search

cct-datascience / organization

Initial repository for organizing group activities
1 stars 0 forks source link

Read new Wood et al. preprint #2091

Closed Aariq closed 1 month ago

Aariq commented 1 month ago

https://doi.org/10.48550/arXiv.2404.16490

Aariq commented 1 month ago

Cannot figure out how to construct nei object for arbitrary spatio-temporal data. This is the difficulty Noam Ross alluded to in his mastodon post: https://ecoevo.social/@noamross/112475869996962443

Aariq commented 1 month ago

Recreation code is said to be supplied as supplementary, but it's not published alongside the pre-print. Might have to wait till it is published to use. But definitely seems important for https://github.com/usa-npn/cales-thermal-calendars/issues/25 since there appears to be issues with over-fitting (temporal smooth and interaction smooth are both extremely wiggly and want even more knots)