[x] add in nu_t argument and use in salmon_sim(), like we have for epsilon; see #9
[x] make the simulations part a new function and split it out, then generate and save all the realisations (looping through m), saving results presumably in a list, to then use for fitting; motivation is to generate all simulations first, then parallelise the fitting
[x] add fit_edm = TRUE/FALSE etc. arguments and if statements, to not always fit edm and both larkins (for testing and developing).
[x] then update test, and do skip_on_ci or whatever for the larkin ones if can't get GHA working with cmdstanr (trying that above).
[x] add switch to sim_and_fit_realisations() and sim_and_fit() [have to do that first] to do R_t or R_prime_t.
[ ] need to loop through different E (so lag combinations) for EDM - currently only testing one. Probably best to add wrapper function to pbsEDM. We have pbsEDM_Evec() already, but need something more to loop through different lags. See also https://github.com/pbs-assess/pbsEDM/issues/28. And see email conversation with Luke regarding his mve implementation.
[ ] look at larkin package to try and fix build errors (might be the examples?)
sim_and_fit_realisations():nu_targument and use insalmon_sim(), like we have for epsilon; see #9fit_edm = TRUE/FALSEetc. arguments and if statements, to not always fit edm and both larkins (for testing and developing).sim_and_fit_realisations()andsim_and_fit()[have to do that first] to doR_torR_prime_t.pbsEDM_Evec()already, but need something more to loop through different lags. See also https://github.com/pbs-assess/pbsEDM/issues/28. And see email conversation with Luke regarding his mve implementation.