[SciPy-User] Working with covariance matrices in toeplitz form?

[Sturla Molden]

"Ryan J. Kinnear" <ryan at kinnear.ca> wrote:

> Has anyone done any work on this?  Could this be a potential feature?

I am not sure, but would it possible to do this with some stride magic?
I.e. can the regularity of the Toeplitz matrix be expressed as a set of
strides that maps from a covariance matrix to a vector? In that case you
could use the NumPy function as_strided to achieve this effect.