gadget_project_output flexibility

[pamelajwoods]

Couple suggestions - 1) would be nice to be able to add custom fleet names (not restricted by the pre_fleet or the '.' connector and 2) would be nice to have automatic printing of total biomass and a definable reference biomass.
Thanks!

[pamelajwoods]

More thoughts for your consideration : ).

• it is necessary to add " ; -- data -- " above the data section in the main file for it to be read properly
• adding the prognosis likelihood to implement an HCR
• rename gadget_project_fleets to gadget_project_fleet and gadget_project_ref_point to gadget_project_ref_points. Is this function general enough it could be used for intermediate catches too? Or is that confusing?
• ablility to separate bloss from btrigger (unless it is there already)
• can pre_fleet be passed as a path attribute from gadget_project_fleet? (so you don't have to keep re-entering it as it).
• Can n_replicates also be passed as a path attribute? Maybe for clarity of usage it would be good to have a very simple helper function like 'gadget_project_errordimensions' where it just defines the error structure (model, implementation, recruitment) and the output structure (in our case I think it would just be setting n_replicates but could be made more flexible). Then there could be one last helper function to compile results that takes the n_replicates into account with the data_frame.
• can harvest_rate in project_advice can be split among fleets? (also consider renaming harvest_rate)
• option to print log in gadget_evaluate would be useful

[bthe]

adding the prognosis likelihood to implement an HCR

This commit https://github.com/Hafro/rgadget/tree/eb7abc0563b1a55d33e8620df62fe72404d949f9 should address that. Here is an example that would work in the ling case (but any other stock should work):

library(Rgadget)

fit <- gadget.fit(wgts = 'WGTS_v2')

ref.cm <- 75


gadget_project_time(variant_dir = 'PRE') %>% 
  gadget_project_stocks(imm.file = 'lingimm',mat.file = 'lingmat') %>% 
  gadget_project_fleets(pre_fleet = 'lln',type = 'prognosis') %>% 
  gadget_project_fleets(pre_fleet = 'bmt',type = 'prognosis') %>% 
  gadget_project_fleets(pre_fleet = 'gil',type = 'prognosis') %>% 
  gadget_project_prognosis_likelihood(stocks = c('lingimm','lingmat'),
                                      pre_fleets = c('lln','bmt','gil'),
                                      biocoeffs = c(-100,ref.cm,1,-100,ref.cm,1),## biocoeffs, this doesn't work atm 
                                      ## only want the mature biomass as the trigger biomass
                                      triggercoeffs=c(0,0,0,0,0,1), 
                                      triggervalues=c(0, sprintf('#%s.btrigger','ling')), ## immature and mature triggers
                                      harvestrates = sprintf(c('0#%s.hr.low','0.2#%s.hr.high'),'ling'), ## linear between hr.low and hr.high when < btrigger
                                      quotasteps = c(3, 4, 5, 6),
                                      ## assume equal fishing intensity for now
                                      quotaproportions = c(0.25, 0.25, 0.25, 0.25)) %>% 
  gadget_evaluate(params.out = paste(attr(.,'variant_dir'),'params.pre',sep='/'),
                  params.in = 'WGTS_v2/params.final') %>% 
  gadget_project_advice(harvest_rate = 1, 
                        pre_fleet = 'lln.bmt.gil',
                        prefix = 'asserr') %>% 
  ## if you want to add implementation error:
  #gadget_project_advice(harvest_rate = 1, 
  #                      pre_fleet = 'lln.bmt.gil',
  #                      prefix = 'implerr') %>% 
  gadget_project_ref_point(tibble(ling.hr.high=c(0.1,0.2),ling.btrigger = c(10e6,10e6)))