Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Alternative scheduling #236

Open
richfitz opened this issue May 26, 2021 · 0 comments
Open

Alternative scheduling #236

richfitz opened this issue May 26, 2021 · 0 comments

Comments

@richfitz
Copy link
Member

Fully exploring this will require changes in mcstate too.

At the moment we do a static schedule so that for 4 threads and 16 particles we'd send them to threads

1111222233334444

however this typically comes after a sorting of likelihood so it's very likely that group 1 has similar parameters to each other and group 4 has similar parameters to each other, but that group 1 and 4 might have quite different parameters. This minimises within-group variance in parameters (and likely time taken) but maximises between-group variance. Not ideal!

So if we did this as

1234123412341234

We'd reduce the between-group variance.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant