Package 'FLRef'

Description

ABItgt() Computes ABI for target F, e.g. ABImsy (Griffith et al. 2023)

Usage

ABItgt(stock, ftgt = NULL, thresh = 0.9, ...)

Arguments

Value

*FLQuant*

Examples

data(ple4)
ABImsy = ABItgt(ple4,ftgt=0.22,thresh=0.9)
plot(ABImsy)+ylim(0,2)+
 geom_hline(yintercept = c(0.8,1),col=c(2,1),linetype=c(2,1))+ylab(expression(ABI[MSY]))

Description

Combines numbers-at-age with an inverse age-length key and conditions the result on length to obtain age proportions within each length class.

Usage

ALK(N_a, iALK)

Arguments

Value

An 'FLPar' containing age proportions by length class.

Examples

## Not run: 
alk <- ALK(N_a = c(100, 80, 50, 20), iALK = ialk)

## End(Not run)

Description

Generates multinomial age samples within length classes from annual age-length keys. The number sampled in each length class is limited by 'nbin' and by the corresponding length-frequency sample.

Usage

alk.sample(lfds, alks, nbin = 20, n.sample = 1)

Arguments

Value

An 'FLPars' object containing sampled annual age-length keys.

Examples

## Not run: 
sampled_alks <- alk.sample(lfds, alks, nbin = 20)

## End(Not run)

Description

Combines annual numbers-at-age with an inverse age-length key and returns age proportions within each length class for every year and iteration.

Usage

ALKs(object, iALK)

Arguments

Value

An 'FLPars' object with one age-length key per year.

Examples

## Not run: 
alks <- ALKs(stock.n(stk), ialk)

## End(Not run)

Description

Applies annual age-length keys to length-frequency data to estimate numbers-at-age by year and iteration.

Usage

applyALK(lfds, alks)

Arguments

Value

An 'FLQuant' containing estimated numbers-at-age.

Examples

## Not run: 
numbers_at_age <- applyALK(lfds, alks)

## End(Not run)

Description

applyAR() Function to apply ICES Advice rule for F < MSY Btrigger

Usage

applyAR(b, btrigger, fmsy, bmin = 0, fmin = 0.001)

Arguments

Value

F advice

Description

Converts equilibrium reference-point information from an FLBRP object into approximate Pella-Tomlinson surplus production model parameters. The function derives the shape parameter m from the equilibrium ratio Bmsy / B0, and derives the intrinsic productivity parameter r from the harvest rate at MSY.

Usage

asem2spm(
  object,
  quant = c("vb", "ssb"),
  fmsy = NULL,
  rel = FALSE,
  spcurve = FALSE
)

asem2spm(
  object,
  quant = c("vb", "ssb"),
  fmsy = NULL,
  rel = FALSE,
  spcurve = FALSE
)

Arguments

Details

The biomass quantity used for the approximation can be either vulnerable biomass, quant = "vb", or spawning-stock biomass, quant = "ssb". Optionally, the function can also return the equilibrium surplus production curve implied by the supplied FLBRP object.

The function approximates the age-structured equilibrium yield curve with a Pella-Tomlinson production function:

$P(B) = B \frac{r}{m - 1} \left[ 1 - \left(\frac{B}{B_0}\right)^{m - 1} \right]$

The shape parameter m is solved numerically from the relationship:

$\frac{B_{MSY}}{B_0} = m^{-1 / (m - 1)}$

and r is calculated from:

$r = H_{MSY} \frac{m - 1}{1 - 1 / m}$

where Hmsy = MSY / Bmsy. For quant = "vb", Bmsy and B0 are based on vulnerable biomass calculated by vbbrp(). For quant = "ssb", they are based on ssb().

This approximation is useful for deriving SPM-compatible prior means for r and m, or for computing SPM-style process deviations from age-structured stochastic projections.

Value

prior means for r and m *FLPar*

If spcurve = FALSE, an FLPar object with approximate surplus production parameters:

r

Intrinsic productivity parameter of the Pella-Tomlinson surplus production model.

m

Pella-Tomlinson shape parameter.

BmsyK

Equilibrium biomass ratio Bmsy / B0.

Hmsy

Harvest rate at MSY, calculated as MSY / Bmsy.

MSY

Maximum sustainable yield.

Bmsy

Biomass at MSY for the selected biomass metric.

B0

Unfished biomass for the selected biomass metric.

Fmsy

Fishing mortality at MSY.

If spcurve = TRUE, a list is returned with:

pars

The FLPar object described above.

curve

A data frame with columns Biomass, SP, and quant. If rel = TRUE, Biomass is scaled by B0 and SP by MSY.

Examples

data(ple4)
ple4 <- merge_catch(ple4)
sr <- srrTMB(as.FLSR(ple4,model=bevholt),spr0=mean(spr0y(ple4)))
brp = brp(FLBRP(ple4,sr))
asem2spm(brp)
plot_pf(brp)
plot_pf(brp,rel=TRUE)
## Not run: 
data(ple4)

ple4 <- merge_catch(ple4)

sr <- srrTMB(
  as.FLSR(ple4, model = bevholt),
  spr0 = mean(spr0y(ple4))
)

brp <- brp(FLBRP(ple4, sr))

# Approximate SPM parameters using vulnerable biomass
pars_vb <- asem2spm(brp, quant = "vb")
pars_vb

# Approximate SPM parameters using spawning biomass
pars_ssb <- asem2spm(brp, quant = "ssb")
pars_ssb

# Return the surplus production curve in absolute units
sp <- asem2spm(brp, quant = "vb", spcurve = TRUE)
head(sp$curve)

# Return the surplus production curve in relative units
sp_rel <- asem2spm(brp, quant = "vb", spcurve = TRUE, rel = TRUE)
head(sp_rel$curve)

plot_pf(brp)
plot_pf(brp, rel = TRUE)

## End(Not run)

Description

generates FLIndexBiomass with random observation error from an FLStock

Usage

bioidx.sim(object, sel = catch.sel(object), sigma = 0.2, q = 0.001, rho = 0)

Arguments

Value

FLIndexBiomass

Examples

data(ple4)
sel = newselex(catch.sel(ple4),FLPar(S50=1.5,S95=2.1,Smax=4.5,Dcv=1,Dmin=0.1))
ggplot(sel)+geom_line(aes(age,data))+ylab("Selectivity")+xlab("Age")
object = propagate(ple4,10)
sel = newselex(catch.sel(object),FLPar(S50=2.5,S95=3.2,Smax=3.5,Dcv=0.6,Dmin=0.2))
idx = bioidx.sim(object,sel=sel,q=0.0001)
# Checks
ggplot(idx@sel.pattern)+geom_line(aes(age,data))+ylab("Selectivity")+xlab("Age")
ggplot(idx@index)+geom_line(aes(year,data,col=ac(iter)))+theme(legend.position = "none")+ylab("Index")

Description

Uses a bisection algorithm to find the value of a forecast control quantity that gives a specified target value for a performance statistic. Typical use cases include finding the constant fbar that gives a chosen risk level, such as $P(SSB < B_{lim}) = 0.05$, over a projection period.

