41) with α 1 = n+1, β 1 = 0. Letting vt 1 ) = r n t 1 ) n 3, then vt 1) satisfies the first member of the Chazy II system [11], v t 1 ) 6v t 1 ) α ) = 4t 1 v t 1 ) 4v 3 t 1 ) α vt 1 ) β ) with α = 8n 3, β = 64n3 7. Proof. We start from expressing r n t 1 ) in terms of R n t 1 ) and R nt 1 ) by.37), r n t 1 ) = 1 4 [ R n t 1 ) R nt 1 )+t 1 R n t 1 ) ],.4) then r nt 1 ) = 1 4 [R nt 1 ) R n t 1 )R nt 1 )+R n t 1 )+tr nt 1 )]..43) 1
13 Substituting.4) and.43) into.36), we obtain the differential equation for R n t 1 ), R nt 1 ) = R n t 1)) R n t 1 ) + 3 R3 nt 1 ) 4t 1 R nt 1 )+t 1 n 1)R n t 1 ). Letting yt 1 ) = R n t 1 ), it is easy to see that yt 1 ) satisfies a particular Painlevé IV, y t 1 ) = y t 1 )) yt 1 ) + 3 y3 t 1 )+4t 1 y t 1 )+t 1 n 1)yt 1). Now we turn to prove the differential equation for r n t 1 ). Viewing.36) as an equation on R n t 1 ), we find the solution is where R n t 1 ) = r n t 1)± n+r n t 1 )), := r n t 1)) +8r 3 n t 1)+8nr n t 1). Then plugging it into.37), we have r n t 1)± ) r n t 1) 1r n t 1) 8nr n t 1 )±t 1 ) n+r n t 1 )) = 0. It follows a second order differential equation for r n t 1 ), r n t 1) 1r n t 1) 8nr n t 1 )±t 1 r n t 1 )) +8r 3 n t 1)+8nr n t 1) = 0, or [ r n t 1)+1rn t 1)+8nr n t 1 ) ] [ = 4t 1 r n t 1 )) +8rn 3 t 1)+8nrn t 1) ]..44) Letting vt 1 ) = r n t 1 ) n 3, or r nt 1 ) = vt 1) n 3, and substituting it into.44), then vt 1) satisfies the first member of the Chazy II system, v t 1 ) 6v t 1 )+ 8n 3 This finishes the proof of Theorem.7. ) ) = 4t 1 v t 1 ) 4v 3 t 1 )+ 16n 3 vt 1)+ 64n3. 7 Theorem.8. σ n t 1 ) satisfies the Jimbo-Miwa-Okamoto σ form of the Painlevé IV equation [18] σ n t 1)) = 4t 1 σ n t 1) σ n t 1 )) 4σ n t 1)+ν 0 )σ n t 1)+ν 1 )σ n t 1)+ν ),.45) with parameters ν 0 = ν 1 = 0, ν = n. 13
14 Proof. By using.6),.1) becomes From.36), we have The difference and sum of.46) and.47) give and respectively. Then the product of.48) and.49) leads to n+r n t 1 ))R n t 1 )+ r n t 1) R n t 1 ) = tr nt 1 ) σ n t 1 )..46) n+r n t 1 ))R n t 1 ) r n t 1) R n t 1 ) = r nt 1 )..47) 4r n t 1) R n t 1 ) = t 1r n t 1 ) σ n t 1 )+r n t 1).48) n+r n t 1 ))R n t 1 ) = t 1 r n t 1 ) σ n t 1 ) r nt 1 ),.49) 8n+r n t 1 ))r nt 1 ) = t 1 r n t 1 ) σ n t 1 )) r nt 1 ))..50) Noting that from.6),.15) and.6), we find Then it follows from.33) that or n 1 σ n t 1 ) = α j = pn,t 1 ). σ n t 1) = d dt 1 pn,t 1 ) = r n t 1 ), r n t 1 ) = 1 σ nt 1 )..51) Substituting.51) into.50), we obtain a second order differential equation satisfied by σ n t 1 ), σ n t 1)) = 4t 1 σ n t 1) σ n t 1 )) 4σ n t 1)) σ n t 1)+n), which is just the Jimbo-Miwa-Okamoto σ form of the Painlevé IV equation, P IV 0,0,n). This result is coincident with Tracy and Widom [3] when they studied the largest eigenvalue distribution in the Gaussian unitary ensemble. Therefore, σ n t 1 ) satisfies both the continuous and discrete σ form of the Painlevé IV. 14
15 .4 Large n Behavior of the Orthogonal Polynomials and Double Scaling Analysis In this subsection, we consider the large n behavior of the monic orthogonal polynomials P n z). We show that, as n, P n z) satisfies the confluent forms of Heun s differential equation. We also give the large n asymptotics of R n t 1 ),r n t 1 ) and σ n t 1 ) under a double scaling, which gives rise to the Painlevé XXXIV equation. Theorem.9. As n, i) if B 1 > 0, then ˆP n u) := P n ) u +t 1 satisfies the biconfluent Heun equation BHE) [9] γ ) ˆP nu) u +δ +u ˆP nu)+ αu q ˆP n u) = 0,.5) u where γ = 1, δ = t 1, α = 0, q = 4 3n 3 ii) if B 1 < 0, then ˆP ) n u) := P n u +t 1 satisfies.5) with parameters γ = 1, δ = t 1, α = 0, q = 4 3n ; Proof. Substituting.11) and.1) into.13), and using.1) to eliminate n 1 R jt 1 ), we obtain + ) P nz)+p nz) R n t 1 ) z t 1 )z t 1 +R n t 1 )) z +P n z) n r nt 1 ) z t 1 ) r n t 1 )R n t 1 ) z t 1 ) z t 1 +R n t 1 )) + r n t 1)+n+r n t 1 ))Rn t ) 1) t 1 r n t 1 )R n t 1 ) z t 1 )R n t 1 ) = 0..53) Replacing r n t 1 ) with the expression of R n t 1 ) from.4),.53) becomes ) P nz)+p nz) R n t 1 ) z t 1 )z t 1 +R n t 1 )) z +P n z) n R n t 1) Rn t 1)+t 1 R n t 1 ) 4z t 1 ) + R nt 1 )R nt 1 ) R nt 1 )+t 1 R n t 1 )) 4z t 1 ) z t 1 +R n t 1 )) ) + R nt 1 )) Rnt 4 1 )+4t 1 Rnt 3 1 )+8n 4t 1)Rnt 1 ) = 0. 8z t 1 )R n t 1 ) Note that the coefficients of P n z) and P n z) only depend on R nt 1 ) and R n t 1) in.54). Now we consider the large n behavior of R n t 1 ). Let ˆR n t 1 ) satisfy a quadratic equation obtained from the non-derivative part of.40), 3ˆR n t 1) 8t 1 ˆRn t 1 )+4t 1 n 1) = )
16 with the solutions i) If B 1 > 0, we choose ˆR n t 1 ) = 3 ˆR n t 1 ) = 3 ) t 1 ± t 1 +6n+3. ) t 1 + t 1 +6n+3. As n, ˆR n t 1 ) = 6n 3 Hence we suppose the expansion, as n, + 4t 1 6t ) 6t ) 6t ) 3 +On 7 n 43n n 5 ). R n t 1 ) = a j t 1 )n 1 j. Substituting the above into.40), we obtain R n t 1 ) = 6n + 4t 1 6t ) 6t t 1 +15) + t 1 6t 6 n 43n 3 18n t t 1 +81) +On 3 ). 5184n 5.55) Plugging.55) into.54), we see that as n, ) 1 P n z)+ z P n z t z)+ 4 6n 3 1 9z t 1 ) P nz) = 0. Let Then ˆP n u) := P n u +t 1 ) z = u +t 1. satisfies the biconfluent Heun equation.5) with parameters γ = 1, δ = t 1, α = 0, q = 4 3n 3 9. ii) If B 1 < 0, we choose ˆR n t 1 ) = 3 ) t 1 t 1 +6n+3. As n, ˆR n t 1 ) = 6n 3 Similarly, we suppose that as n, + 4t 1 6t ) 6t ) 6t 1 +3) 3 +On 7 n 43n n 5 ). R n t 1 ) = b j t 1 )n 1 j. 16
17 Substituting it into.40), we obtain R n t 1 ) = 6n + 4t 1 6t ) 6t 4 18 n t 1 +15) + t 1 6t 6 43n 3 18n 1 +9t t 1 +81) +On 3 ). 5184n 5.56) Plugging.56) into.54), we see that as n, Let ) 1 P n z)+ z P n z t z) 4 6n 3 1 9z t 1 ) P nz) = 0. Then ˆP ) u n u) := P n +t 1 z = u +t 1. satisfies the biconfluent Heun equation.5) with parameters γ = 1, δ = t 1, α = 0, q = 4 3n 3 9. Remark. There are four standard confluent forms of Heun s equation: confluent Heun equation CHE), doubly confluent Heun equationdche), biconfluent Heun equationbhe) and triconfluent Heun equation THE) [9]. The Heun s equation and its confluent forms play an important role in mathematical physics. Many known special functions, such as hypergeometric functions, Mathieu functions and spheroidal functions are solutions of Heun-class equations [30]. Theorem.10. Assume that n, t 1 = n + 1 n 1 6s and s is fixed. Then the large n asymptotics of R n t 1 ),r n t 1 ) and σ n t 1 ) are given by R n t 1 ) = n 1 6 v1 s)+n 1 v s)+n 5 6 v3 s)+on 7 6 ), 3v r n t 1 ) = n1 1 s) sv1 s) v + 4 v 1 1s)+v s))+ s)+ v s)+ v 3 s) +On 4n 1 3 ),.57) 3 and σ n t 1 ) = n 1 6 sv 1 s) v 1 s) ) v 1 s)) 4v 1 s) + n 1 s v 1 s) v 1s)v 3s)+v s)) 4v 1 s) 4 +n 1 6 sv s) v 1 s)v s) v 1 s)v s) ) + v 1 s)) v s) v 1 s) 4v1s) v 1 s)v 3 s) sv 1 s) + v3 1 s) v s) +sv 3 s) 8 + v 1s)v 1s)v 3 s)+v s)v s)) 4v 1 s) ) v 1s)) vs) 4v1 3s) +On 5 6 ),.58) 17
