hindcast_jabba: hindcast_jabba()

In jabbamodel/JABBA: Just Another Bayesian Biomass Assessment

hindcast_jabba()

Description

Wrapper to fit retrospectives

Usage

hindcast_jabba(
  jbinput,
  fit,
  ni = NULL,
  nt = NULL,
  nb = NULL,
  nc = NULL,
  quickmcmc = TRUE,
  init.values = TRUE,
  peels = 1:5,
  verbose = FALSE
)

Arguments

jbinput	object from build_jabba()
ni	number of iterations
nt	thinning interval of saved iterations
nb	burn-in
nc	number of mcmc chains
quickmcmc	Reduces MCMC iters for hindcasting reference run Initial values
init.values	if TRUE init values from fit are used
peels	sequence of retrospective peels default 1:5
verbose	if FALSE run silent
jbfit	fitted model from fit_jabba MCMC settings

Value

hc containing estimates of key joint results from all hindcast run

Examples

data(iccat)
whm = iccat$whm
# ICCAT white marlin setup
jb = build_jabba(catch=whm$catch,cpue=whm$cpue,se=whm$se,assessment="WHM",scenario = "BaseCase",model.type = "Pella",r.prior = c(0.181,0.18),BmsyK = 0.39,igamma = c(0.001,0.001))
fit = fit_jabba(jb,quickmcmc=TRUE,verbose=TRUE)
hc = hindcast_jabba(jbinput=jb,fit=fit,peels=1:5)
jbplot_retro(hc)
jbplot_hcxval(hc,index=c(8,11))
hc.ar1 = jbhcxval(hc,AR1=TRUE) # do hindcasting with AR1
jbplot_hcxval(hc.ar1,index=c(8,11))

jabbamodel/JABBA documentation built on Nov. 2, 2024, 12:50 p.m.