In the example of multivariate normal distribution,

$$ \begin{bmatrix} \mathbf{x}_1 \\ \mathbf{x}_2 \end{bmatrix} \sim \mathcal{N}\left(\begin{bmatrix} \mu_1 \\ \mu_2 \end{bmatrix}, \begin{bmatrix} \Sigma_{11} & \Sigma_{12} \\ \Sigma_{21} & \Sigma_{22} \end{bmatrix}\right) $$

Then it is known that

$$ \mathbf{x}_2 \mid \mathbf{x}_1 \sim \mathcal{N}(\mathbf{m}, \mathbf{S}) $$


$$ \mathbf{m} = \mu_2 + \Sigma_{21} \Sigma_{11}^{-1} (\mathbf{x}_1 - \mu_1) $$

$$ \mathbf{S} = \Sigma_{22} - \Sigma_{21} \Sigma_{11}^{-1} \Sigma_{12} $$

Next, how can I express the conditional distribution $\mathbf{x}_2 \mid \mathbf{x}_1$ when they are multivariate cauchy distributed random variables?

$$ \begin{bmatrix} \mathbf{x}_1 \\ \mathbf{x}_2 \end{bmatrix} \sim \mathcal{CAUCHY}\left(\begin{bmatrix} l_1 \\ l_2 \end{bmatrix}, \begin{bmatrix} \Gamma_{11} & \Gamma_{12} \\ \Gamma_{21} & \Gamma_{22} \end{bmatrix}\right) $$

If it can't be expressed analytically, I also want to know how about in

$$ \begin{bmatrix} \mathbf{x}_1 \\ \mathbf{x}_2 \end{bmatrix} \sim \mathcal{CAUCHY}\left(\begin{bmatrix} \mathbf{0} \\ \mathbf{0} \end{bmatrix}, \begin{bmatrix} \Gamma_{11} & \Gamma_{12} \\ \Gamma_{21} & \Gamma_{22} \end{bmatrix}\right) $$

  • 1
    $\begingroup$ You might find my account of the bivariate case at stats.stackexchange.com/a/71303/919 to be helpful, especially at the remark "the vertical skew transformation rescales each conditional distribution by $\sqrt{1-\rho^2}$ and then recenters it by $\rho x.$" $\endgroup$
    – whuber
    Aug 2, 2021 at 0:52
  • 1
    $\begingroup$ You write down the joint density and then treat $x_1$ as a constant. What do you obtain? $\endgroup$
    – Xi'an
    Aug 2, 2021 at 6:44

1 Answer 1


I found that there is a formula of conditional multivariate t distribution.

https://arxiv.org/pdf/1402.4306.pdf https://sigfin.org/?plugin=attach&refer=021-18&openfile=SIG-FIN-021-18.pdf

Then we can get the formula of conditional multivariate cauchy distribution by setting $\nu = 1$.

In conclusion, the conditional distribution of multivariate cauchy distribution may be a multivariate t distribution.


Your Answer

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

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