18 respectively. Here v 1 s),v s) and v 3 s) satisfy the differential equations.6),.63) and.64), and the large s asymptotics are given by.65),.66) and.67). In addition, ˆvs) := v 1s) satisfies the Painlevé XXXIV equation [4, 16] ˆv s) = 4ˆv s)+s ˆvs)+ ˆv s)) ˆvs)..59) Proof. By changing variable t 1 to s from the relation t 1 = n+ 1 n 1 6s, and denoting R n t 1 ) = R n n+ 1 n 1 6 s) =: Rn s),.40) becomes r n t 1 ) = r n n+ 1 n 1 6 s) =: rn s), σ n t 1 ) = σ n n+ 1 n 1 6 s) =: σn s), ) R R n s) = n s) R n s) + 3 4n 1 3 R 3 n s) 1 ) n 1 6 +n 1 s s R n s)+ n 3 +s n 1 3 ) R n s)..60) Let Řns) satisfy the quadratic equation by neglecting the derivative terms in.60), 3Ř ns) 4 n s+n )Řn s)+n 1 3 s +8n 1 3 s = 4. The solution is Ř n s) = n+ ) n 16 s± 1+3n+n 1 3s +8n 1 3s. As n, t 1, then wx,t 1 ) Ae x. It makes α n t 1 ) 0 since wx,t 1 ) tends to an even weight function [10]. In view of R n t 1 ) = α n t 1 ), we see that R n s) 0 as n. So we should choose Ř n s) = n+ ) n 16 s 1+3n+n 1 3s +8n 1 3s. It follows that as n, Hence we suppose that Ř n s) = s n n s 16 n 5 6 ) 1 +O. n 7 6 R n s) = n 1 6 v1 s)+n 1 v s)+n 5 6 v3 s)+on 7 6 )..61) 18
19 Substituting.61) into.60), we find as n, and v 1s) 1 s)) v 1 s) + v1s) sv 1 s) = 0,.6) ) v 1s)v s) v + 1 s)) v 1 s) v1 s) +4 v 1 s) s v s)+v 1 s) = 0,.63) v 3 s) v 1s)v 3s) 1 + v 1 s) v1 s) +4 v 1 s) s )v 3 s) v s)) v 1 s) + v 1s)v s)v s) v1 s) v 1 s)) v s) v 3 1s) 1 s v 1 s)+ sv1 s) 3 4 v3 1 s)+v s)+ v s) = 0..64) We obtain the large s asymptotic of v 1 s) from.6). As s, v 1 s) = s s 9 8 s s s 11 +os 11 )..65) Substituting.65) into.63), we have the large s behavior of v s), v s) = s s s s 1 +os 1 )..66) Then substituting.65) and.66) into.64), we obtain the large s asymptotic of v 3 s), v 3 s) = s s s s s 13 +os 13 )..67) In addition, letting v 1 s) = ˆvs) and substituting into.6), we readily see that ˆvs) satisfies the Painlevé XXXIV equation.59). Now we consider the large n behavior of r n s) and σ n s). We find from.4) that r n s) = 1 4 [ 1 1 n 6 R n s) R ] n s)+ n+ 1 n 1 6 s) Rn s). Substituting.61) into the above, we arrive at.57). From.46) we have σ n t 1 ) = t 1 r n t 1 ) n+r n t 1 ))R n t 1 ) r n t 1) R n t 1 ). Replacing r n t 1 ) with the expression of R n t 1 ) from.4), we find σ n t 1 ) = R4 n t 1) 4t 1 Rn 3t 1)+4t 1 8n)R n t 1) R n t 1)). 8R n t 1 ) 19
20 It follows that σ n s) = 1 8 R 3 ns) n1 +n 1 6s Using.61), we obtain.58). This completes the proof. ) 3 R R ns)+s n n 1 3 s Rn s) n1 n s)). 4 R n s) Remark. The results of Theorem.10 coincide with [16]. See also [4, 34] on the asymptotics of the recurrence coefficients α n,β n and the Painlevé XXXIV equation. In addition, Perret and Schehr [6] also obtained the Painlevé XXXIV in the study of the gap probability distribution between the first two largest eigenvalues in the Gaussian unitary ensemble. Proposition.11. Assume that n, t 1 = n+ 1 n 1 6s and s is fixed. Then as n, P n z) := P n z, n+ 1 n 1 6s) satisfies the Hermite s differential equation P nz) z P nz)+n P n z) = 0. Proof. By changing variable t 1 to s,.54) becomes ) P n z)+ P n z) R n s) z s)z s+ R n s) z + R n s) 1 n 1 6 R n s) R ) ns)+ s R n s) 4z s) z s+ R n s)) where R n s) = R n n+ 1 n 1 6s) and s := n+ 1 n 1 6s. + P n 1 n z) n 1 6 R n s) R ns)+ s R n s) 4z s) 3 R + n1 ns)) R ) ns)+4 s R 4 ns)+8n 4 s 3 ) R ns) = 0, 8z s) R n s) Substituting.61) into the above, we find as n, the coefficient of P n z) is and the coefficient of P n z) is It follows that as n, z + v 1s) 4n 7 6 +On 3 ) n v 1 s)) 4sv 1 s)+ v 3 1 s) 4 n 1 3v 1 s) P nz) z P nz)+n P n z) = 0. +On 3 ). 0
21 3 Gaussian Weight with Two Jumps In this section, we suppose B 1 0 and B 0, which corresponds to the Gaussian weight with two jumps wx,t 1,t ) := e x A+B 1 θx t 1 )+B θx t )), t 1 < t. This is an extension of [8] which considered the Gaussian weight with a single jump. For example, e µ, x < t 1 ; wx,t 1,t ) = e x 1, t 1 < x < t ; e µ, x > t, corresponds to A = e µ, B 1 = 1 e µ, B = e µ Ladder Operators Let P n x,t 1,t ) be themonic polynomials of degree n orthogonalwith respect to the weight function wx,t 1,t ), P j x,t 1,t )P k x,t 1,t )wx,t 1,t )dx = h j t 1,t )δ jk, j,k = 0,1,,..., 3.1) where wx,t 1,t ) := w 0 x)a 1 +B 1 θx t 1 )+B θx t )), w 0 x) := e v0x), v 0 x) = x. Similarly as the previous section, the monic polynomials P n x,t 1,t ) can be written in the form P n x,t 1,t ) = x n +pn,t 1,t )x n 1 + +P n 0,t 1,t ), 3.) and the recurrence relation reads xp n x,t 1,t ) = P n+1 x,t 1,t )+α n t 1,t )P n x,t 1,t )+β n t 1,t )P n 1 x,t 1,t ) with the initial conditions P 0 x,t 1,t ) = 1, β 0 t 1,t )P 1 x,t 1,t ) = 0. We also have the expressions of α n t 1,t ) and β n t 1,t ): α n t 1,t ) = pn,t 1,t ) pn+1,t 1,t ), 3.3) 1
22 A telescopic sum of 3.3) gives We also denote and we have β n t 1,t ) = h nt 1,t ) h n 1 t 1,t ). n 1 α j t 1,t ) = pn,t 1,t ). D n t 1,t ) := D n [w] = 1 n!, ) n 1 j<k n n x k x j ) wx j,t 1,t )dx j ) n 1 n 1 D n t 1,t ) = det x i+j wx,t 1,t )dx = h j t 1,t ). 3.4) i, To simplify notations, we suppress the t 1,t dependence in P n x), h n, α n and β n in the following discussions. From Lemma 1 and Remark in [], we have the following theorem. j=1 Theorem 3.1. The lowering and raising operators for monic polynomials orthogonal with respect to wx,t 1,t ) are P nz) = β n A n z)p n 1 z) B n z)p n z), P n 1 z) = B nz)+v 0 z))p n 1z) A n 1 z)p n z), where and A n z) = + R n,1t 1,t ) z t 1 + R n,t 1,t ) z t, 3.5) B n z) = r n,1t 1,t ) z t 1 + r n,t 1,t ) z t, 3.6) R n,1 t 1,t ) = B 1P n t 1)e t 1 h n, R n, t 1,t ) = B P n t )e t 1 h n, r n,1 t 1,t ) = B 1P n t 1 )P n 1 t 1 )e t 1 h n 1 Here P n t 1 ) = P n x,t 1,t ) x=t1, P n t ) = P n x,t 1,t ) x=t., r n, t 1,t ) = B P n t )P n 1 t )e t h n 1. From [, 8], we see that Lemma.,.3 and.4 are still valid for the weight with two jumps. Substituting 3.5) and 3.6) into S 1 ), we obtain r n+1,1 t 1,t )+r n,1 t 1,t ) z t 1 + r n+1,t 1,t )+r n, t 1,t ) z t = z α n)r n,1 t 1,t ) z t 1 + z α n)r n, t 1,t ) z t α n.
