ben18785

Thanks @chonlei -- I really like this! I think this makes setting hyperparameters much neater. My preference would be to go for proposal 1 since it's safer.

Haha. I have no idea why I wrote @chonlei there! Sorry @MichaelClerx (I think I must have written "@C" then chose "h" rather than "l).

It does! Thanks Ferguz On Tue, Mar 16, 2021 at 10:11 AM Fergus Cooper ***@***.***> wrote: > @MichaelClerx @ben18785 > does this look about right? > 1225905 > > >...

Also, the `sigma_inv` calculation in the method in `GaussianLogPDF` can be cached, which can be addressed in same PR, I reckon.

Also, add note about result being from chapter 2 of matrix cookbook (plus maybe https://stats.stackexchange.com/questions/27436/how-to-take-derivative-of-multivariate-normal-density).

Do you have a reference for those optimisation methods? I’m trying to get my head around them all... Or a book recommendation for optimisation in general? > On 6 Apr...

Cheers for all these -- just going through these references now!

Sounds like a plan. I'd keep the diagnostics outside of pints.mcmc as can imagine having diagnostics for other methods later on (and it's annoying-er to do import pints.mcmc.diagnostics). On Tue,...

So, I think we still want our existing diagnostics methods but that any new ones come from Arviz... On Tue, Feb 9, 2021 at 3:47 PM David Augustin wrote: >...

Hi David, happy for you to make those changes if you like. Perhaps it'd just be better though to move wholesale to Arviz? There are a host of changes to...