run_proj: run_proj

In nikolaifish/bamExtras: Functions and Data in Support of Stock Assessments with the Beaufort Assessment Model (BAM)

run_proj

Description

Project BAM forward in time to conduct deterministic or stochastic projections.

Usage

run_proj(
  rdat = NULL,
  run_bam_args = NULL,
  bam2r_args = NULL,
  stochastic = FALSE,
  pstar = NULL,
  styr_proj = NULL,
  nyb_rcn = list(L = 3, U = 3, R = 5, comp = 3),
  nyp_cur = 3,
  nyp = 5,
  nm_yr_p = c("endyr", "endyr_rec_dev", "endyr_rec_phase2", "endyr_rec_spr", "styr_regs",
    "endyr_proj"),
  F_cur = NULL,
  F_proj = NULL,
  L_cur = NULL,
  ages = NULL,
  N_styr_proj = NULL,
  S_styr_proj = NULL,
  sel_L = NULL,
  sel_D = NULL,
  sel_tot = NULL,
  sel_F_flt = NULL,
  wgt_mt = NULL,
  wgt_L_klb = NULL,
  wgt_D_klb = NULL,
  wgt_F_flt_klb = NULL,
  len_F_flt_mm = NULL,
  reprod = NULL,
  M = NULL,
  spawn_time = NULL,
  SR_par = NULL,
  SR_method = "BH",
  age_error = NULL,
  project_bam = FALSE,
  ia_args = list(U_nm = NULL),
  obs_error = list(cv_U_sc = 0),
  plot = FALSE,
  nm_comp_sfx = c(""),
  t_series_na_vals = c(-9999, -99999, -999999),
  key_lenprob = c(all = "^len.cv")
)

Arguments

rdat	BAM output rdat (list) object read in with dget().
run_bam_args	list of arguments passed internally to run_bam if any are provided
bam2r_args	list of arguments passed internally to bam2r if any are provided
stochastic	logical. Is this a stochastic projection? Mostly runs the same code either way, but stochastic projections add normal errors to recruitment. Stochastic setting is intended to be run with rdat input from MCBE.
pstar	Value to use for applying pstar scaling of Fmsy. If set to NULL, it is not used and pstar code will not be run.
styr_proj	start year (i.e. first calendar year) of projection
nyb_rcn	number of years in calculations that include recent landings (L) or discards and their cvs, index cvs (U), recruitment (R), or numbers of samples in age or length compositions (comp). Fleets without L or D within the last nyb_rcn years of the base model will not be projected forward. named list
nyp_cur	number of years to maintain current conditions before implementing management
nyp	number of years of projection. Must be >nyp_cur
nm_yr_p	names of init objects to project forward by nyp
F_cur	current fully selected fishing mortality rate
F_proj	fully selected fishing mortality rate
L_cur	current level of landings
ages	ages. numeric vector
N_styr_proj	abundance at age in styr_proj. numeric vector
S_styr_proj	spawning stock size at age in styr_proj. Often biomass in mt, sometimes eggs in n. numeric vector
sel_L	selectivity at age to compute landings. numeric vector
sel_D	selectivity at age to compute dead discards. numeric vector
sel_tot	selectivity at age to compute Z (includes landings and discards). numeric vector
sel_F_flt	list of selectivity vectors (by age) or matrices (by year,age) for each fleet associated with landings or discards. numeric vector or matrix
wgt_mt	weight at age of population in metric tons (mt). numeric vector
wgt_L_klb	weight at age of landings in thousand pounds (klb). numeric vector
wgt_D_klb	weight at age of discards in thousand pounds (klb). numeric vector
wgt_F_flt_klb	list of weight vectors (klb by age) or matrices (by year,age) for each fleet associated with landings or discards. numeric vector or matrix
len_F_flt_mm	list of length vectors (mm by age) or matrices (by year,age) for each fleet associated with landings or discards. numeric vector or matrix
reprod	reproductive contribution at age to SSB. numeric vector
M	natural mortality rate, at age
spawn_time	time of year for peak spawning
SR_par	list of parameters associated with the stock-recruit (SR) relationship. Currently only works with a Beverton-Holt SR relationship for which the parameters are: h = steepness of spawner-recruit function, R0 = virgin recruitment of spawner-recruit function, Phi0 = virgin spawners per recruit, biascorr = bias correction
SR_method	Spawner-recruit function BH = Beverton-Holt, R0 = model estimated mean recruitment (R0), GM = constant geometric mean of nyb_rcn$R recent years of t.series$recruits. By default, run_proj will try to use BH, or R0, but if it can't find all the parameters, it will resort to GM. Note GM method has not been reviewed and should be used with caution.
age_error	age error matrix to use for the projection years. By default, the function uses the same age_error matrix from the base model
project_bam	logical. After the projection is run, should the function build a new set of projected bam files?
ia_args	list of arguments to pass to interim adjustment function. See Details section below.
obs_error	list of optional arguments for incorporating observation error into projections. See Details section below.
plot	logical. Produce plots of extended base model including projected data
nm_comp_sfx	Character vector. Possible suffix added to fleet abbreviation in comps (e.g. c("a","b") in SEDAR 82 gray triggerfish). Used to match comps with catch when projecting comp data when project_bam=TRUE
t_series_na_vals	Numeric vector of values to consider as NA in t.series object.
key_lenprob	Optional list for specifying cv values for length-at-age, for each set of length compositions. (e.g. key_lenprob = c("D.rHD"="len.cv.L", "L.rGN"="len.cv.L")). Use keyword 'all', as in default to set all values to a particular value in parm.cons.

