Check all functions work with cmdstanr backend about bayesnec

[beckyfisher]

All exported functions have been checked with the cmdstanr backend, and appear to be working. The only difference between cmdstanr and rstan appears to be related to the fact that a brmsfit model fitted using cmdstanr does not retain initial values. For cmdstanr modfit$fit$fit@inits returns and empty list, whereas a model fitted using rstan will return a list of the initial values. Within bayesnec this is only being used in function amend, and will potentially reduce reproducibility when using cmdstanr. As there is no apparent work around this issue, this potential lack of reproducibility has now been added to the text of the article.