search

pfmc-assessments / lingcod

2 stars 2 forks source link

STAR panel request #17 #160

Closed kellijohnson-NOAA closed 3 years ago

kellijohnson-NOAA commented 3 years ago

Request

Provide a retrospective analysis that goes back 5 years for the northern and southern models. Report the Mohn’s rho values –(Woods Hole and Alaskan/Hurtado-Ferro).

Rationale:

There are significant scaling issues associated with the retrospectives that merit a closer investigation

kellijohnson-NOAA commented 3 years ago

Matt C. asked for uncertainty in all retros in an email, but I think we will only have time to provide uncertainty wrt the base.

See r4ss pngtime branch for a more informed way to do retrospectives that accounts for rec devs and blocks. Might have bugs :(

kellijohnson-NOAA commented 3 years ago

Code for Mohn's rho that needs more automation eventually.

mydir <- "models/2021.s.017.001_triextrasdreweight_retro"
retroModels <- r4ss::SSgetoutput(
  dirvec = file.path(mydir, "retro", paste("retro", -1:-5, sep = ""))
)
retroSummary <- r4ss::SSsummarize(c(list(mod.2021.n.023.001), retroModels))
endyrvec <- retroSummary[["endyrs"]] + 0:-5

r4ss::SSmohnsrho(retroSummary,
           endyrvec = endyrvec,
           startyr = 2015,
           verbose = TRUE
)