Closed nahorp closed 2 years ago
No, this is just a change in how the summary output displays the links. The links of non-mu parameters only apply if those parameters are predicted (via a separate formula), which is now reflected in the summary output.
Hello,
I fitted 2 models, 1-2 years apart, using the
shifted_lognormal()
family argument (with default links). The earlier one is as below,This is congruent with the
?brmsfamily
help documentation (as below),However, the newer one shows identity link functions (vs. log previously) for sigma and ndt parameters. Am I missing something?
Thank you for reading :)