23 It follows that R n,1 t 1,t )+R n, t 1,t ) = α n, 3.7) r n+1,1 t 1,t )+r n,1 t 1,t ) = t 1 α n )R n,1 t 1,t ), and r n+1, t 1,t )+r n, t 1,t ) = t α n )R n, t 1,t ). Similarly, plugging 3.5) and 3.6) into S ), we obtain rn,1 t 1,t ) + r ) n,t 1,t ) rn,1 t 1,t ) +z z t 1 z t + n = β n + R n,1t 1,t ) z t 1 + R n,t 1,t ) z t It implies the following equalities: n 1 R j,1t 1,t ) + z t 1 + r ) n,t 1,t ) + z t 1 z t ) + R n 1,1t 1,t ) + R n 1,t 1,t ) z t 1 z t n 1 R j,t 1,t ) z t ). 3.8) β n = n+r n,1t 1,t )+r n, t 1,t ), 3.9) r n,1t 1,t ) = β n R n,1 t 1,t )R n 1,1 t 1,t ), 3.10) r n, t 1,t ) = β n R n, t 1,t )R n 1, t 1,t ), 3.11) r n,1 t 1,t )r n, t 1,t ) n 1 +t 1 r n,1 t 1,t )+ R j,1 t 1,t ) t 1 t = β n Rn,1 t 1,t )R n 1, t 1,t )+R n, t 1,t )R n 1,1 t 1,t ) t 1 t +R n,1 t 1,t )+R n 1,1 t 1,t ) ), 3.1) r n,1 t 1,t )r n, t 1,t ) n 1 +t r n, t 1,t )+ R j, t 1,t ) t t 1 = β n Rn,1 t 1,t )R n 1, t 1,t )+R n, t 1,t )R n 1,1 t 1,t ) t t 1 +R n, t 1,t )+R n 1, t 1,t ) ). 3.13) The sum of 3.1) and 3.13) gives n 1 n 1 t 1 r n,1 t 1,t )+t r n, t 1,t )+ R j,1 t 1,t )+ R j, t 1,t ) = β n R n,1 t 1,t )+R n, t 1,t )+R n 1,1 t 1,t )+R n 1, t 1,t )). 3.14) 3
24 3. Toda Evolution in t 1 and t Taking a derivative with respect to t 1 and t in the equation respectively, we obtain and It follows that and Hence, P nx,t 1,t )e x A 1 +B 1 θx t 1 )+B θx t ))dx = h n t 1,t ), n = 0,1,,..., t1 h n t 1,t ) = B 1 P nt 1 )e t 1 t h n t 1,t ) = B P nt )e t. t1 lnh n t 1,t ) = R n,1 t 1,t ), 3.15) t lnh n t 1,t ) = R n, t 1,t ), 3.16) t1 [lnβ n t 1,t )] = t1 lnh n t 1,t ) t1 lnh n 1 t 1,t ) = R n 1,1 t 1,t ) R n,1 t 1,t ), t [lnβ n t 1,t )] = t lnh n t 1,t ) t lnh n 1 t 1,t ) = R n 1, t 1,t ) R n, t 1,t ). t1 β n t 1,t ) = β n R n 1,1 t 1,t ) β n R n,1 t 1,t ) = r n,1t 1,t ) R n,1 t 1,t ) β nr n,1 t 1,t ), 3.17) t β n t 1,t ) = β n R n 1, t 1,t ) β n R n, t 1,t ) = r n, t 1,t ) R n, t 1,t ) β nr n, t 1,t ). 3.18) On the other hand, differentiating with respect to t 1 and t in the equation respectively gives and P n x,t 1,t )P n 1 x,t 1,t )e x A 1 +B 1 θx t 1 )+B θx t ))dx = 0, n = 0,1,,..., Now we have the two variables Toda equations on α n and β n. t1 pn,t 1,t ) = r n,1 t 1,t ), 3.19) t pn,t 1,t ) = r n, t 1,t ). 3.0) 4
25 Proposition 3.. t1 β n + t β n = β n α n 1 α n ), 3.1) t1 α n + t α n = β n β n+1 )+1. 3.) Proof. The sum of 3.17) and 3.18) gives t1 β n t 1,t )+ t β n t 1,t ) = β n R n 1,1 t 1,t )+R n 1, t 1,t ) R n,1 t 1,t ) R n, t 1,t )). Using 3.7), we arrive at 3.1). From 3.3), 3.19) and 3.0) we have t1 α n + t α n = r n,1 t 1,t )+r n, t 1,t ) r n+1,1 t 1,t ) r n+1, t 1,t ). With the aid of 3.9), we readily obtain 3.). 3.3 Generalized Jimbo-Miwa-Okamoto σ Form of Painlevé IV We define a quantity related to the Hankel determinant D n t 1,t ), σ n t 1,t ) := t1 lnd n t 1,t )+ t lnd n t 1,t ). It is easy from 3.4), 3.15) and 3.16) to see that n 1 n 1 σ n t 1,t ) = R j,1 t 1,t ) R j, t 1,t ). Then 3.14) becomes t 1 r n,1 t 1,t )+t r n, t 1,t ) σ n t 1,t ) = 4β n R n,1 t 1,t )+R n, t 1,t ))+ t1 β n + t β n, 3.3) where we have made use of 3.17) and 3.18). From 3.7) we have n 1 σ n t 1,t ) = α j = pn,t 1,t ). It follows from 3.19) and 3.0) that t1 σ n t 1,t ) = r n,1 t 1,t ), 3.4) t σ n t 1,t ) = r n, t 1,t ). 3.5) 5
26 Then we see from 3.9) that β n = n+ t 1 σ n t 1,t )+ t σ n t 1,t ). 3.6) 4 Substituting 3.4), 3.6) into 3.17) and 3.5), 3.6) into 3.18), we obtain t 1 σ n t 1,t )+ t1 t σ n t 1,t ) = t 1 σ n t 1,t )) n+ t1 σ n t 1,t )+ t σ n t 1,t ))R n,1 t 1,t ) 3.7) R n,1 t 1,t ) and t σ n t 1,t )+ t t1 σ n t 1,t ) = t σ n t 1,t )) n+ t1 σ n t 1,t )+ t σ n t 1,t ))R n, t 1,t ) 3.8) R n, t 1,t ) respectively. We regard 3.7) and 3.8) as quadratic equations on R n,1 t 1,t ) and R n, t 1,t ), respectively. The solutions are and where we write σ n t 1,t ) as σ n for short, and R n,1 t 1,t ) = t 1 σ n t1 t σ n ± 1 n+ t1 σ n + t σ n ) 3.9) R n, t 1,t ) = t σ n t t1 σ n ±, 3.30) n+ t1 σ n + t σ n ) 1 := t 1 σ n + t1 t σ n ) +4 t1 σ n ) n+ t1 σ n + t σ n ), := t σ n + t t1 σ n ) +4 t σ n ) n+ t1 σ n + t σ n ). Substituting 3.4), 3.5), 3.6), 3.9) and 3.30) into 3.3), we obtain a second order differential equation on σ n, which is equivalent to We summarize it into the following theorem. t 1 t1 σ n +t t σ n σ n ) = 1 + ± 1, [t 1 t1 σ n +t t σ n σ n ) 1 ] = 4 1. Theorem 3.3. σ n := σ n t 1,t ) satisfies the second order partial differential equation: [t 1 t1 σ n +t t σ n σ n ) t 1 σ n + t1 t σ n ) t σ n + t t1 σ n ) 4 t1 σ n ) n+ t1 σ n + t σ n ) 4 t σ n ) n+ t1 σ n + t σ n )] = 4[ t 1 σ n + t1 t σ n ) +4 t1 σ n ) n+ t1 σ n + t σ n )] [ t σ n + t t1 σ n ) +4 t σ n ) n+ t1 σ n + t σ n )]. 3.31) 6
27 Actually, 3.31) is a two variables generalization of the Jimbo-Miwa-Okamoto σ form of the Painlevé IV. If σ n is independent of t, then 3.31) is reduced to t 1 σ n ) = 4t 1 t1 σ n σ n ) 4 t1 σ n ) n+ t1 σ n ), 3.3) which is the Jimbo-Miwa-Okamoto σ form of the Painlevé IV. Note that 3.3) is the same with.45). Similarly, if σ n is independent of t 1, then 3.31) becomes t σ n ) = 4t t σ n σ n ) 4 t σ n ) n+ t σ n ), which is also the Jimbo-Miwa-Okamoto σ form of the Painlevé IV. The results of this section coincide with [3], which is a special case of our problem with A = 0, B 1 = 1, B = 1. Corollary 3.4. Assume that n, t 1 = n+ 1 n 1 6s 1, t = n+ 1 n 1 6s and s 1,s are fixed. Then as n, σ n s 1,s ) := σ n n+ 1 n 1 6s 1, n+ 1 n 1 6s ) satisfies the following second order partial differential equation: s1 σ n s σ n + s s1 σ n ) + s σ n s1 σ n + s1 s σ n ) = 4 s1 σ n s σ n s 1 s1 σ n +s s σ n σ n ). Proof. Substituting t 1 = n+ 1 n 1 6s 1 and t = n+ 1 n 1 6s into 3.31), and noting that σ n s 1,s ) := σ n n+ 1 n 1 6s 1, n+ 1 n 1 6s ), we obtain [ ) s1 σ n + s σ n ) s1 σ n ) s σ n + s s1 σ n + s σ n s1 σ n + s1 s σ n ] 4 s1 σ n s σ n s 1 s1 σ n +s s σ n σ n ) + n 1 6 s1 σ n s σ n s1 σ n + s σ n ) 3 +On 3 ) = 0. Since the terms of On 3) are very complicated, we do not write down the detailed result here. Letting n and disregarding the terms of On 1 6) and higher order terms, we have s1 σ n s σ n + s s1 σ n ) + s σ n s1 σ n + s1 s σ n ) = 4 s1 σ n s σ n s 1 s1 σ n +s s σ n σ n ). In the end, we mention a special case, namely B 1 = B and t 1 = t. This case makes the weight wx,t 1,t ) an even function, and is a generalization of [0] on the symmetric gap probability distribution of the Gaussian unitary ensemble. 7