Usage

bisect(
  stock,
  sr,
  deviances = rec(stock) %=% 1,
  metrics,
  refpts,
  statistic,
  years,
  pyears = years,
  tune,
  prob,
  tol = 0.01,
  maxit = 15,
  verbose = TRUE
)

Arguments

Details

The function evaluates the forecast at the lower and upper bounds supplied in tune. If the requested probability lies between the two outcomes, the interval is repeatedly bisected until the performance statistic is within tol of prob, or until maxit iterations are reached.

The bisection search requires that the target objective is bracketed by the two values supplied in tune. In practice, this means that:

$\left[g(x_{min}) - p\right] \left[g(x_{max}) - p\right] < 0$

where $g(x)$ is the performance statistic produced by forecasting with control value $x$, and $p$ is the target value supplied by prob. If both bounds give performance statistics on the same side of the target, the function cannot identify a unique bracketed solution.

The function is most useful for risk-based advice calculations, such as finding the fishing mortality that gives a pre-specified probability of falling below Blim over a future evaluation period.

Value

If a solution is found, an FLStock object returned by fwd() or ffwd() for the tuned control value. If the requested target is already achieved at the lower or upper bound, the corresponding forecast object is returned. If the supplied range does not bracket the target statistic, a list with elements min and max is returned, containing the forecasts at the lower and upper bounds.

References

Burden, R. L. and Faires, J. D. (1985). Numerical Analysis. 3rd edition. PWS Publishers. Section 2.1: The Bisection Algorithm.

Examples

## Not run: 
data(ple4)

# Extend and propagate stock
stock <- propagate(stf(ple4, end = 2118), 100)

# Define a Beverton-Holt stock-recruitment relationship
srr <- predictModel(
  model = rec ~ a * ssb * exp(-b * ssb),
  params = FLPar(a = 5.20, b = 1.65e-6)
)

# Generate autocorrelated recruitment deviances
devs <- ar1rlnorm(
  rho = 0.4,
  years = 2018:2118,
  iters = 100,
  meanlog = 0,
  sdlog = 0.5
)

# Define an ICES-style risk statistic
statistic <- list(
  FP05 = list(
    ~ yearMeans((SB / SBlim) < 1),
    name = "P.05",
    desc = "ICES P.05"
  )
)

# Find fbar that gives P(SSB < Blim) = 0.05
fp05fwd <- bisect(
  stock = stock,
  sr = srr,
  deviances = devs,
  metrics = list(SB = ssb),
  refpts = FLPar(SBlim = 150000),
  statistic = statistic,
  years = 2018:2118,
  pyears = 2069:2118,
  tune = list(fbar = c(0.1, 1.0)),
  prob = 0.05
)

## End(Not run)

Description

function to assign B[y+1] to B[y]. Warning correlation structure of B[y+1] and F[y] is meaningless

Usage

blag(mvn, verbose = TRUE)

Arguments

Value

output list of quant posteriors and mle's

Author(s)

Henning Winker (GFCM)

Description

Constructs, for each gear of a multigear operating model, the selectivity-at-length, the ALK-integrated selectivity-at-age, the capture-conditioned inverse age-length key, and (optionally) the Fbar-standardised selectivity and the gear's share of total F-at-age.

Usage

build_gear(
  gear_cfg,
  lhpar,
  age,
  amin = min(age),
  amax = max(age),
  lmin = 5,
  lmax_mult = 1.2,
  bin = 1,
  cv = 0.1,
  timing = 0,
  fbar_range = NULL,
  scale = TRUE
)

Arguments

Details

With $P(l \mid a,t)$ the biological inverse ALK at 'timing' $t$:

$s_{g,a}=\sum_l P(l\mid a,t)\,s_g(l),\qquad P(l\mid a,g)=\frac{P(l\mid a,t)\,s_g(l)}{s_{g,a}}.$

F partition. With $\mathcal A$ the Fbar ages and $w_g$ = 'f_mult', gear Fbar trajectories in the OM are $\bar F_{g,y}=w_g \bar F_y$. Gear F-at-age and the share of total F are then

$F_{g,a,y}=w_g\tilde s_{g,a}\bar F_y,\qquad \phi_{g,a}=\frac{w_g\tilde s_{g,a}}{\sum_h w_h\tilde s_{h,a}},$

which does not depend on year or iteration: as long as all gears follow proportional Fbar trajectories, the F-share (and hence the gear split of catch-at-age) is a property of the gear configuration and belongs here. If gear trajectories diverge, compute time-varying shares with [f_share_gear()] instead.

All quantities live on one length grid, [len_grid()] (= [iALK()]'s grid, last bin a plus group); selectivity is evaluated at its bin mid-points with [sel_logistic_len()], [sel_normal_len()] or [sel_dnormal_len()]. 'amin'/'amax' are kept for backward compatibility and are unused.

Value

A named list, one element per gear, each a list with

name

Gear name.

type

Selectivity type.

sel_len

'FLQuant' selectivity-at-length $s_g(l)$ on the reconciled length grid (lower bin limits as dimnames).

sel_a

'FLQuant' ALK-integrated selectivity-at-age $s_{g,a}$, max-scaled if 'scale = TRUE'.

sel_a_fbar

'FLQuant' $\tilde s_{g,a}=s_{g,a}/\bar s_{g,\mathcal A}$, i.e. selectivity scaled to mean one over the Fbar ages, so that $F_{g,a,y}=\tilde s_{g,a}\,\bar F_{g,y}$. 'NULL' without 'fbar_range'.

f_share

'FLQuant' static share of total F-at-age taken by this gear, $\phi_{g,a}$ (see Details). Sums to one over gears at every age with non-zero F. 'NULL' without 'fbar_range'.

len_bins

Data frame of lower, upper and mid lengths.

condALK

Matrix $P(l \mid a, g)$, ages in rows.

f_mult, ess_age, ess_len

As supplied in 'gear_cfg'.

obs, n_seasons

Sampling design (defaults '"fishery"', '1').

lhpar

'lhpar', used as the default 'params' of [lfd.sim()].

lw

Length-weight parameters 'c(a, b)' taken from 'lhpar' ('NULL' if 'lhpar' has no 'a'/'b').

wt_len

Named vector of weight at length, $W(l)=a\,\bar l^{\,b}$ at bin mid-points (same units as 'a'/'b'), used by [lfd_sop()] to raise length samples to catch weight.

grid

List of 'lmin', 'lmax_mult', 'bin', 'cv', 'timing', 'fbar_range' used to build the gear. Downstream helpers ([plot_condALK_bias()], [lfd.sim()] when rebuilding the ALK) should default to these instead of repeating the arguments.

