| Title: | A Simple and Robust Statistical Catch at Age Model |
|---|---|
| Description: | A simple and robust statistical Catch at Age model that is specifically designed for stocks with intermediate levels of data quantity and quality. |
| Authors: | Colin P. Millar [aut], Ernesto Jardim [aut, cre] |
| Maintainer: | Ernesto Jardim <[email protected]> |
| License: | EUPL |
| Version: | 1.9.0 |
| Built: | 2024-11-22 06:08:59 UTC |
| Source: | https://github.com/flr/FLa4a |
Update FLStock and FLIndex objects with simulations from stock assessment fits.
## S4 method for signature 'FLStock,a4aFitSA' e1 * e2 ## S4 method for signature 'FLStock,SCAPars' e1 * e2 ## S4 method for signature 'FLIndices,a4aFitSA' e1 * e2 ## S4 method for signature 'FLIndices,SCAPars' e1 * e2## S4 method for signature 'FLStock,a4aFitSA' e1 * e2 ## S4 method for signature 'FLStock,SCAPars' e1 * e2 ## S4 method for signature 'FLIndices,a4aFitSA' e1 * e2 ## S4 method for signature 'FLIndices,SCAPars' e1 * e2
e1 |
the original |
e2 |
a |
a4aFit
The a4aFit class was built to store the a4a stock assessment fits.
a4aFit(...) a4aFit(...) clock(object, ...) ## S4 method for signature 'a4aFit' clock(object) fitSumm(object, ...) ## S4 method for signature 'a4aFit' fitSumm(object) ## S4 method for signature 'a4aFit' stock.n(object) ## S4 method for signature 'a4aFit,ANY' harvest(object) ## S4 method for signature 'a4aFit' catch.n(object) ## S4 method for signature 'a4aFit' index(object) ## S4 method for signature 'a4aFit' show(object) ## S4 method for signature 'a4aFit' logLik(object, ...) ## S4 method for signature 'a4aFit' iter(obj, it) ## S4 method for signature 'a4aFit' computeCatchDiagnostics(object, stock, ...)a4aFit(...) a4aFit(...) clock(object, ...) ## S4 method for signature 'a4aFit' clock(object) fitSumm(object, ...) ## S4 method for signature 'a4aFit' fitSumm(object) ## S4 method for signature 'a4aFit' stock.n(object) ## S4 method for signature 'a4aFit,ANY' harvest(object) ## S4 method for signature 'a4aFit' catch.n(object) ## S4 method for signature 'a4aFit' index(object) ## S4 method for signature 'a4aFit' show(object) ## S4 method for signature 'a4aFit' logLik(object, ...) ## S4 method for signature 'a4aFit' iter(obj, it) ## S4 method for signature 'a4aFit' computeCatchDiagnostics(object, stock, ...)
... |
additional argument list that might never be used |
object |
object of relevant class (see signature of method) |
obj |
the object to be subset |
it |
iteration to be extracted |
The function call
Information on call duration
Fit summary
Estimates of stock numbers-at-age
Estimates of fishing mortality at age
Estimates of catch numbers-at-age
Estimates of survey or CPUE indices-at-age
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj)data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj)
a4aFitCatchDiagn
The a4aFitCatchDiagn class extends FLQuants to store information to run diagnostics on aggregated catch estimated by the a4a stock assessment fit.
computeCatchDiagnostics(object, ...)computeCatchDiagnostics(object, ...)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
stock |
|
indices |
|
data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index))data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index))
a4aFitMCMC
The a4aFitMCMC class extends a4aFitSA to store information about the MCMC run.
a4aFitMCMC(...) a4aFitMCMC(...) ## S4 method for signature 'a4aFitMCMC' a4aFitSA(object, ...) ## S4 method for signature 'a4aFitMCMC' a4aFit(object, ...) as.mcmc(x, ...) ## S4 method for signature 'a4aFitMCMC' as.mcmc(x, ...) burnin(object, ...) ## S4 method for signature 'a4aFitMCMC' burnin(object, burnin)a4aFitMCMC(...) a4aFitMCMC(...) ## S4 method for signature 'a4aFitMCMC' a4aFitSA(object, ...) ## S4 method for signature 'a4aFitMCMC' a4aFit(object, ...) as.mcmc(x, ...) ## S4 method for signature 'a4aFitMCMC' as.mcmc(x, ...) burnin(object, ...) ## S4 method for signature 'a4aFitMCMC' burnin(object, burnin)
... |
additional argument list that might never be used |
object |
object of relevant class (see signature of method) |
x |
an object to be coerced into mcmc |
burnin |
a numeric with the number of iterations to be removed |
A character vector for the object name.
A textual description of the object contents.
A named numeric vector with various values of quant and year ranges, plusgroup, fishing mortality ranges, etc.
The function call
Information on call duration
Fit summary
Estimates of stock numbers-at-age
Estimates of fishing mortality at age
Estimates of catch numbers-at-age
Estimates of survey or CPUE indices-at-age
An object of class SCAMCMC with information about the MCMC run
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index), fit="assessment") obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj) is(pars(obj))data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index), fit="assessment") obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj) is(pars(obj))
a4aFitResiduals
The a4aFitResiduals class extends FLQuants to store residuals of the a4a stock assessment fit. By default, these should be log residuals of catches and indices.
