Wish list

[PaulRegular]

• Tailor to other species and assess sampling protocol
• Add a time component to assess trade-offs / efficiency (may require simulations of multiple species; determine the priorities of the survey)
• Assess the effects of missing strata
• Assess the effects of variable trawl area
• Assess the effects of random vs. stratified age sampling (initial tests indicates better performance of length-stratified approach)
• Assess the effects of applying finer-scale age-length keys (may decrease bias)
• Use this as a platform to test spatial models (e.g. ogmap, geostatistical models)
• Use as an operating model for an MSE
• Add error to age reading
• Add stock-recruit relationship (add sex, maturity and weight relationships)
• Consider making growth sex-specific
• Consider simulating M and Z separately
• Add sampling of other info (e.g. sex, maturity)
• Explore alternate approaches to introducing sampling error (e.g. instead of Binomial, try Poisson or negative binomial); look to the comparative fishing literature for inspiration here
• Apply sub-sampling to actual survey data and compare to sub-sampling of simulated data as a means to further ground-truth the simulation approach
• Revise set allocation to better reflect within-strata allocations applied in the DFO survey
• Use the units package
• Figure out a computationally efficient way to simulate population at length (currently limited to population at age) and compare strat estimates at length to the true values