See Also

[joint_sel()], [f_age_gear()], [catch_n_gear()], [lfd.sim()]

Examples

## Not run: 
gear_cfg <- list(
  Trawl   = list(type = "dnormal", Lpeak = 23.4, sd_left = 3.2,
                 sd_right = 11.2, f_mult = 0.9, ess_age = 50, ess_len = 150),
  Gillnet = list(type = "logistic", L50 = 29.7, L95 = 34.1,
                 f_mult = 0.1, ess_age = 30, ess_len = 100)
)
om_gears <- build_gear(gear_cfg, lhpar = lhpars, age = 0:20,
                       timing = 0.5, fbar_range = c(3, 19))
gear_par(om_gears, "f_mult")
joint_sel(om_gears)

## End(Not run)

Description

Generate catch.n with lognormal annual and multinomial age composition observation error

Usage

ca.sim(
  object,
  ess = 200,
  u = NULL,
  what = c("catch", "landings", "discards")[1],
  rescale = FALSE
)

Arguments

Value

'FLQuant', same dimensions as 'object', with sampled catch-at-age (or rescaled proportions x true total, if 'rescale=TRUE').

Examples

## Not run: 
data(ple4)
object <- propagate(catch.n(ple4), 10)

## original behaviour, unchanged
ca1 <- ca.sim(object, ess = 200)

## CRN-safe: same u -> same sample, even if called twice
u <- array(runif(200 * dim(object)[2] * 10),
           dim = c(200, dim(object)[2], 10))
ca2 <- ca.sim(object, ess = 200, u = u)
ca3 <- ca.sim(object, ess = 200, u = u)
stopifnot(identical(ca2, ca3))

## End(Not run)

Description

Sums gear catches over a window of years and normalises across gears:

$\pi_g=\frac{\sum_{y\in\mathcal Y}C_{g,y}} {\sum_h\sum_{y\in\mathcal Y}C_{h,y}}.$

Supplying the realised catch split over the LFD window (rather than the nominal 'f_mult') mirrors what an assessment would actually be given, and reflects that a gear's catch share differs from its Fbar share whenever gears select different ages.

Usage

catch_by_gear(
  catch_n,
  wt = NULL,
  years = NULL,
  iter = NULL,
  type = c("weight", "numbers")
)

Arguments

Value

Named numeric vector or matrix of catch shares (rows sum to one).

Examples

## Not run: 
catch_by_gear(catch_n, om_hist, years = lfd_years, iter = 1)
cbg <- catch_by_gear(catch_n, om_hist, years = lfd_years)  # all iters
fiticc(lfd_i, stklen_i, sel_fun = sel_fun, catch_by_gear = cbg[i, ])

## End(Not run)

Description

Catch by gear and year

Usage

catch_gear(catch_n, wt = NULL, type = c("weight", "numbers"))

Arguments

Value

'FLQuants' of annual catch by gear (age summed).

Description

Splits the catch implied by the stock's realised total F among gears:

$C_{g,a,y}=N_{a,y}\frac{\phi_{g,a,y}F_{a,y}}{Z_{a,y}} \left(1-e^{-Z_{a,y}}\right),\qquad Z_{a,y}=F_{a,y}+M_{a,y}.$

Because shares sum to one, $\sum_g C_{g,a,y}$ equals the Baranov total catch exactly. This is the canonical definition used by the OM and by the OEM ('sampling.lfd.oem()'), which should call this function rather than reproduce it.

Usage

catch_n_gear(stock, shares, years = dimnames(stock)$year, check = TRUE)

Arguments

Value

'FLQuants' of catch numbers-at-age by gear. The Baranov total is attached as 'attr(, "total")'.

Examples

## Not run: 
cn <- catch_n_gear(om_hist, om_gears, years = 1972:2024)  # static shares
cn <- catch_n_gear(om_hist, f_share_gear(f_age))          # time-varying

## End(Not run)

Description

Combines gear-specific selectivity-at-age curves into a single joint selectivity-at-age curve using relative apical partial fishing mortality weights by gear.

Usage

combine_sel_age(object, fapic_rel, normalise_fapic = FALSE, scale = TRUE)

Arguments

Details

Gear-specific selectivity curves are assumed to be scaled to a maximum of one. The combined raw fishing-mortality shape is calculated as:

'sel[a] = sum_g fapic_rel[g] * sel_g[a]'

If 'scale = TRUE', the combined curve is scaled to a maximum of one:

'sel[a] = sel[a] / max(sel)'

The weights are interpreted as relative apical partial F contributions by gear, not relative Fbar contributions. This avoids dependence on an arbitrary Fbar age range and is consistent with constructing fleet-specific fishing mortality as:

'F_g[a, y] = Fapic_g[y] * sel_g[a]'

before summing across gears.

Value

An 'FLQuant' containing the combined selectivity-at-age.

Examples

## Not run: 
## Preferred: named vector
fapic_rel <- c(traps = 1, gillnet = 0.3, linefish = 0.1)

sel_joint <- combine_sel_age(
  object = sels$siganus_sutor,
  fapic_rel = fapic_rel
)

plot_sel_age(sel_joint)

## Convenience: unnamed vector matched to names(object)
sel_joint <- combine_sel_age(
  object = sels$siganus_sutor,
  fapic_rel = c(1, 0.3, 0.1)
)

## End(Not run)

Description

computeFbrp() Computes biological reference points corresponding to the proxy Fbrp

Usage

computeFbrp(
  stock,
  sr = "missing",
  proxy = NULL,
  x = NULL,
  blim = 0.1,
  type = c("b0", "btgt", "value"),
  btri = "missing",
  bpa = "missing",
  bthresh = "missing",
  verbose = T,
  fmax = 10,
  ...
)

Arguments

Value

brp object of class FLBRP with computed Fbrp reference points

Examples

data(ple4)
srr = srrTMB(as.FLSR(ple4,model=rickerSV),spr0=spr0y(ple4))
brp = computeFbrp(stock=ple4,sr=srr,proxy=c("sprx","f0.1"),blim=0.1,type="b0")
ploteq(brp,obs=TRUE,refpts="msy")

Description

computeFbrps() Computes biological reference points corresponding to the proxy Fbrp

Usage

computeFbrps(
  stock,
  sr = "missing",
  proxy = c("sprx", "bx", "all"),
  fmsy = FALSE,
  f0.1 = TRUE,
  fmax = 5,
  verbose = T,
  ...
)

Arguments

Value

brp object of class FLBRP with computed Fbrp reference points

Description

Reweights an inverse age-length key by selectivity-at-length and normalises each age row to obtain the conditional distribution of length among fish caught by a gear.

Usage

condition_alk(ialk_mat, sel_len_vec)

Arguments

Value

A numeric matrix with the same dimensions as 'ialk_mat'. Rows sum to one, except ages with zero selected probability, which are returned as zero rows.

Description

$F_{g,a,y}=\tilde s_{g,a}\,\bar F_{g,y}$

with $\tilde s_{g,a}$ the Fbar-standardised selectivity stored by [build_gear()]. By construction ‘fbar()' of each gear’s F-at-age equals the input trajectory.

Usage

f_age_gear(stock, om_gears, fbar_gear)

Arguments

Value

'FLQuants' of F-at-age by gear over the years of 'fbar_gear'.