28 4 Conclusion We study the Hankel determinants for a Gaussian weight with one and two jumps. For the single jump case, by using the ladder operator approach, we obtain three auxiliary quantities r n t 1 ), R n t 1 ), σ n t 1 ), related to the Hankel determinant D n t 1 ). From.3),.51) and.49) we know the relations are σ n t 1 ) = d dt 1 lnd n t 1 ), r n t 1 ) = 1 σ n t 1), and R n t 1 ) = t 1σ n t 1) σ n t 1 ) σ n t 1) 4n+σ n t. 1) We show that σ n t 1 ) satisfies both the continuous and discrete σ form of the Painlevé IV. yt 1 ) = R n t 1 ) satisfies a Painlevé IV and vt 1 ) = r n t 1 ) n satisfies a Chazy II. R 3 nt 1 ) and r n t 1 ) also satisfy a second order non-linear difference equations respectively. Moreover, we consider the large n behavior of the corresponding monic orthogonal polynomials and prove that they satisfy the biconfluent Heun equation. We also consider the large n asymptotics of the auxiliary quantities under a double scaling, which gives rise to the Painlevé XXXIV equation. For the general case with two jumps, we also use the ladder operator approach to obtain a partial differential equation to describe the Hankel determinant D n t 1,t ), which is a two variables generalization of the Jimbo- Miwa-Okamoto σ form of the Painlevé IV. Acknowledgments This work of Chao Min was supported by the Scientific Research Funds of Huaqiao University under grant number Z17Y0054. Yang Chen was supported by the Macau Science and Technology Development Fund under grant numbers FDCT 130/014/A3, FDCT 03/017/A1 and the University of Macau through MYRG FST, MYRG FST. 8
29 References [1] E. L. Basor, Distribution functions for random variables for ensembles of positive Hermitian matrices, Commun. Math. Phys ) [] E. L. Basor, Y. Chen, P ainlevé V and the distribution function of a discontinuous linear statistic in the Laguerre unitary ensembles, J. Phys. A: Math. Theor ) [3] E. L. Basor, Y. Chen, L. Zhang, PDEs satisfied by extreme eigenvalues distributions of GUE and LUE, Random Matrices: Theor. Appl. 1 01) [4] A. Bogatskiy, T. Claeys, A. Its, Hankel determinant and orthogonal polynomials for a Gaussian weight with a discontinuity at the edge, Commun. Math. Phys ) [5] Y. Chen, M. E. H. Ismail, Ladder operators and differential equations for orthogonal polynomials, J. Phys. A: Math. Gen ) [6] Y. Chen, N. Lawrence, On the linear statistics of Hermitian random matrices, J. Phys. A: Math. Gen ) [7] Y. Chen, S. M. Manning, Distribution of linear statistics in random matrix models, J. Phys.: Condes. Matter ) [8] Y. Chen, G. Pruessner, Orthogonal polynomials with discontinuous weights, J. Phys. A: Math. Gen ) L191 L198. [9] Y. Chen, L. Zhang, Painlevé VI and the unitary Jacobi ensembles, Stud. Appl. Math ) [10] T. S. Chihara, An introduction to orthogonal polynomials, Dover Publications, New York, [11] C. M. Cosgrove, Chazy s second-degree Painlevé equations, J. Phys. A: Math. Gen ) [1] D. Dai, L. Zhang, Painlevé VI and Hankel determinants for the generalized Jacobi weight, J. Phys. A: Math. Theor )
30 [13] P. J. Forrester, N. S. Witte, Discrete Painlevé equations and random matrix averages, Nonlinearity ) [14] V. I. Gromak, I. Laine, S. Shimomura, Painlevé Differential Equations in the Complex Plane, Walter de Gruyter, Berlin, New York, 00. [15] L. Haine, J.-P. Semengue, The Jacobi polynomial ensemble and the Painlevé VI equation, J. Math. Phys ) [16] A. Its, Discrete Painlevé Equations and Orthogonal Polynomials, In Symmetrices and Integrability of Difference Equations, Ed. by D. Levi, P. Olver, Z. Thomova, and P. Winternitz, London Mathematical Society Lecture Note Series 381, Cambridge University Press 011) [17] A. Its, I. Krasovsky, Hankel determinant and orthogonal polynomials for the Gaussian weight with a jump, In Integrable Systems and Random Matrices, Contemp. Math. Amer. Math. Soc., vol. 458, Providence, RI 008) [18] M. Jimbo, T. Miwa, Monodromy perserving deformation of linear ordinary differential equations with rational coefficients. II, Physica D 1981) [19] K. Johansson, On fluctuations of eigenvalues of random Hermitian matrices, Duke Math. J ) [0] S. Lyu, Y. Chen, E. Fan, Asymptotic gap probability distributions of the Gaussian unitary ensembles and Jacobi unitary ensembles, Nucl. Phys. B ) [1] A. P. Magnus, Painlevé-type differential equations for the recurrence coefficients of semiclassical orthogonal polynomials, J. Comput. Appl. Math ) [] M. L. Mehta, Random Matrices, 3rd edn., Elsevier, New York, 004. [3] C. Min, Y. Chen, Linear statistics of matrix ensembles in classical background, Math. Meth. Appl. Sci ) [4] C. Min, Y. Chen, On the variance of linear statistics of Hermitian random matrices, Acta Phys. Pol. B )
31 [5] C. Min, Y. Chen, Gap probability distribution of the Jacobi unitary ensemble: an elementary treatment, from finite n to double scaling, Stud. Appl. Math ) 0 0. [6] A. Perret, G. Schehr, Near-extreme eigenvalues and the first gap of Hermitian random matrices, J. Stat. Phys ) [7] A. Ramani, B. Grammaticos, J. Hietarinta, Discrete versions of the Painlevé equations, Phys. Rev. Lett ) [8] A. Ramani, B. Grammaticos, Discrete Painlevé equations: coalescences, limits and degeneracies, Physica A ) [9] A. Ronveaux, Heun s Differential Equations, Oxford Science Publications, Oxford, [30] S. Yu. Slavyanov, W. Lay, Special Functions: A Unified Theory Based on Singularities, Oxford University Press, Oxford, 000. [31] G. Szegő, Orthogonal Polynomials, 4th edn., AMS Colloquium Publications, Vol. 3, Providence, RI, [3] C. A. Tracy, H. Widom, Fredholm determinants, differential equations and matrix models, Commun. Math. Phys ) [33] N. S. Witte, P. J. Forrester, C. M. Cosgrove, Gap probabilities for edge intervals in finite Gaussian and Jacobi unitary matrix ensembles, Nonlinearity ) [34] S.-X. Xu, Y.-Q. Zhao, Painlevé XXXIV asymptotics of orthogonal polynomials for the Gaussian weight with a jump at the edge, Stud. Appl. Math )
Painlevé VI and Hankel determinants for the generalized Jacobi weight
arxiv:98.558v2 [math.ca] 3 Nov 29 Painlevé VI and Hankel determinants for the generalized Jacobi weight D. Dai Department of Mathematics, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong
More informationPainlevé VI and the Unitary Jacobi ensembles
Painlevé VI and the Unitary Jacobi ensembles arxiv:0911.5636v3 [math.ca] 23 Dec 2009 Yang Chen Department of Mathematics, Imperial College London, 180 Queen s Gates, London SW7 2BZ, UK ychen@imperial.ac.uk
More informationThe Density Matrix for the Ground State of 1-d Impenetrable Bosons in a Harmonic Trap
The Density Matrix for the Ground State of 1-d Impenetrable Bosons in a Harmonic Trap Institute of Fundamental Sciences Massey University New Zealand 29 August 2017 A. A. Kapaev Memorial Workshop Michigan
More informationOrthogonal Polynomials, Perturbed Hankel Determinants. and. Random Matrix Models
Orthogonal Polynomials, Perturbed Hankel Determinants and Random Matrix Models A thesis presented for the degree of Doctor of Philosophy of Imperial College London and the Diploma of Imperial College by
More informationNear extreme eigenvalues and the first gap of Hermitian random matrices
Near extreme eigenvalues and the first gap of Hermitian random matrices Grégory Schehr LPTMS, CNRS-Université Paris-Sud XI Sydney Random Matrix Theory Workshop 13-16 January 2014 Anthony Perret, G. S.,
