Debugging Tricks

Jump to bottom

Grant Adams edited this page Feb 7, 2021 · 2 revisions

Gradient is of length 1 or similar error

Population is crashing a n-at-age < 0, or dividing by zero

The model is not converging:

Check mod$identified to see what parameters are having difficulty estimating if this printed
Change initial values of the "bad" parameters

Recreating another model

Step 1 - input n-at-age into fixed_natage from original model
Step 2 - input selectivity-at-age into emp_sel and set selectivity