Examples

## Not run: 
fbar_gear <- FLQuants(lapply(gear_par(om_gears, "f_mult"),
                             function(w) f_hist_base * w))
f_age <- f_age_gear(stki, om_gears, fbar_gear)
harvest(stki)[, dimnames(fbar_gear[[1]])$year] <- Reduce(`+`, f_age)

## End(Not run)

Description

$\phi_{g,a,y}=F_{g,a,y}\Big/\sum_h F_{h,a,y}$

General form of 'om_gears[[g]]$f_share', needed when gear Fbar trajectories are not proportional (e.g. an MP that manages one gear only).

Usage

f_share_gear(f_age)

Arguments

Value

'FLQuants' of shares (0 where total F is 0).

Description

Fbrp() Extract Fbrp based reference points from output of computeFbrp

Usage

Fbrp(brp)

Arguments

Value

FLPar object with computed Fbrp reference points

Description

Fe40() Patterson estimator for Fmsy

Usage

Fe40(stock, nyears = 3)

Arguments

Value

value

Description

flr2stars()

Usage

flr2stars(object, uncertainty = NULL, quantiles = c(0.05, 0.95))

Arguments

Value

STARS list with $timeseris and $refpts

Description

Fmmy() Uses opt.bisect to derive the F at Maximum Median Yield from stochastic simulations

Usage

Fmmy(
  brp,
  sigmaR = 0.5,
  rho = 0,
  nyears = 100,
  iters = 250,
  yrs.eval = NULL,
  range = "missing",
  tol = 0.001,
  maxit = 15,
  verbose = TRUE
)

Arguments

Value

list of FLPar, FLStock and FLBRP objects

Examples

data(ple4)
bh = srrTMB(as.FLSR(ple4,model=bevholtSV),spr0=spr0y(ple4))
brp = computeFbrp(ple4,bh,proxy=c("bx","msy"),x=35,blim=0.1)
fmmy = Fmmy(brp,sigmaR=0.7,rho=0.3)
getF(fmmy) # FMMY value 
plotFsim(fmmy)
brpfmmy = computeFbrp(ple4,bh,proxy=getF(fmmy),blim=0.1)
fsim = Fsim(brpfmmy,sigmaR=0.7,rho=0.3)
plotFsim(fsim)

Description

Calculates the Fbar value giving a maximum probability of ssb being below Blim of 5 percent

Usage

Fp05(
  object,
  iters = "missing",
  range = "missing",
  tol = 0.001,
  maxit = 20,
  verbose = TRUE
)

Arguments

Value

list

Examples

data(ple4)
bh = srrTMB(as.FLSR(ple4,model=bevholtSV),spr0=spr0y(ple4))
brp = computeFbrp(ple4,bh,proxy="bx",x=35,blim=0.2) # set Blim higher
fsim = Fsim(brp,sigmaR=0.7,rho=0.3,iters=500)
plotFsim(fsim)
fp.05 = Fp05(fsim)
plotFsim(fp.05,panels=c(2,4)) # black line is Fp0.05
getF(fp.05)

Description

Fsim() Simulates stochastic stock dynamics under under constant Fbrp

Usage

Fsim(
  brp,
  Ftgt = NULL,
  sigmaR = 0.5,
  rho = 0,
  nyears = 100,
  iters = 250,
  yrs.eval = NULL,
  verbose = TRUE
)

Arguments

Value

list of FLPar, FLStock and FLBRP objects

Examples

data(ple4)
hs = srrTMB(as.FLSR(ple4,model=segreg),spr0=spr0y(ple4),lplim=0.05,uplim=0.25)
blim = params(hs)[[2]]
brp = computeFbrp(ple4,hs,proxy=c("sprx","f0.1","msy"),x=40,blim=blim)
ploteq(brp)
fsim = Fsim(brp,sigmaR=0.7,rho=0.3)
plotFsim(fsim)
plotFsim(fsim,panels=2)

Description

Runs stochastic FLR forward projections with ffwd() over a grid of fishing mortality values. By default, the grid is taken from fbar(brp), with the first value replaced by Fmsy. For each F value, a corresponding FLBRP, FLStock, and FLSR object is constructed, propagated over iterations, and projected forward using recruitment deviances generated by rlnormar1().

Usage

Fsim_grid(
  brp,
  Fgrid = NULL,
  sigmaR = 0.5,
  rho = 0,
  nyears = 100,
  iters = 10,
  yrs.eval = NULL,
  parallel = FALSE,
  workers = NULL,
  progress = TRUE,
  verbose = TRUE
)

Arguments

Details

The function is intended for exploring the emergent biomass process error implied by an age-structured operating model under different constant fishing mortality levels. The output can be passed to pe_sampler() to obtain a data frame of depletion and process-error estimates.

The function constructs one FLBRP object per F value by replacing fbar(brp) with a constant F. These objects are then converted to FLStock and FLSR objects before stochastic forward projection.

Recruitment deviances are generated once using rlnormar1() and then passed to each ffwd() call. This allows the F scenarios to be compared under the same recruitment-deviation structure.

Value

An FLStocks object containing one projected FLStock per F value in the grid. The stocks are named using the corresponding F values.

Examples

## Not run: 
data(ple4)
ple4 <- merge_catch(ple4)
bh <- srrTMB(
  as.FLSR(ple4, model = bevholt),
  spr0 = spr0y(ple4),
  r0 = NULL
)

brp <- computeFbrp(ple4, bh, proxy = c("msy"))

sim <- Fsim_grid(
  brp,
  sigmaR = 0.5,
  rho = 0.3,
  nyears = 100,
  iters = 20,
  parallel = TRUE,
  workers= 12
)

class(sim)
names(sim)

## End(Not run)

Description

generates an up-down-constant F-pattern

Usage

fudc(
  object,
  fref = 0.2,
  fhi = 2.5,
  flo = 0.8,
  sigmaF = 0.2,
  breaks = c(0.5, 0.75)
)

Arguments

Value

FLQuant

Examples

data(ple4)
sr <- srrTMB(as.FLSR(ple4,model=bevholtSV),spr0=mean(spr0y(ple4)))
brp = computeFbrp(ple4,sr,proxy="msy")
fmsy = Fbrp(brp)["Fmsy"]
stki = propagate(ple4,100)
fy = fudc(ple4,fhi=2,flo=0.9,fref=fmsy,sigmaF=0)
fyi = fudc(stki,fhi=2,flo=0.9,fref=fmsy,sigmaF=0.2)
plot(fy,fyi)+ylab("F")
#Forcasting
om <- FLStockR(ffwd(stki,sr,fbar=fyi))
om@refpts = Fbrp(brp)
plotAdvice(window(om,start=1960))

Description

Function to summarise forecast results

Function to summarise ICES fishing opportunities

Usage

fwd2ices(stock, uncertainty, eval.yrs = NULL, dB = NULL, refyr = NULL)

fwd2ices(stock, uncertainty, eval.yrs = NULL, dB = NULL, refyr = NULL)

