Cole Monnahan

I think that example script is outdated. It's better to try the models on the wiki, e.g., https://github.com/James-Thorson-NOAA/VAST/wiki/Simple-example Does that run for you?

I've been using V7.0.0... I thought the numerical improvements were in v8.0.0?

No I pulled the newest version of 'development' and per your instruction this morning started using v8.0.0. Should I not do that? It's running except for those two minor issues....

I'm using v8.0.0.cpp. This is paired with the most recent version of the development branch. On Thu, Mar 14, 2019 at 2:51 PM Jim Thorson wrote: > I'm confused --...

OK thanks. If V7.0.0 has the stability updates you mentioned today then I will not use v8.0.0 to simplify my life. Does it? On Thu, Mar 14, 2019 at 2:59...

I'm going to piggyback on this issue since it's still open and related. When I use `format_covariates` the output is an array of dimension `(n_x, n_t, n_p)` where n_p is...

I'm having this issue, not locally but when I do a TravisCI build on my package that depends on rstan. Is there a recommended way to setup an rstan dependency...

Nevermind, after much digging I needed to add this to my travis config file: ```` # from https://discourse.mc-stan.org/t/building-rstanarm-like-packagee-on-travis-with-c-14/6054/10? before_install: - mkdir -p ~/.R - echo "CXX14FLAGS=-O3 -mtune=native -march=native -Wno-unused-variable -Wno-unused-function...

Allan was referring to me displaying the output live, not updating it. We don't see any of the output in ss3sim so I'm probably not the right person to update...

To be slightly pedantic the linear correlations have no impact on the MCMC efficiency. They are negated by the mass matrix transformation of the posterior surface. It's good to report...