A Julia package containing Distributions.jl
type specifications for various distributions arising from random matrix theory.

SpikedWigner(beta, n, spikes; scaled=false)
: Wigner distribution with an added spike matrix.spikes
is an array[s1, ..., sr]
such that the diagonal matrix with diagonal sqrt(n)(s_{1}, ... , s_{r}, 0, ..., 0) is added to a white Wigner matrix. 
SpikedWishart(beta, n, p, spikes; scaled=false)
: Wishart distribution with spiked covariance [1].spikes
is an array[s1, ..., sr]
such that the Wishart covariance is diagonal with entries (1 + s_{1}, ... , 1 + s_{r}, 1, ..., 1). 
Jacobi(beta, n1, n2, p)
: Random matrices of the form E(E+H)^{1}. Here E and H are (n_{1}, p) and (n_{2}, p) white Wisharts respectively. [2]
Specifying scaled=true
in SpikedWigner
and SpikedWishart
scales the matrices by an appropriate function of n so that the corresponding bulks converge to the semicircle and MarchenkoPastur laws respectively.
Due to the inverse in the definition of the Jacobi ensemble, no scaling is necessary for Jacobi
,
Normal entries in Gaussian ensembles are scaled to have variance 1.
The package implements the following types:
MarchenkoPastur(gamma)
: Limiting empirical spectral density of a real white Wishart matrix with p/n > gamma as long as 0 < gamma < 1.TracyWidom(beta)
: Limiting distribution of the maximum eigenvalue of many random matrix ensembles with Dyson parameter beta [3].Wachter(gamma1, gamma2)
: Limiting empirical spectral density of S_{1} S_{2}^{1}. Here S_{1} and S_{2} are sample covariance matrices with n_{1}/p > gamma_{1} and n_{2}/p > gamma_{2}.
It also implements the following functions for computing eigenvalue distributions:
supercrit_dist(E)
: Approximate distribution of the supercritical eigenvalues of a matrix drawn from the ensemble E. Currently implemented for Wishart with beta = 1 [4], beta = 2 [5] and Wigner with beta = 1 [6], beta = 2 [7]
The function randeigvals
efficiently samples from the distribution of eigenvalues of the implemented random matrix distributions. It does this by generating a tridiagonal or banded matrix with eigenvalue equal in distribution to the specified model.
See the documentation.
[1] Dumitriu & Edelman, "Matrix Models for beta ensembles," Journal of Mathematical Physics, 11 (2002).
[2] Killip & Nenciu, "Matrix Models for Circular Ensembles," International Mathematics Research Notices, 50 (2004).
[3] Bornemann, "On the numerical evaluation of distributions in random matrix theory: a review," (2010).
[4] Baik, Ben Arous & Peche, "Phase transition of the largest eigenvalue for nonnull complex sample covariance matrices," The Annals of Probability, 33 (2005).
[5] Paul, "Asymptotics of sample eigenstructure for a large dimensional spiked covariance model," Statistica Sinica, 17 (2007).
[6] Feral, Peche, "The largest eigenvalue of rank one deformation of large wigner matrices," Commun. Math. Phys., 272 (2007).
[7] Peche, "The largest eigenvalue of small rank perturbations of Hermitian random matrices," Probability Theory and Related Fields, 134 (2006).