Arguments

Value

data.frame

data.frame

Description

Function to summarise forecast results

Usage

fwd2stars(object, eval.yrs = NULL, rel = NULL, dB = NULL, refyr = NULL)

Arguments

Value

data.frame

Description

Function to summarise Type 1 GFCM fishing opportunities

Usage

fwd2type1(
  stock,
  uncertainty = NULL,
  eval.yrs = NULL,
  dB = NULL,
  refyr = NULL,
  rel = NULL,
  osa = FALSE,
  maxdec = 2
)

Arguments

Value

data.frame

Description

fwdF4B () Bisection Function to search F for a given biomass

Usage

fwdF4B(
  stock,
  sr,
  btgt,
  nfy = 3,
  niy = 1,
  ival = niy,
  imet = "TAC",
  ftune = c(0, 2),
  tol = 0.001,
  verbose = TRUE,
  ssbQ = 1,
  recQ = 1
)

Arguments

Value

*FLStock*

Description

Builds $P(l \mid a, t)$ with exactly the grid, CV and timing stored in 'gear$grid', so its columns always match 'gear$sel_len' and 'gear$condALK'. The upper limit is passed to [iALK()] explicitly via ‘max', so 'iALK()'’s own 'lmax' default can no longer diverge.

Usage

gear_ialk(
  gear,
  age = as.numeric(rownames(gear$condALK)),
  timing = gear$grid$timing %||% 0,
  params = gear$lhpar,
  model = vonbert,
  reflen = NULL
)

Arguments

Value

Age x length matrix (rows sum to one).

Description

Extract a scalar gear entry across all gears

Usage

gear_par(om_gears, what)

Arguments

Value

Named vector (numeric, or character for '"type"').

Examples

## Not run: gear_par(om_gears, "f_mult")

Description

Helper functio to extract F from various FLRef output

Usage

getF(x)

Arguments

Description

huecol

Usage

huecol(n, alpha = 1)

Arguments

Description

Constructs the probability distribution of length conditional on age from a growth model and a coefficient of variation in length-at-age.

Usage

iALK(
  params,
  model = vonbert,
  age,
  cv = 0.1,
  lmin = 5,
  lmax = 1.2,
  bin = 1,
  max = ceiling(linf * lmax),
  reflen = NULL
)

Arguments

Value

An 'FLPar' containing an age-by-length inverse age-length key. Rows represent ages, columns represent length classes, and each age row sums to one.

Examples

## Not run: 
ialk <- iALK(
  params = FLPar(linf = 45, k = 0.4, t0 = -0.3),
  age = 0:5,
  lmin = 5,
  lmax = 1.2,
  bin = 1
)

## End(Not run)

Description

generates FLIndex with lognormal annual and multinomial age composition observation error

Usage

idx.sim(
  object,
  sel = catch.sel(object),
  ages = NULL,
  years = NULL,
  ess = 200,
  sigma = 0.2,
  q = 0.01
)

Arguments

Value

FLIndex

Examples

data(ple4)
sel = newselex(catch.sel(ple4),FLPar(S50=1.5,S95=2.1,Smax=4.5,Dcv=1,Dmin=0.1))
ggplot(sel)+geom_line(aes(age,data))+ylab("Selectivity")+xlab("Age")
object = propagate(ple4,10)
idx = idx.sim(object,sel=sel,ess=200,sigma=0.2,q=0.01,years=1994:2017)
# Checks
ggplot(idx@sel.pattern)+geom_line(aes(age,data))+ylab("Selectivity")+xlab("Age")
ggplot(idx@index)+geom_line(aes(year,data,col=ac(iter)))+facet_wrap(~age,scales="free_y")+
theme(legend.position = "none")+ylab("Index")

Description

jabba2FLStockR()

jabba2FLStockR()

Usage

jabba2FLStockR(jabba, blim = 0.3, bthr = 0.5, thin = 10, rel = FALSE)

jabba2FLStockR(jabba, blim = 0.3, bthr = 0.5, thin = 10, rel = FALSE)

Arguments

Value

FLStockR with refpts

FLStockR with refpts

Description

jabba2stars()

Usage

jabba2stars(jabba, quantiles = c(0.05, 0.95), blim = 0.3, bthr = 0.5)

Arguments

Value

STARS list with $timeseris and $refpts

Description

Combines gear selectivities on the Fbar scale, weighted by each gear's share of Fbar ('f_mult'):

$S_a=\frac{\sum_g w_g\tilde s_{g,a}}{\max_a\sum_g w_g\tilde s_{g,a}}.$

This is the selectivity of total F implied by [f_age_gear()], and is what the aggregate 'FLBRP'/'FLStock' must carry for the equilibrium reference points to be consistent with the multigear dynamics.

Usage

joint_sel(om_gears, scale = TRUE)

Arguments

Details

Note that [combine_sel_age()] weights *max-scaled* curves by relative *apical* F; the two agree only if the weights are converted between bases. Use 'joint_sel()' whenever 'f_mult' is defined as a share of Fbar.

Value

'FLQuant' of joint selectivity-at-age.

Description

Lower bin limits 'seq(lmin, ceiling(linf * lmax_mult), bin)', exactly as [iALK()] builds them; the last bin is the plus group $[l_{max}, \infty)$.

Usage

len_grid(linf, lmin = 5, lmax_mult = 1.2, bin = 1)

Arguments

Details

Background: before this helper, 'sel_la()' (exclusive upper limit) and 'iALK()' (plus-group upper limit) built their grids differently, and every helper that rebuilt an 'iALK()' ('plot_condALK()', 'plot_condALK_bias()', 'lfd.sim(timing = )', 'sel_a_seasonal_avg()', 'wt_a_seasonal()', 'plot_sel_a_seasonal()') used its own default 'lmax_mult = 1.2'. With 'build_gear(lmax_mult = 1.3)' those helpers built a shorter grid than the gear's (e.g. 63 vs 69 bins for Linf = 55.7), and 'ialk take the grid from 'gear$grid' (via [gear_ialk()] or an explicit 'max').

Value

Numeric vector of lower bin limits.

Description

Projects numbers-at-age through an inverse age-length key at one or more within-year sampling times and draws a multinomial length sample. When a gear is supplied, the inverse key is conditioned on its selectivity-at-length.

Usage

len.sim(
  N_a,
  params,
  model = vonbert,
  ess = 250,
  timing = seq(0, 11/12, 1/12),
  unit = "cm",
  scale = TRUE,
  reflen = NULL,
  bin = gear$grid$bin %||% 1,
  cv = gear$grid$cv %||% 0.1,
  lmin = gear$grid$lmin %||% 5,
  lmax = gear$grid$lmax_mult %||% 1.2,
  gear = NULL,
  u = NULL
)

Arguments

Details

This is a single-stage length sampler. In contrast, [lfd.sim()] first samples catch-at-age at low effective sample size and then conditions the length draw on that sampled age composition.

Value

An 'FLQuant' containing sampled length frequencies by year and iteration.