More informationPainlevé IV and degenerate Gaussian Unitary Ensembles
Painlevé IV and degenerate Gaussian Unitary Ensembles arxiv:math-ph/0606064v1 8 Jun 006 Yang Chen Department of Mathematics Imperial College London 180 Queen s Gates London SW7 BZ UK M. V. Feigin Department
More informationFredholm determinant with the confluent hypergeometric kernel
Fredholm determinant with the confluent hypergeometric kernel J. Vasylevska joint work with I. Krasovsky Brunel University Dec 19, 2008 Problem Statement Let K s be the integral operator acting on L 2
More informationExceptional solutions to the Painlevé VI equation associated with the generalized Jacobi weight
Random Matrices: Theory and Applications Vol. 6, No. (07) 75000 ( pages) c World Scientific Publishing Company DOI: 0.4/S006750004 Exceptional solutions to the Painlevé VI equation associated with the
More informationPainlevé equations and orthogonal polynomials
KU Leuven, Belgium Kapaev workshop, Ann Arbor MI, 28 August 2017 Contents Painlevé equations (discrete and continuous) appear at various places in the theory of orthogonal polynomials: Discrete Painlevé
More informationOPSF, Random Matrices and Riemann-Hilbert problems
OPSF, Random Matrices and Riemann-Hilbert problems School on Orthogonal Polynomials in Approximation Theory and Mathematical Physics, ICMAT 23 27 October, 2017 Plan of the course lecture 1: Orthogonal
More informationRANDOM MATRIX THEORY AND TOEPLITZ DETERMINANTS
RANDOM MATRIX THEORY AND TOEPLITZ DETERMINANTS David García-García May 13, 2016 Faculdade de Ciências da Universidade de Lisboa OVERVIEW Random Matrix Theory Introduction Matrix ensembles A sample computation:
More informationarxiv:hep-th/ v1 14 Oct 1992
ITD 92/93 11 Level-Spacing Distributions and the Airy Kernel Craig A. Tracy Department of Mathematics and Institute of Theoretical Dynamics, University of California, Davis, CA 95616, USA arxiv:hep-th/9210074v1
More informationOn the singularities of non-linear ODEs
On the singularities of non-linear ODEs Galina Filipuk Institute of Mathematics University of Warsaw G.Filipuk@mimuw.edu.pl Collaborators: R. Halburd (London), R. Vidunas (Tokyo), R. Kycia (Kraków) 1 Plan
More informationarxiv: v2 [nlin.si] 3 Feb 2016
On Airy Solutions of the Second Painlevé Equation Peter A. Clarkson School of Mathematics, Statistics and Actuarial Science, University of Kent, Canterbury, CT2 7NF, UK P.A.Clarkson@kent.ac.uk October
More informationMarkov operators, classical orthogonal polynomial ensembles, and random matrices
Markov operators, classical orthogonal polynomial ensembles, and random matrices M. Ledoux, Institut de Mathématiques de Toulouse, France 5ecm Amsterdam, July 2008 recent study of random matrix and random
More informationExponential tail inequalities for eigenvalues of random matrices
Exponential tail inequalities for eigenvalues of random matrices M. Ledoux Institut de Mathématiques de Toulouse, France exponential tail inequalities classical theme in probability and statistics quantify
More informationThe Generating Functions for Pochhammer
The Generating Functions for Pochhammer Symbol { }, n N Aleksandar Petoević University of Novi Sad Teacher Training Faculty, Department of Mathematics Podgorička 4, 25000 Sombor SERBIA and MONTENEGRO Email
More informationReferences 167. dx n x2 =2
References 1. G. Akemann, J. Baik, P. Di Francesco (editors), The Oxford Handbook of Random Matrix Theory, Oxford University Press, Oxford, 2011. 2. G. Anderson, A. Guionnet, O. Zeitouni, An Introduction
More informationPainlevé Transcendents and the Information Theory of MIMO Systems
Painlevé Transcendents and the Information Theory of MIMO Systems Yang Chen Department of Mathematics Imperial College, London [Joint with Matthew R. McKay, Department of Elec. and Comp. Engineering, HKUST]
More informationNUMERICAL CALCULATION OF RANDOM MATRIX DISTRIBUTIONS AND ORTHOGONAL POLYNOMIALS. Sheehan Olver NA Group, Oxford
NUMERICAL CALCULATION OF RANDOM MATRIX DISTRIBUTIONS AND ORTHOGONAL POLYNOMIALS Sheehan Olver NA Group, Oxford We are interested in numerically computing eigenvalue statistics of the GUE ensembles, i.e.,
More informationA determinantal formula for the GOE Tracy-Widom distribution
A determinantal formula for the GOE Tracy-Widom distribution Patrik L. Ferrari and Herbert Spohn Technische Universität München Zentrum Mathematik and Physik Department e-mails: ferrari@ma.tum.de, spohn@ma.tum.de
More informationRecurrence Coef f icients of a New Generalization of the Meixner Polynomials
Symmetry, Integrability and Geometry: Methods and Applications SIGMA 7 (011), 068, 11 pages Recurrence Coef f icients of a New Generalization of the Meixner Polynomials Galina FILIPUK and Walter VAN ASSCHE
More informationRandom Matrices and Multivariate Statistical Analysis
Random Matrices and Multivariate Statistical Analysis Iain Johnstone, Statistics, Stanford imj@stanford.edu SEA 06@MIT p.1 Agenda Classical multivariate techniques Principal Component Analysis Canonical
More informationQUADRATIC EQUATIONS AND MONODROMY EVOLVING DEFORMATIONS
QUADRATIC EQUATIONS AND MONODROMY EVOLVING DEFORMATIONS YOUSUKE OHYAMA 1. Introduction In this paper we study a special class of monodromy evolving deformations (MED). Chakravarty and Ablowitz [4] showed
More informationDifferential Equations for Dyson Processes
Differential Equations for Dyson Processes Joint work with Harold Widom I. Overview We call Dyson process any invariant process on ensembles of matrices in which the entries undergo diffusion. Dyson Brownian
More informationCenter of Mass Distribution of The Jacobi Unitary Ensembles: Painlevé V, Asymptotic Expansions
Center of Mass Distribution of The Jacobi Unitary Ensembles: Painlevé V, Asymptotic Expansions arxiv:80.07454v [math.ca] 3 Jan 08 Longjun Zhan a,, Gordon Blower b,, Yang Chen a, Mengkun Zhu a, a Department
More informationarxiv:solv-int/ v1 20 Oct 1993
Casorati Determinant Solutions for the Discrete Painlevé-II Equation Kenji Kajiwara, Yasuhiro Ohta, Junkichi Satsuma, Basil Grammaticos and Alfred Ramani arxiv:solv-int/9310002v1 20 Oct 1993 Department
More informationA Note about the Pochhammer Symbol
Mathematica Moravica Vol. 12-1 (2008), 37 42 A Note about the Pochhammer Symbol Aleksandar Petoević Abstract. In this paper we give elementary proofs of the generating functions for the Pochhammer symbol
More informationUniversality of distribution functions in random matrix theory Arno Kuijlaars Katholieke Universiteit Leuven, Belgium
Universality of distribution functions in random matrix theory Arno Kuijlaars Katholieke Universiteit Leuven, Belgium SEA 06@MIT, Workshop on Stochastic Eigen-Analysis and its Applications, MIT, Cambridge,
More informationarxiv:nlin/ v1 [nlin.si] 29 May 2002
arxiv:nlin/0205063v1 [nlinsi] 29 May 2002 On a q-difference Painlevé III Equation: II Rational Solutions Kenji KAJIWARA Graduate School of Mathematics, Kyushu University 6-10-1 Hakozaki, Higashi-ku, Fukuoka
More informationNew explicit solitary wave solutions for (2 + 1)-dimensional Boussinesq equation and (3 + 1)-dimensional KP equation
Physics Letters A 07 (00) 107 11 www.elsevier.com/locate/pla New explicit solitary wave solutions for ( + 1)-dimensional Boussinesq equation and ( + 1)-dimensional KP equation Yong Chen, Zhenya Yan, Honging
More informationTriangular matrices and biorthogonal ensembles
/26 Triangular matrices and biorthogonal ensembles Dimitris Cheliotis Department of Mathematics University of Athens UK Easter Probability Meeting April 8, 206 2/26 Special densities on R n Example. n
More informationSTAT C206A / MATH C223A : Stein s method and applications 1. Lecture 31
STAT C26A / MATH C223A : Stein s method and applications Lecture 3 Lecture date: Nov. 7, 27 Scribe: Anand Sarwate Gaussian concentration recap If W, T ) is a pair of random variables such that for all
More informationMultiple orthogonal polynomials. Bessel weights
for modified Bessel weights KU Leuven, Belgium Madison WI, December 7, 2013 Classical orthogonal polynomials The (very) classical orthogonal polynomials are those of Jacobi, Laguerre and Hermite. Classical
More informationUniversal Fluctuation Formulae for one-cut β-ensembles
Universal Fluctuation Formulae for one-cut β-ensembles with a combinatorial touch Pierpaolo Vivo with F. D. Cunden Phys. Rev. Lett. 113, 070202 (2014) with F.D. Cunden and F. Mezzadri J. Phys. A 48, 315204
More informationGap probability of a one-dimensional gas model
Gap probability of a one-dimensional gas model arxiv:cond-mat/9605049v 9 May 1996 Y Chen 1, and S. M. Manning, 1 Department of Mathematics, Imperial College 180 Queen s Gate,London SW7 BZ UK Department
More informationPainlevé III asymptotics of Hankel determinants for a perturbed Jacobi weight
Painlevé III asymptotics of Hankel determinants for a perturbed Jacobi weight arxiv:4.8586v [math-ph] 9 May 05 Zhao-Yun Zeng a, Shuai-Xia Xu b, and Yu-Qiu Zhao a a Department of Mathematics, Sun Yat-sen
More informationA Remark on Hypercontractivity and Tail Inequalities for the Largest Eigenvalues of Random Matrices
A Remark on Hypercontractivity and Tail Inequalities for the Largest Eigenvalues of Random Matrices Michel Ledoux Institut de Mathématiques, Université Paul Sabatier, 31062 Toulouse, France E-mail: ledoux@math.ups-tlse.fr
More informationarxiv: v2 [math-ph] 13 Feb 2013
FACTORIZATIONS OF RATIONAL MATRIX FUNCTIONS WITH APPLICATION TO DISCRETE ISOMONODROMIC TRANSFORMATIONS AND DIFFERENCE PAINLEVÉ EQUATIONS ANTON DZHAMAY SCHOOL OF MATHEMATICAL SCIENCES UNIVERSITY OF NORTHERN
More informationThis ODE arises in many physical systems that we shall investigate. + ( + 1)u = 0. (λ + s)x λ + s + ( + 1) a λ. (s + 1)(s + 2) a 0
Legendre equation This ODE arises in many physical systems that we shall investigate We choose We then have Substitution gives ( x 2 ) d 2 u du 2x 2 dx dx + ( + )u u x s a λ x λ a du dx λ a λ (λ + s)x
More informationarxiv: v3 [math-ph] 23 May 2017
arxiv:161.8561v3 [math-ph] 3 May 17 On the probability of positive-definiteness in the ggue via semi-classical Laguerre polynomials Alfredo Deaño and icholas J. Simm May 5, 17 Abstract In this paper, we
More informationExtreme eigenvalue fluctutations for GUE
Extreme eigenvalue fluctutations for GUE C. Donati-Martin 204 Program Women and Mathematics, IAS Introduction andom matrices were introduced in multivariate statistics, in the thirties by Wishart [Wis]
More informationarxiv: v9 [math-ph] 6 Nov 2014
Generalization of the three-term recurrence formula and its applications Yoon Seok Choun Baruch College, The City University of New York, Natural Science Department, A506, 7 Lexington Avenue, New York,
More informationDouble contour integral formulas for the sum of GUE and one matrix model
Double contour integral formulas for the sum of GUE and one matrix model Based on arxiv:1608.05870 with Tom Claeys, Arno Kuijlaars, and Karl Liechty Dong Wang National University of Singapore Workshop
More informationNonlinear Integral Equation Formulation of Orthogonal Polynomials
Nonlinear Integral Equation Formulation of Orthogonal Polynomials Eli Ben-Naim Theory Division, Los Alamos National Laboratory with: Carl Bender (Washington University, St. Louis) C.M. Bender and E. Ben-Naim,
More informationUNIFORM BOUNDS FOR BESSEL FUNCTIONS
Journal of Applied Analysis Vol. 1, No. 1 (006), pp. 83 91 UNIFORM BOUNDS FOR BESSEL FUNCTIONS I. KRASIKOV Received October 8, 001 and, in revised form, July 6, 004 Abstract. For ν > 1/ and x real we shall
More informationBurgers equation in the complex plane. Govind Menon Division of Applied Mathematics Brown University
Burgers equation in the complex plane Govind Menon Division of Applied Mathematics Brown University What this talk contains Interesting instances of the appearance of Burgers equation in the complex plane
More informationRectangular Young tableaux and the Jacobi ensemble
Rectangular Young tableaux and the Jacobi ensemble Philippe Marchal October 20, 2015 Abstract It has been shown by Pittel and Romik that the random surface associated with a large rectangular Young tableau
More informationRecurrence Relations and Fast Algorithms
Recurrence Relations and Fast Algorithms Mark Tygert Research Report YALEU/DCS/RR-343 December 29, 2005 Abstract We construct fast algorithms for decomposing into and reconstructing from linear combinations
More informationarxiv:nlin/ v2 [nlin.si] 9 Oct 2002
Journal of Nonlinear Mathematical Physics Volume 9, Number 1 2002), 21 25 Letter On Integrability of Differential Constraints Arising from the Singularity Analysis S Yu SAKOVICH Institute of Physics, National
More informationCentral Limit Theorems for linear statistics for Biorthogonal Ensembles
Central Limit Theorems for linear statistics for Biorthogonal Ensembles Maurice Duits, Stockholm University Based on joint work with Jonathan Breuer (HUJI) Princeton, April 2, 2014 M. Duits (SU) CLT s
More informationJINHO BAIK Department of Mathematics University of Michigan Ann Arbor, MI USA
LIMITING DISTRIBUTION OF LAST PASSAGE PERCOLATION MODELS JINHO BAIK Department of Mathematics University of Michigan Ann Arbor, MI 48109 USA E-mail: baik@umich.edu We survey some results and applications
More informationThe Spectral Theory of the X 1 -Laguerre Polynomials
Advances in Dynamical Systems and Applications ISSN 973-5321, Volume x, Number x, pp. 25 36 (2xx) http://campus.mst.edu/adsa The Spectral Theory of the X 1 -Laguerre Polynomials Mohamed J. Atia a, Lance
More informationOn the Transformations of the Sixth Painlevé Equation
Journal of Nonlinear Mathematical Physics Volume 10, Supplement 2 (2003), 57 68 SIDE V On the Transformations of the Sixth Painlevé Equation Valery I GROMAK andgalinafilipuk Department of differential
