search

pbs-assess / sdmTMB

:earth_americas: An R package for spatial and spatiotemporal GLMMs with TMB
https://pbs-assess.github.io/sdmTMB/
187 stars 27 forks source link

Add unit testing against VAST #85

Open seananderson opened 2 years ago

seananderson commented 2 years ago

Will have to be outside the usual testthat folder (for now in a folder in inst?) or skipped in all but local runs, which might be ideal assuming it doesn't upset CRAN to have packages loaded in non-run tests that aren't in 'suggests'.

See https://github.com/seananderson/sdmTMB-paper/blob/main/doc/appendix-pcod-VAST.Rmd (delta-Gamma) https://github.com/seananderson/sdmTMB-paper/blob/main/doc/appendix-pcod-VAST-tweedie.Rmd (Tweedie) for a starting point.

Test Tweedie, delta-Gamma, delta-lognormal, anything else easily matching? Test parameter estimates, predictions, COG, and index values Test bias correction Make sure to use exact same mesh as is done above Test various spatiotemporal structures (IID, AR1, RW)

jindivero commented 2 years ago

Just wanted to let you know I'm working on this! https://github.com/pbs-assess/sdmTMB/blob/bc0f43658a8daf41ee0cd174297e16e18046d265/julia-tests/test-1-vast.R

seananderson commented 2 years ago

There has been a bunch of work over here on this: https://github.com/kellijohnson-NOAA/indexwc/blob/main/sandbox/run_index_comparison.R

We have everything matching exactly so far. It's complicated enough, we may want a separate repo that runs a bunch of comparisons. Or a folder in this repo that is ignored when building the package.

kellijohnson-NOAA commented 2 years ago

My suggestion would be to create a new repository that runs comparisons on a schedule. That way it can have its own README file etc.