Let's begin with some termes:

  1. For $1\le i<j<\infty$ let $X_{i,j}$ be independant and indentically distributed (i.i.d) real random variables with mean $0$ and variance $1$ and set $X_{j,i}=X_{i,j}$. Let $X_{i,i}$ be i.i.d. real random variables (with possibly a different distribution) with mean $0$ and variance $1$. Then $M_n=\left[ X_{i,j}\right]_{i,j=1}^n$ will be a random $n\times n$ symmetric matrix.

  1. There are $n$ random eigenvalues which we will denote by $$ \lambda_1\le \lambda_2\le \dots \lambda_n. $$ The empirical spectral measure is $$ \nu^*_n=\frac1n \sum_{i=1}^n \delta_{\lambda_i}. $$ This is a random discrete probability measure which puts $n^{-1}$ mass to each (random) eigenvalue.

  1. theorem:(Wigner's semicircle law) Let $$ \nu_n=\frac1n \sum_{i=1}^n \delta_{\frac{\lambda_i}{\sqrt{n}}}. $$ be the normalized empirical spectral measure. Then as $n\to \infty$ we have $$ \nu_n \Rightarrow \nu \qquad \text{a.s.} $$ where $\nu$ has density $$ \frac{d\nu}{dx}=\frac{1}{2\pi} \sqrt{4-x^2}\,\chi_{\{|x|\le 2\}}. $$ ($\chi_A$ denote the indicatrice function of $A$)

Question Using simulation to understand this theorem in other word: Do the same thing that was done for the central limit theorem as it is shown in this page CLT I have already posted here about the same subject. Any help is welcome.

  • 2
    What is your question? – rm -rf Mar 9 '15 at 4:09
  • 1
    In the linked question, you already seem to have a simulation. So what is the Mathematica issue you want to address here? – Jens Mar 9 '15 at 5:14
  • My question is to do the same thing that was done for the central limit theorem as it is shown in this page demonstrations.wolfram.com/TheCentralLimitTheorem – Zbigniew Mar 9 '15 at 6:51

Your Answer

 
discard

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.