More informationOn Tricomi and Hermite-Tricomi Matrix Functions of Complex Variable
Communications in Mathematics and Applications Volume (0), Numbers -3, pp. 97 09 RGN Publications http://www.rgnpublications.com On Tricomi and Hermite-Tricomi Matrix Functions of Complex Variable A. Shehata
More informationPositivity of Turán determinants for orthogonal polynomials
Positivity of Turán determinants for orthogonal polynomials Ryszard Szwarc Abstract The orthogonal polynomials p n satisfy Turán s inequality if p 2 n (x) p n 1 (x)p n+1 (x) 0 for n 1 and for all x in
More informationMATRIX INTEGRALS AND MAP ENUMERATION 2
MATRIX ITEGRALS AD MAP EUMERATIO 2 IVA CORWI Abstract. We prove the generating function formula for one face maps and for plane diagrams using techniques from Random Matrix Theory and orthogonal polynomials.
More informationConcentration Inequalities for Random Matrices
Concentration Inequalities for Random Matrices M. Ledoux Institut de Mathématiques de Toulouse, France exponential tail inequalities classical theme in probability and statistics quantify the asymptotic
More informationHermite Interpolation and Sobolev Orthogonality
Acta Applicandae Mathematicae 61: 87 99, 2000 2000 Kluwer Academic Publishers Printed in the Netherlands 87 Hermite Interpolation and Sobolev Orthogonality ESTHER M GARCÍA-CABALLERO 1,, TERESA E PÉREZ
More informationMATRIX KERNELS FOR THE GAUSSIAN ORTHOGONAL AND SYMPLECTIC ENSEMBLES
Ann. Inst. Fourier, Grenoble 55, 6 (5), 197 7 MATRIX KERNELS FOR THE GAUSSIAN ORTHOGONAL AND SYMPLECTIC ENSEMBLES b Craig A. TRACY & Harold WIDOM I. Introduction. For a large class of finite N determinantal
More informationConstrained Leja points and the numerical solution of the constrained energy problem
Journal of Computational and Applied Mathematics 131 (2001) 427 444 www.elsevier.nl/locate/cam Constrained Leja points and the numerical solution of the constrained energy problem Dan I. Coroian, Peter
More informationMultiple Orthogonal Polynomials
Summer school on OPSF, University of Kent 26 30 June, 2017 Introduction For this course I assume everybody is familiar with the basic theory of orthogonal polynomials: Introduction For this course I assume
More informationMaximal height of non-intersecting Brownian motions
Maximal height of non-intersecting Brownian motions G. Schehr Laboratoire de Physique Théorique et Modèles Statistiques CNRS-Université Paris Sud-XI, Orsay Collaborators: A. Comtet (LPTMS, Orsay) P. J.
More informationRational Form Solitary Wave Solutions and Doubly Periodic Wave Solutions to (1+1)-Dimensional Dispersive Long Wave Equation
Commun. Theor. Phys. (Beijing, China) 43 (005) pp. 975 98 c International Academic Publishers Vol. 43, No. 6, June 15, 005 Rational Form Solitary Wave Solutions and Doubly Periodic Wave Solutions to (1+1)-Dimensional
More informationULTRASPHERICAL TYPE GENERATING FUNCTIONS FOR ORTHOGONAL POLYNOMIALS
ULTRASPHERICAL TYPE GENERATING FUNCTIONS FOR ORTHOGONAL POLYNOMIALS arxiv:083666v [mathpr] 8 Jan 009 Abstract We characterize, up to a conjecture, probability distributions of finite all order moments
More informationA remark on a variable-coefficient Bernoulli equation based on auxiliary -equation method for nonlinear physical systems
A remark on a variable-coefficient Bernoulli equation based on auxiliary -equation method for nonlinear physical systems Zehra Pınar a Turgut Öziş b a Namık Kemal University, Faculty of Arts and Science,
More informationRecurrence coefficients of generalized Meixner polynomials and Painlevé equations
Recurrence coefficients of generalized Meixner polynomials and Painlevé equations Lies Boelen, Galina Filipuk, Walter Van Assche October 0, 010 Abstract We consider a semi-classical version of the Meixner
More informationNumerical Evaluation of Standard Distributions in Random Matrix Theory
Numerical Evaluation of Standard Distributions in Random Matrix Theory A Review of Folkmar Bornemann s MATLAB Package and Paper Matt Redmond Department of Mathematics Massachusetts Institute of Technology
More informationSpecial solutions of the third and fifth Painlevé equations and vortex solutions of the complex Sine-Gordon equations
Special solutions of the third and fifth Painlevé equations and vortex solutions of the complex Sine-Gordon equations Peter A Clarkson School of Mathematics, Statistics and Actuarial Science University
More informationOrthogonal polynomials
Orthogonal polynomials Gérard MEURANT October, 2008 1 Definition 2 Moments 3 Existence 4 Three-term recurrences 5 Jacobi matrices 6 Christoffel-Darboux relation 7 Examples of orthogonal polynomials 8 Variable-signed
More informationA New Integrable Couplings of Classical-Boussinesq Hierarchy with Self-Consistent Sources
Commun. Theor. Phys. Beijing, China 54 21 pp. 1 6 c Chinese Physical Society and IOP Publishing Ltd Vol. 54, No. 1, July 15, 21 A New Integrable Couplings of Classical-Boussinesq Hierarchy with Self-Consistent
More informationBessel Functions Michael Taylor. Lecture Notes for Math 524
Bessel Functions Michael Taylor Lecture Notes for Math 54 Contents 1. Introduction. Conversion to first order systems 3. The Bessel functions J ν 4. The Bessel functions Y ν 5. Relations between J ν and
More informationPainlevé Representations for Distribution Functions for Next-Largest, Next-Next-Largest, etc., Eigenvalues of GOE, GUE and GSE
Painlevé Representations for Distribution Functions for Next-Largest, Next-Next-Largest, etc., Eigenvalues of GOE, GUE and GSE Craig A. Tracy UC Davis RHPIA 2005 SISSA, Trieste 1 Figure 1: Paul Painlevé,
More informationNumerical analysis and random matrix theory. Tom Trogdon UC Irvine
Numerical analysis and random matrix theory Tom Trogdon ttrogdon@math.uci.edu UC Irvine Acknowledgements This is joint work with: Percy Deift Govind Menon Sheehan Olver Raj Rao Numerical analysis and random
More informationLinear and nonlinear ODEs and middle convolution
Linear and nonlinear ODEs and middle convolution Galina Filipuk Institute of Mathematics University of Warsaw G.Filipuk@mimuw.edu.pl Collaborator: Y. Haraoka (Kumamoto University) 1 Plan of the Talk: Linear
More informationIntroduction to the Hirota bilinear method
Introduction to the Hirota bilinear method arxiv:solv-int/9708006v1 14 Aug 1997 J. Hietarinta Department of Physics, University of Turku FIN-20014 Turku, Finland e-mail: hietarin@utu.fi Abstract We give
More informationSpectral Theory of X 1 -Laguerre Polynomials
Advances in Dynamical Systems and Applications ISSN 973-5321, Volume 8, Number 2, pp. 181 192 (213) http://campus.mst.edu/adsa Spectral Theory of X 1 -Laguerre Polynomials Mohamed J. Atia Université de
More informationSpecial Function Solutions of a Class of Certain Non-autonomous Nonlinear Ordinary Differential Equations IJSER. where % "!