Examples

## Not run: 
lfd_survey <- len.sim(stock.n(om)[, "2024"], params = lhpar, ess = 300,
                       timing = 0.5)
u_len <- rUnif_len(trawl, timing = seq(0, 11/12, 1/12),
                    years = an(dimnames(N_a)$year), nsim = dims(N_a)$iter,
                    seed = 456)
lfd_survey_g <- len.sim(N_a, params = lhpar, ess = trawl$ess_len,
                         timing = seq(0, 11/12, 1/12),
                         gear = trawl, u = u_len)

## End(Not run)

Description

Builds 'observations$lfd' and 'observations$catch_gear' for [sampling.lfd.oem()] over 'years' (typically 'y0:fy') and fills the historical years with samples drawn by [sample_lfd_gears()], i.e. with the same code and, when 'u_lfd' is given, the same common random numbers the OEM will use. Templates have the season dimension implied by 'season', so they always match the OEM's output.

Usage

lfd_obs_init(
  stock,
  om_gears,
  years,
  hist_years,
  u_lfd = NULL,
  season = FALSE,
  f_share = NULL
)

Arguments

Value

A list with 'lfd' and 'catch_gear' ('FLQuants' by gear).

Description

Length analogue of 'FLRef::sops()' (which raises catch-at-age with 'catch.wt'): scales each year/unit/iteration of a length sample so that its sum of products with weight-at-length equals the catch weight,

$\tilde N_{l,y,u}=N_{l,y,u}\, \frac{C_{y,u}}{\sum_{l'}N_{l',y,u}\,W(l')}.$

Years/units with zero sample or zero catch return zero. For a seasonal sample (season dimension > 1) and annual catch, one raising factor per year is computed from the sample pooled over seasons and applied to every season.

Usage

lfd_sop(lfd, wt_len, catch)

Arguments

Value

'FLQuant' of raised numbers-at-length, with 'attr(, "raising")' the raising factor per year/unit/iter.

See Also

[sampling.lfd.oem()] ('lfd.raising = TRUE')

Description

Empty length-sample template for a gear

Usage

lfd_template(gear, years, nsim, units = "unique", season = FALSE)

Arguments

Value

'FLQuant' of zeros, len x year x unit x season x area x iter.

Description

Generates fishery-dependent length-frequency data from gear-specific catch-at-age using a two-stage sampling design.

Usage

lfd.sim(
  object,
  gear,
  ess_age = gear$ess_age,
  ess_len = gear$ess_len,
  u_age = NULL,
  u_len = NULL,
  scale = TRUE,
  timing = NULL,
  params = gear$lhpar,
  model = vonbert,
  reflen = NULL,
  cv = gear$grid$cv %||% 0.1,
  lmin = gear$grid$lmin %||% 5,
  lmax_mult = gear$grid$lmax_mult %||% 1.2,
  bin = gear$grid$bin %||% 1
)

Arguments

Details

Sampling proceeds as follows:

1. A catch-at-age composition is sampled with effective sample size 'ess_age' using [ca.sim()].

2. The sampled age composition is projected through $P(l \mid a, g)$ and sampled at effective sample size 'ess_len'.

Value

An 'FLQuant' containing sampled length frequencies by year and iteration.

Examples

## Not run: 
lfd_trawl <- lfd.sim(catch_n_gear$Trawl, trawl)
devs <- rUnif_lfd(om_gears, years = an(dimnames(catch_n_gear$Trawl)$year),
                   nsim = dim(catch_n_gear$Trawl)[6], seed = 456)
lfd_trawl_crn <- lfd.sim(
  catch_n_gear$Trawl,
  trawl,
  u_age = devs$age$Trawl,
  u_len = devs$len$Trawl
)

## End(Not run)

Description

Applies [lfd.sim()] separately to each season of a season-structured catch-at-age object. Seasonal midpoint timing is '(season - 0.5) / n_seasons'.

Usage

lfd.sim.season(
  object,
  gear,
  n_seasons = dim(object)[4],
  ess_age = gear$ess_age,
  ess_len = gear$ess_len,
  params = gear$lhpar,
  u_age = NULL,
  u_len = NULL,
  season = TRUE,
  ...
)

Arguments

Details

Each season is an independent sampling event: a stage-1 age sample of size 'ess_age[s]' and a stage-2 length sample of size 'ess_len[s]', with the conditional ALK rebuilt at the season mid-point. Together with [season_fraction()] this emulates continuous (monthly or quarterly) length sampling of a fishery from an annual operating model.

Scalar sample sizes were previously split with 'round(ess / n_seasons)', which does not preserve the total (e.g. 50 / 12 gave 48); they are now split exactly.

Value

An 'FLQuant' of sampled length frequencies by season (or pooled over seasons when 'season = FALSE'), year and iteration.

Examples

## Not run: 
lfd.sim.season(obj, trawl, n_seasons = 4, params = lhpars)
lfd.sim.season(obj, trawl, n_seasons = 4, ess_len = c(0, 200, 200, 100),
                params = lhpars)

## End(Not run)

Description

Converts an FLStock object with total catch numbers into a landings-only stock by assigning catch.n to landings.n and setting discards.n to zero. Landings and discards weights are then recomputed.

Usage

merge_catch(stock)

Arguments

Details

This is a convenience function for cases where catch should be treated as landed catch, for example before forward projection or when preparing simplified single-fleet inputs. The function modifies:

• landings.n(stock): set equal to catch.n(stock)

• discards.n(stock): set to zero

• landings(stock): recomputed with computeLandings()

• discards(stock): recomputed with computeDiscards()

Value

An FLStock object where all catch numbers are assigned to landings and discards are set to zero.

Examples

## Not run: 
stock <- merge_catch(stock)

# Check that discards have been removed
range(discards.n(stock))
discards(stock)

## End(Not run)

Description

computes Lorenzen M with scaling option

Usage

Mlorenzen(object, Mref = "missing", Aref = 2)

Arguments

Value

FLQuant m()

Examples

data(ple4)
Ml = Mlorenzen(stock.wt(ple4))
# Scale
Ms = Mlorenzen(stock.wt(ple4),Mref=0.2,Aref=2)
flqs = FLQuants(Lorenzen=Ml,Scaled=Ms)

Description

generates flexible 5-paramater selex curves

Usage

newselex(object, selexpars)

Arguments

Value

FLquant with selectivity pattern

Examples

data(ple4)
sel = newselex(catch.sel(ple4),FLPar(S50=2,S95=3,Smax=4.5,Dcv=0.6,Dmin=0.3))
ggplot(sel)+geom_line(aes(age,data))+ylab("Selectivity")+xlab("Age")
# Simulate
harvest(ple4)[] = sel
sr <- srrTMB(as.FLSR(ple4,model=bevholtSV),spr0=mean(spr0y(ple4)))
brp = computeFbrp(ple4,sr,proxy="msy") 
fbar(brp) = FLQuant(rep(0.01,70))
stk = as(brp,"FLStock")
units(stk) = standardUnits(stk)
its = 100
stk <- FLStockR(propagate(stk, its))
stk@refpts= Fbrp(brp)
b0=an(Fbrp(brp)["B0"])
control = FLPar(Feq=0.15,Frate=0.1,Fsigma=0.15,SB0=b0,minyear=2,maxyear=70,its=its)
run <- rffwd(stk, sr=sr,control=control,deviances=ar1rlnorm(0.3, 1:70, its, 0, 0.6))
plotAdvice(run)

Description

The plain bisection algorithm (Burden & Douglas, 1985) is employed here to find the value of a given forecast target quantity (e.g. 'fbar') for which a selected value of a performance statistic is obtained over a chosen period.

