Open lentinj opened 8 months ago
I guess we can do one more, t0
can't be greater than minimum age otherwise we get negative growths. Might be worthwhile to either to set a default upper bound to 0 or set t0
as -exp(log_t0)
.
t0 can't be greater than minimum age otherwise we get negative growths
Certainly sounds sensible but it's not a solution in itself. In initialconditions we do (age - cur_step_size) - t0
, so if (like anch) you have ages 0..3 then you in effect considering age == -1
. But it's quite possible that this in itself is nonsensical.
Could the solution to this dilemma be to restrict the initial conditions to run only for ages minage + 1
and above and let the recruitment do its thing when that happens.
I seem to remember @willbutler42 doing something similar to avoid double recruitment in the first step, although not sure where now.
I seem to remember @willbutler42 doing something similar to avoid double recruitment in the first step, although not sure where now.
Will get back to you tomorrow regarding this
Whist messing with minimal modelwizard models.
Somewhat obviously in retrospect, if
t0
is equal to one of our ages, then at that age the mean length will be 0. However, this is resulting in NaN's for that age.I think the problem is
g3a_initialconditions_normalcv()
, and we should be avoid_zero()ingmean_f * cv_f
. But I've not experimented further.