Wikipedia has the following definition of a normal random vector:

A real random vector $\mathbf{X} = (X_1,\ldots,X_k)^{\mathrm T}$ is called a normal random vector if there exists a random $\ell$-vector $\mathbf{Z}$, which is a standard normal random vector, a $k$-vector $\mathbf{\mu}$, and a $k \times \ell$ matrix $\boldsymbol{A}$, such that $\mathbf{X}=\boldsymbol{A} \mathbf{Z} + \mathbf{\mu}$. Then one writes $X \sim \mathcal N_k(\mathbf \mu, \mathbf \Sigma)$ with $\mathbf \Sigma = \mathbf A \mathbf A^T$.

Does an equivalent representation exist for a normal random matrix, i.e. a normal random matrix as a linear transformation of a standard normal random matrix?

  • $\begingroup$ There's nothing to show or prove, because a Normal random matrix simply is a Normal random vector arranged in a tabular form: the distinction is merely a matter of notation. That's what the Wikipedia article is trying to tell you at the line "The matrix normal is related to the multivariate normal distribution..." $\endgroup$
    – whuber
    Apr 4, 2020 at 22:02

1 Answer 1


A normal random matrix $\mathbf{X} \sim \mathcal{MN}_{n\times p}(\mathbf{M}, \mathbf{U}, \mathbf{V})$ is transformed as

$$\mathbf{DXC}\sim \mathcal{MN}_{r\times s}(\mathbf{DMC}, \mathbf{DUD}^T, \mathbf{C}^T\mathbf{VC})$$

for $\mathbf D, \mathbf C$ full rank.

So with the "standard normal random matrix" $\mathbf Z \sim \mathcal N_{n\times p}(0, I_n, I_p)$ one can write

$$\mathbf X = \mathbf M + \mathbf A \mathbf Z \mathbf B \sim \mathcal N_{n\times p}(\mathbf M + \mathbf A\mathbf 0\mathbf B, \mathbf A\mathbf I_n\mathbf A^T, \mathbf B \mathbf I_p\mathbf B^T) = \mathcal N_{n \times p}(\mathbf M, \underbrace{\mathbf U}_{=\mathbf A\mathbf A^T}, \underbrace{\mathbf V}_{=\mathbf B\mathbf B^T}).$$

On second read, the above was more or less stated on Wikipedia.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.