Usage

opt.bisect(
  stock,
  sr,
  deviances = rec(stock) %=% 1,
  metrics,
  statistic,
  years,
  pyears = years,
  tune,
  tol = 0.001,
  maxit = 15,
  log = TRUE,
  verbose = TRUE
)

Arguments

Author(s)

Credits to Iago Mosqueira

References

Burden, Richard L.; Faires, J. Douglas (1985), "2.1 The Bisection Algorithm", Numerical Analysis (3rd ed.), PWS Publishers, ISBN 0-87150-857-5

Examples

data(ple4)
stock <- propagate(stf(ple4, end=2118), 200)
srr <- predictModel(model=rec ~ ifelse(ssb <= b, a * ssb, a * b), params=FLPar(a=1.29, b=1.35e+06))
# GENERATE SRR deviances
devs <- ar1rlnorm(rho=0.4, 2018:2118, iters=200, meanlog=0, sdlog=0.5)
# DEFINE MMY statistic
statistic <- list(MMY=list(~apply(L,1,median), name="MMY",
  desc="ICES Maximum Median Yield"))
# CALL bisect over 100 years, Fmmy calculated over last 50.
fmmy <- opt.bisect(stock, sr=srr, deviances=devs, metrics=list(L=landings), 
statistic=statistic, years=2018:2118,
pyears=2069:2118, tune=list(fbar=c(0.01, 0.2)))
# fmmy
mean(fbar(fmmy)[,ac(2069:2118)])

Description

Applies sample_proc_dev() to each projected stock in an FLStocks object and returns a compact data frame of F, depletion, and process-error estimates.

Usage

pe_sampler(stks, brp, biomass = c("vb", "ssb"), eps = 1e-12)

Arguments

Details

This is a convenience wrapper around sample_proc_dev(). It is useful for producing the main diagnostic relationship:

$\sigma_{proc} \sim B / B_{MSY}$

The output can be plotted directly to visualise how the emergent biomass process error changes with stock depletion or fishing pressure.

Value

A data.frame with one row per F scenario and columns:

f

Median fishing mortality in the projected stock.

sigma

Sampled biomass process-error standard deviation.

bbmsy

Median biomass relative to Bmsy.

bb0

Median biomass relative to B0.

Examples

## Not run: 
data(ple4)

bh <- srrTMB(
  as.FLSR(ple4, model = bevholt),
  spr0 = spr0y(ple4),
  r0 = NULL
)

brp <- computeFbrp(ple4, bh, proxy = c("msy"))

sim <- Fsim_grid(
  brp,
  sigmaR = 0.5,
  rho = 0.3,
  nyears = 100,
  iters = 100,
  parallel = TRUE,
  workers = 4
)

ps <- pe_sampler(sim, brp, biomass = "vb")

head(ps)

plot_pe(ps)

## End(Not run)

Description

sets plus group on FLQuant

Usage

pgquant(object, pg)

Arguments

Value

FLQuant

Description

UPDATED: the optional mean-length overlay is built with [gear_ialk()], so it uses the gear's own grid (any 'lmax_mult') and timing. Previously a hard-coded 'lmax_mult = 1.2' gave a non-conformable ' was built with a different 'lmax_mult', and 't = 0' biology was compared to 't = timing' catch.

Usage

plot_condALK(gear, params = NULL, low = "#FFFFCC", high = "#7F0000", ...)

Arguments

Value

A 'ggplot' object.

Description

UPDATED: the raw biological line is now evaluated at the same within-year 'timing' (and grid) the gears were built with. Previously it was evaluated at 'timing = 0' while 'condALK' used 't0 - timing', so for 'timing = 0.5' half a year of growth appeared as "selectivity bias".

Usage

plot_condALK_bias(
  om_gears,
  params = om_gears[[1]]$lhpar,
  timing = om_gears[[1]]$grid$timing %||% 0,
  cv = om_gears[[1]]$grid$cv %||% 0.1,
  lmin = om_gears[[1]]$grid$lmin %||% 5,
  lmax_mult = om_gears[[1]]$grid$lmax_mult %||% 1.2,
  bin = om_gears[[1]]$grid$bin %||% 1
)

Arguments

Value

A 'ggplot' object.

Description

Overlays the annual length composition (proportions) obtained under different sampling designs, e.g. one mid-year sample against monthly sampling pooled over the year, optionally with the expected composition. Seasonal samples are pooled with 'seasonSums()'.

Usage

plot_lfd_designs(designs, year, iter = 1, expected = NULL)

Arguments

Value

A 'ggplot' object.

Description

Displays annual or seasonal length-frequency distributions along a continuous time axis, with one panel per gear. Optional von Bertalanffy cohort trajectories can be overlaid for reference.

Usage

plot_lfd_ridgeline(
  panels,
  lhpar = NULL,
  scale = 0.9,
  cohort_pad = 15,
  annual = FALSE
)

Arguments

Details

For seasonal data, ridge positions are calculated as 'year + (season - 0.5) / n_seasons'. When 'annual = TRUE', seasonal data are aggregated with 'seasonSums()'.

Value

A 'ggplot' object.

Examples

## Not run: 
plot_lfd_ridgeline(lfds, lhpar = lhpars)
plot_lfd_ridgeline(lfd_season, annual = TRUE, lhpar = lhpars)

## End(Not run)

Description

Plots seasonal length-frequency distributions for a selected year, with seasons distinguished by colour and gears shown in separate panels.

Usage

plot_lfd_season(lfd_season, year, iter = 1)

Arguments

Value

A 'ggplot' object.

Examples

## Not run: 
plot_lfd_season(list(Trawl = lfd_season_trawl, Gillnet = lfd_season_gillnet),
                 year = "2020")

## End(Not run)

Description

Plots sampled biomass process error against biomass relative to Bmsy. Optionally both axes can be normalised relative to a reference row, typically row 1, which corresponds to the Fmsy scenario in Fsim_grid().

Usage

plot_pe(ps, rel = TRUE, ref = 1, line = TRUE, points = TRUE, reverse_x = FALSE)

Arguments

Value

A ggplot object.