International Journal of Scientific & Engineering Research, Volume 6, Issue 3, March-2015 Special Function Solutions of a Class of Certain Non-autonomous Nonlinear Ordinary Differential Equations,, $ Abstract
More informationDeterminants of Hankel Matrices
Determinants of Hankel Matrices arxiv:math/67v1 [math.ca] 8 Jun 2 Estelle L. Basor Department of Mathematics California Polytechnic State University San Luis Obispo, CA 9347, USA Yang Chen Department of
More informationEigenvalue PDFs. Peter Forrester, M&S, University of Melbourne
Outline Eigenvalue PDFs Peter Forrester, M&S, University of Melbourne Hermitian matrices with real, complex or real quaternion elements Circular ensembles and classical groups Products of random matrices
More informationSingle-User MIMO System, Painlevé Transcendents and Double Scaling
Single-User MIMO System, Painlevé Transcendents and Double Scaling Hongmei Chen, Min Chen, Gordon Blower, Yang Chen Faculty of Science and Technology, Department of Mathematics, University of Macau. Department
More informationDifference Equations for Multiple Charlier and Meixner Polynomials 1
Difference Equations for Multiple Charlier and Meixner Polynomials 1 WALTER VAN ASSCHE Department of Mathematics Katholieke Universiteit Leuven B-3001 Leuven, Belgium E-mail: walter@wis.kuleuven.ac.be
More informationOrthogonal Symmetric Toeplitz Matrices
Orthogonal Symmetric Toeplitz Matrices Albrecht Böttcher In Memory of Georgii Litvinchuk (1931-2006 Abstract We show that the number of orthogonal and symmetric Toeplitz matrices of a given order is finite
More informationUniversality in Numerical Computations with Random Data. Case Studies.
Universality in Numerical Computations with Random Data. Case Studies. Percy Deift and Thomas Trogdon Courant Institute of Mathematical Sciences Govind Menon Brown University Sheehan Olver The University
More informationIntegrability, Nonintegrability and the Poly-Painlevé Test
Integrability, Nonintegrability and the Poly-Painlevé Test Rodica D. Costin Sept. 2005 Nonlinearity 2003, 1997, preprint, Meth.Appl.An. 1997, Inv.Math. 2001 Talk dedicated to my professor, Martin Kruskal.
More informationCOMPUTATION OF BESSEL AND AIRY FUNCTIONS AND OF RELATED GAUSSIAN QUADRATURE FORMULAE
BIT 6-85//41-11 $16., Vol. 4, No. 1, pp. 11 118 c Swets & Zeitlinger COMPUTATION OF BESSEL AND AIRY FUNCTIONS AND OF RELATED GAUSSIAN QUADRATURE FORMULAE WALTER GAUTSCHI Department of Computer Sciences,
More informationAnalogues for Bessel Functions of the Christoffel-Darboux Identity
Analogues for Bessel Functions of the Christoffel-Darboux Identity Mark Tygert Research Report YALEU/DCS/RR-1351 March 30, 2006 Abstract We derive analogues for Bessel functions of what is known as the
More informationQuadratures and integral transforms arising from generating functions
Quadratures and integral transforms arising from generating functions Rafael G. Campos Facultad de Ciencias Físico-Matemáticas, Universidad Michoacana, 58060, Morelia, México. rcampos@umich.mx Francisco
More informationPolyexponentials. Khristo N. Boyadzhiev Ohio Northern University Departnment of Mathematics Ada, OH
Polyexponentials Khristo N. Boyadzhiev Ohio Northern University Departnment of Mathematics Ada, OH 45810 k-boyadzhiev@onu.edu 1. Introduction. The polylogarithmic function [15] (1.1) and the more general
More informationThe elliptic sinh-gordon equation in the half plane
Available online at www.tjnsa.com J. Nonlinear Sci. Appl. 8 25), 63 73 Research Article The elliptic sinh-gordon equation in the half plane Guenbo Hwang Department of Mathematics, Daegu University, Gyeongsan
More informationUltraspherical moments on a set of disjoint intervals
Ultraspherical moments on a set of disjoint intervals arxiv:90.049v [math.ca] 4 Jan 09 Hashem Alsabi Université des Sciences et Technologies, Lille, France hashem.alsabi@gmail.com James Griffin Department
More informationIrina Pchelintseva A FIRST-ORDER SPECTRAL PHASE TRANSITION IN A CLASS OF PERIODICALLY MODULATED HERMITIAN JACOBI MATRICES
Opuscula Mathematica Vol. 8 No. 008 Irina Pchelintseva A FIRST-ORDER SPECTRAL PHASE TRANSITION IN A CLASS OF PERIODICALLY MODULATED HERMITIAN JACOBI MATRICES Abstract. We consider self-adjoint unbounded
More informationON THE SYMMETRIES OF INTEGRABLE PARTIAL DIFFERENCE EQUATIONS
Proceedings of the International Conference on Difference Equations, Special Functions and Orthogonal Polynomials, World Scientific (2007 ON THE SYMMETRIES OF INTEGRABLE PARTIAL DIFFERENCE EQUATIONS ANASTASIOS
More informationFluctuations from the Semicircle Law Lecture 4
Fluctuations from the Semicircle Law Lecture 4 Ioana Dumitriu University of Washington Women and Math, IAS 2014 May 23, 2014 Ioana Dumitriu (UW) Fluctuations from the Semicircle Law Lecture 4 May 23, 2014
More informationLocal and global finite branching of solutions of ODEs in the complex plane
Local and global finite branching of solutions of ODEs in the complex plane Workshop on Singularities and Nonlinear ODEs Thomas Kecker University College London / University of Portsmouth Warszawa, 7 9.11.2014
More informationOPSF, Random Matrices and Riemann-Hilbert problems
OPSF, Random Matrices and Riemann-Hilbert problems School on Orthogonal Polynomials in Approximation Theory and Mathematical Physics, ICMAT 23 27 October, 207 Plan of the course lecture : Orthogonal Polynomials
More informationSpectral analysis of two doubly infinite Jacobi operators
Spectral analysis of two doubly infinite Jacobi operators František Štampach jointly with Mourad E. H. Ismail Stockholm University Spectral Theory and Applications conference in memory of Boris Pavlov
More informationTHREE LECTURES ON QUASIDETERMINANTS
THREE LECTURES ON QUASIDETERMINANTS Robert Lee Wilson Department of Mathematics Rutgers University The determinant of a matrix with entries in a commutative ring is a main organizing tool in commutative
More informationBäcklund transformations for fourth-order Painlevé-type equations
Bäcklund transformations for fourth-order Painlevé-type equations A. H. Sakka 1 and S. R. Elshamy Department of Mathematics, Islamic University of Gaza P.O.Box 108, Rimae, Gaza, Palestine 1 e-mail: asakka@mail.iugaza.edu
More information