Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $H$ be a positive definite Hermitian matrix. How to prove

$$\int_{\mathbb R^n}\int_{\mathbb R^n}e^{-z^*Hz}dxdy=\frac{\pi^n}{\det H},$$ where $z=x+iy$ and $dx$, $dy$ denote the integration over real $n$ dimensional region?

share|cite|improve this question
Would you mind explaining to me what is $z^*Hz$? – Américo Tavares Aug 16 '11 at 20:03
If one defines the covariance between two complex-valued random column vectors $Z\in\mathbb{C}^n$, $W\in \mathbb{C}^m$ as the $n\times m$ matrix $E((Z-E(Z))(W-E(W)^*)$ where $^*$ is the conjugate-transpose, then when $Z$ and $W$ are the same random vector, this is like the variance. If $H$ is the variance, then with a Gaussian distribution, the density would be a constant times $e^{-z^*Hz/2}$, differing from what you've got only by that factor of $2$. This is probably ultimately reducible to $\int_{-\infty}^\infty e^{-x^2/2}\,dx = \sqrt{2\pi}$. – Michael Hardy Aug 16 '11 at 20:07
@Américo: Note that $z \in \mathbb{C}^n$ and $H \in \mathbb{C}^{n \times n}$. The row vector $z^{\ast}$ is the conjugate transpose of $z$ and the product $z^{\ast}Hz$ is simply the matrix product (you can write this as the scalar product $\langle Hz, z \rangle$ with the standard Hermitian scalar product on $\mathbb{C}^n$ if you prefer). – t.b. Aug 16 '11 at 20:49
@Theo: I was unsure about the order of $H$ and the meaning you clarified of $z^*Hz$. – Américo Tavares Aug 16 '11 at 21:21
@Américo: Oh, sorry... No offense intended by further explaining and providing links. – t.b. Aug 16 '11 at 21:23
up vote 8 down vote accepted

Since $H$ is Hermitian, it admits Cholesky decomposition, i.e. $H = U^\dagger U$. Notice that $\det H = \vert \det U \vert^2$.

Then $-z^\ast H z = (U z)^\ast (U z)$. Introducing $w = U z$, and noticing that $\mathrm{d} x \mathrm{d} y = \frac{1}{\vert \det U \vert^2 } \mathrm{d}w_x \mathrm{d} w_y$ we have ( here $w_x = \mathop{Re}(U z)$ and $w_y = \mathop{Im}(U z)$):

$$ \int_{\mathbb{R}^n \times \mathbb{R}^n} \mathrm{d} x \mathrm{d} y \,\, \exp(-z^\ast H z) = \int_{\mathbb{R}^n \times \mathbb{R}^n} \frac{\mathrm{d} w_x \mathrm{d} w_y}{\vert \det U \vert^2} \,\, \exp(-w^\ast w) = \frac{\pi^n}{\det H }. $$

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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