Description

plots production functions

Usage

plot_pf(object, quant = c("vb", "ssb"), fmsy = NULL, rel = FALSE)

Arguments

Value

ggplot

Examples

data(ple4)
ple4 <- merge_catch(ple4)
sr <- srrTMB(as.FLSR(ple4,model=bevholtSV),spr0=mean(spr0y(ple4)))
brp = brp(FLBRP(ple4,sr))
asem2spm(brp)[1:4]
plot_pf(brp)
plot_pf(brp,rel=TRUE)

Description

Plots the selectivity-at-age stored in a gear object together with the corresponding seasonally averaged curve. Individual seasonal curves can be displayed as reference lines.

Usage

plot_sel_a_seasonal(
  gear,
  lhpar,
  age,
  n_seasons,
  show_seasons = TRUE,
  model = vonbert,
  reflen = NULL,
  cv = gear$grid$cv %||% 0.1,
  lmin = gear$grid$lmin %||% 5,
  lmax_mult = gear$grid$lmax_mult %||% 1.2,
  bin = gear$grid$bin %||% 1
)

Arguments

Value

A 'ggplot' object.

Examples

## Not run: 
plot_sel_a_seasonal(
  om_gears$Trawl,
  lhpars,
  age = 0:5,
  n_seasons = 12
)

## End(Not run)

Description

Plots one or more selectivity-at-age curves stored as 'FLQuant' objects.

Usage

plot_sel_age(
  object,
  age_by = 1,
  points = TRUE,
  title = NULL,
  line_width = 0.7,
  point_size = 1.5,
  colours = NULL,
  ylab = "Selectivity",
  xlab = "Age"
)

Arguments

Details

The function accepts either a single 'FLQuant' or a named list of 'FLQuant' objects. It is intended for comparing joint or gear-specific selectivity-at-age curves used in operating models.

The age dimension is extracted from the 'age' dimension of each 'FLQuant'. Plus-group labels such as '"10+"' are converted to numeric values for plotting.

If a single 'FLQuant' is provided, the legend is suppressed. If a list of 'FLQuant' objects is provided, the legend is shown using the list names.

Value

A 'ggplot' object.

Examples

## Not run: 
## Single curve
plot_sel_age(sel_joint)

## Multiple species
sel_joint <- list(
  siganus = sel_joint_siganus,
  lethrinus = sel_joint_lethrinus,
  lutjanus = sel_joint_lutjanus,
  shark = sel_joint_shark
)

plot_sel_age(sel_joint)

## End(Not run)

Description

Plots selectivity-at-length and selectivity-at-age side by side for one selectivity object or for a named list of gear-specific selectivity objects.

Usage

plot_sel_la(
  object,
  ncol = 2,
  points_age = TRUE,
  title = NULL,
  line_width = 0.7,
  point_size = 1.5,
  len_by = 5,
  age_by = 1,
  colours = NULL
)

Arguments

Details

The function is intended as a diagnostic plot for operating-model selectivity assumptions. It accepts objects returned by 'sel_la()', where selectivity-at-length and selectivity-at-age are stored as 'FLQuant' objects. If a named list is supplied, each element is treated as a gear and overlaid in both panels.

Length selectivity is plotted against length-bin midpoints when these are available through 'object$len_bins' or the '"len_bins"' attribute of the length 'FLQuant'. Otherwise, midpoints are inferred from the lower length-bin labels and the median bin width.

Value

A 'ggplot' object.

Examples

## Not run: 
plot_sel_la(sels$lutjanus_bohar, len_by = 10, age_by = 1)

plot_sel_la(
  sels$carcharhinus_melanopterus,
  len_by = 20,
  age_by = 2,
  title = expression(italic("Carcharhinus melanopterus"))
)

## End(Not run)

Description

plotAdvice Plots stochastic stock dynamics against refpts for constant Fsim()

Usage

plotAdvice(
  object,
  rpts = "missing",
  type = NULL,
  yield = c("catch", "landings")[1],
  plotrefs = TRUE,
  probs = c(0.05, 0.2, 0.5, 0.8, 0.95),
  colour = "dodgerblue",
  ncol = NULL,
  osa = FALSE,
  label.size = 2.5,
  ssbQ = 1,
  recQ = 1
)

Arguments

Value

ggplot

Examples

data(ple4)
srr = srrTMB(as.FLSR(ple4,model=rickerSV),spr0=spr0y(ple4))
brp = computeFbrp(stock=ple4,sr=srr,proxy=c("sprx","f0.1","fe40"),blim=0.1,type="b0")
plotAdvice (ple4,brp)

Description

plotAR Plots the new proposed ICES advice rule

Usage

plotAR(
  pars,
  ftgt = 1,
  btrigger = "missing",
  bpa = "missing",
  bthresh = "missing",
  fpa = "missing",
  fthresh = "missing",
  bclose = 0,
  fmin = 0,
  obs = "missing",
  kobe = TRUE,
  alpha = 1,
  xmax = 1.2,
  ymax = 1.5,
  ylab = "missing",
  xlab = "missing",
  rel = FALSE,
  expand = TRUE,
  labels = TRUE,
  label.cex = 3.5,
  critical = TRUE
)

Arguments

Value

ggplot

Examples

data(ple4)
srr = srrTMB(as.FLSR(ple4,model=segreg),spr0=spr0y(ple4))
blim = params(srr)[[2]]
brp = computeFbrp(stock=ple4,sr=srr,proxy="f0.1",blim=blim)
rpt = Fbrp(brp)
plotAR(rpt,btrigger=an(0.8*rpt["Btgt"]))
# Use Bpa as trigger (ICES style)
plotAR(rpt,obs=ple4,bpa=1.4)
# Change kobe to greyscale
plotAR(rpt,obs=ple4,bpa=1.4,kobe=FALSE)
# add fishing closure with minimum unavoidable F and Btrigger
plotAR(rpt,obs=ple4,bpa=1.4,btrigger=0.7,kobe=TRUE,bclose=1,fmin=0.01)
# show a relative
plotAR(rpt,obs=ple4,rel=TRUE,bpa=1.4,btrigger=0.7,kobe=TRUE,bclose=1,fmin=0.02)

Description

plotbioage() Plots stock N_a, W_a, M_a and Mat_a by year

Usage

plotbioage(
  stk,
  metrics = c("Weight", "Maturity", "M", "Selectivity"),
  bysex = FALSE,
  ncol = 2
)

Arguments

Value

ggplot

Examples

data(ple4)
plotbioage(ple4)

Description

plotbioyr() Plots stock N_a, W_a, M_a and Mat_a across years

Usage

plotbioyr(stk, ncol = 2)

Arguments

Value

ggplot

Examples

data(ple4)
plotbioyr(ple4)

Description

Produces candidate GFCM advice plot

Usage

plotCECAF(
  ftgt = 1,
  btgt = 1,
  blim = 0.3,
  btrigger = "missing",
  bthr = 0.8,
  fthr = 1.2,
  bclose = 0.3,
  fmin = 0.2,
  fadv = 0.8,
  obs = "missing",
  kobe = TRUE,
  alpha = 1,
  xmax = 1.3,
  ymax = 1.5,
  ylab = "missing",
  xlab = "missing",
  rel = TRUE,
  expand = TRUE,
  labels = TRUE,
  critical = kobe,
  text = TRUE
)