search

thefaylab / hydra_sim

Code-base for the hydra multispecies simulation model
2 stars 0 forks source link

Predicted values for survey observations #4

Open thefaylab opened 3 years ago

thefaylab commented 3 years ago

While trying to find the predicted values for survey observations, I notice that there is currently no survey selectivity. (or rather, selectivity is 1.0 for all size classes) I also notice the 'estimated survey biomass' quantitites are using average biomass over the year, with a time-invariant area and species specific catchability.

Is there aneed/want to move to implementing size-specific selectivity, with specific accounting for the timing of survey during the year?

Also @andybeet for insight in case I am reading things wrong.

sgaichas commented 3 years ago

The survey portion of hydra was really just a placeholder. So I think we will want to implement survey size selectivity and timing for the estimation model. We do have spring and fall surveys and the simulated data also has within-year timing.

I'm not sure this is necessary for the simulation model though. This may be another case where we will want a flag to skip portions of the code for simulation or estimation mode.

gavinfay commented 2 years ago

Size selectivity is now implemented in the predicted values for the surveys that get compared to the data. (e.g. see calculation of predicted_survey_index) The selectivity is currently hard-coded at 1 for all species and sizes, but the array is there in the calculations.

Expected values for the surveys are currently the average over the time steps within the year, rather than at specific times of the year. I also (simply) made the assumption that surveys cover all areas (OK for our initial examples as there's only 1 area). Will need to think about how to specify timing of surveys within year because the number of time steps is dynamic based on other parameters. I think the average is OK for now.