Details

ia_args:

• U_nmThe fleet abbreviation for an index of abundance that you want to use to adjust F (e.g. sTV, sCT). If set to NULL, no interim adjustment will be used.

• typeMethod used for computing F adjustment. "Uprop" computes the mean U from recent years, divided by U from reference year. The reference year is the terminal year of the last full stock assessment. The set of recent years is computed as the current projection year plus yr_U_lag. By default, type = "Uprop"

• yr_U_lagA vector of negative integers used to compute the set of years used for computing an interim adjustment from the reference index (comma separated example values: -1, -1:-3, -2:4). By default, yr_U_lag = -(2:4).

• yr_ia_byFrequency of interim adjustments, in year units. Used to determine the years between stock assessments when interim adjustments will be computed and applied to F. The interim adjustment years are computed as yrlim_p <- endyr + c(1,nyp); yr_ia <- seq(yrlim_p[1], yrlim_p[2], by = yr_ia_by). By default, yr_ia_by = 5

• include_yrlimLogical. Should interim analysis be conducted in the first and last years of the projection period? By default, include_yrlim_p = FALSE

obs_error:

• cv_U_scNumber to multiply by cv values of abundance indices in the projection period.

Author(s)

Kyle Shertzer, Erik Williams, and Nikolai Klibansky

Examples

## Not run: 
# Run deterministic projection with defaults
proj_VeSn <- run_proj(rdat_VermilionSnapper,plot=TRUE)

# Run deterministic projection with defaults
# and project bam inputs, by adding the bam files with bam2r_args
proj_VeSn <- run_proj(rdat_VermilionSnapper,
project_bam=TRUE,
bam2r_args = list(
dat_obj=dat_VermilionSnapper,
tpl_obj=tpl_VermilionSnapper,
cxx_obj=cxx_VermilionSnapper
))
# The projections work for most of the assessments
proj_BlSb <- run_proj(rdat_BlackSeaBass,plot=TRUE)
proj_GaGr <- run_proj(rdat_GagGrouper,plot=TRUE)
proj_RdGr <- run_proj(rdat_RedGrouper,plot=TRUE)
proj_RdSn <- run_proj(rdat_RedSnapper,plot=TRUE)
proj_VeSn <- run_proj(rdat_VermilionSnapper,plot=TRUE)
# Some need additional arguments to run correctly
proj_GrTr <- run_proj(rdat_GrayTriggerfish,
nm_comp_sfx = c("a","b"),
plot=TRUE,
key_lenprob = c("D.rHDs"="len.cv.L", "L.rGNs"="len.cv.L")
)

## End(Not run)

nikolaifish/bamExtras documentation built on April 17, 2025, 9:44 p.m.