The Matrix Cookbook


I was browsing through the doc for the Eigen matrix package (C++ templated, which I need for automatic differentiation) and they had a pointer to the following extended (70+ page) matrix “cheat sheet”:

It contains all kinds of useful definitions and derivations, most relevantly for me, gradients of all the matrix operations (inverses, determinants, traces, eigenvalues). They even have the gradient for the multivariate normal density function (see section 8), which has a pleasingly simple form. Now if I could just get them to do the Wishart and multivariate-t, which they define, but don’t differentiate.

And how great is that pair of author names, “Petersen” and “Pedersen”? They’re reminiscent of Tintin’s Thomson and Thompson.

The authors even have a blog about The Matrix Cookbook, though it hasn’t been updated in almost a year.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: