Computes the probability distribution of fish lengths across different
age classes. AD-compatible: uses RTMB::pnorm and ADoverload so that
gradients propagate if growth parameters (mu_a, sd_a) are estimated.
Usage
get_pla(len_lower, len_upper, mu_a, sd_a)
Arguments
- len_lower
Numeric vector of lower bounds of length bins (length L). Data only.
- len_upper
Numeric vector of upper bounds of length bins (length L). Data only.
- mu_a
Numeric vector of mean length at age (length A). May be AD.
- sd_a
Numeric vector of SD of length at age (length A). May be AD.
Value
Matrix of dimensions L x A where columns sum to 1.
Details
Bins are assumed contiguous and non-overlapping, i.e.
len_upper[i] == len_lower[i + 1] for all i. This allows a
single vectorized pnorm call per age on the n_len + 1 unique
bin edges rather than two calls on the lower and upper bounds separately.