## S4 method for signature 'a4aFit' residuals(object, stock, indices, ...)## S4 method for signature 'a4aFit' residuals(object, stock, indices, ...)
object |
object of relevant class (see signature of method) |
stock |
|
indices |
|
... |
additional argument list that might never be used |
data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index))data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index))
a4aFitSA
The a4aFitSA class extends a4aFit to store information about the parameters of the model.
a4aFitSA(...) a4aFitSA(...) ## S4 method for signature 'a4aFitSA' a4aFit(object, ...) pars(object) ## S4 method for signature 'a4aFitSA' pars(object) ## S4 method for signature 'a4aFitSA' m(object) ## S4 method for signature 'a4aFitSA' wt(object) ## S4 method for signature 'a4aFitSA' qmodel(object) ## S4 method for signature 'a4aFitSA' fmodel(object) ## S4 method for signature 'a4aFitSA' srmodel(object) ## S4 method for signature 'a4aFitSA' n1model(object) ## S4 method for signature 'a4aFitSA' vmodel(object) ## S4 method for signature 'a4aFitSA' stkmodel(object) ## S4 method for signature 'a4aFitSA' show(object) ## S4 method for signature 'a4aFitSA' submodels(object, ...) ## S4 method for signature 'a4aFitSA' iter(obj, it) a4aFitSAs(object, ...) ## S4 method for signature 'list' a4aFitSAs(object, ...) ## S4 method for signature 'a4aFitSA' a4aFitSAs(object, ...) ## S4 method for signature 'missing' a4aFitSAs(object, ...)a4aFitSA(...) a4aFitSA(...) ## S4 method for signature 'a4aFitSA' a4aFit(object, ...) pars(object) ## S4 method for signature 'a4aFitSA' pars(object) ## S4 method for signature 'a4aFitSA' m(object) ## S4 method for signature 'a4aFitSA' wt(object) ## S4 method for signature 'a4aFitSA' qmodel(object) ## S4 method for signature 'a4aFitSA' fmodel(object) ## S4 method for signature 'a4aFitSA' srmodel(object) ## S4 method for signature 'a4aFitSA' n1model(object) ## S4 method for signature 'a4aFitSA' vmodel(object) ## S4 method for signature 'a4aFitSA' stkmodel(object) ## S4 method for signature 'a4aFitSA' show(object) ## S4 method for signature 'a4aFitSA' submodels(object, ...) ## S4 method for signature 'a4aFitSA' iter(obj, it) a4aFitSAs(object, ...) ## S4 method for signature 'list' a4aFitSAs(object, ...) ## S4 method for signature 'a4aFitSA' a4aFitSAs(object, ...) ## S4 method for signature 'missing' a4aFitSAs(object, ...)
... |
additional argument list that might never be used |
object |
object of relevant class (see signature of method) |
obj |
the object to be subset |
it |
iteration to be extracted |
The function call
Information on call duration
Fit summary
Estimates of stock numbers-at-age
Estimates of fishing mortality at age
Estimates of catch numbers-at-age
Estimates of survey or CPUE indices-at-age
an object of class SCAPars with information about model parameters
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index), fit="assessment") obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj) is(pars(obj))data(ple4) data(ple4.index) obj <- sca(stock=ple4, indices=FLIndices(ple4.index), fit="assessment") obj slotNames(obj) clock(obj) fitSumm(obj) flq <- stock.n(obj) is(flq) flq <- index(obj) is(flq) logLik(obj) AIC(obj) BIC(obj) is(pars(obj))
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
a4aGr(object, ...) ## S4 method for signature 'missing' a4aGr(object, ...) grMod(object, ...) ## S4 method for signature 'a4aGr' grMod(object) grMod(object) <- value ## S4 replacement method for signature 'a4aGr,formula' grMod(object) <- value grInvMod(object, ...) ## S4 method for signature 'a4aGr' grInvMod(object) grInvMod(object) <- value ## S4 replacement method for signature 'a4aGr,formula' grInvMod(object) <- value ## S4 method for signature 'a4aGr' params(object) ## S4 replacement method for signature 'a4aGr,FLPar' params(object) <- value ## S4 method for signature 'a4aGr' distr(object) ## S4 replacement method for signature 'a4aGr,character' distr(object) <- value ## S4 method for signature 'a4aGr' vcov(object) ## S4 replacement method for signature 'a4aGr,numeric' vcov(object) <- valuea4aGr(object, ...) ## S4 method for signature 'missing' a4aGr(object, ...) grMod(object, ...) ## S4 method for signature 'a4aGr' grMod(object) grMod(object) <- value ## S4 replacement method for signature 'a4aGr,formula' grMod(object) <- value grInvMod(object, ...) ## S4 method for signature 'a4aGr' grInvMod(object) grInvMod(object) <- value ## S4 replacement method for signature 'a4aGr,formula' grInvMod(object) <- value ## S4 method for signature 'a4aGr' params(object) ## S4 replacement method for signature 'a4aGr,FLPar' params(object) <- value ## S4 method for signature 'a4aGr' distr(object) ## S4 replacement method for signature 'a4aGr,character' distr(object) <- value ## S4 method for signature 'a4aGr' vcov(object) ## S4 replacement method for signature 'a4aGr,numeric' vcov(object) <- value
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
value |
the new object |
grModthe formula for the growth model, e.g. von Bertallanffy
grInvModthe formula for the inverse of the growth model, having length as the independent variable
paramsan FLPar object with the parameters of the model; must match the equations in the models
vcovan array with the variance covariance matrix of the parameters
distra character with the parameters' statistical distribution; it must match a known distribution for R (e.g. "norm" for gaussian), so that rnorm can be called
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm")mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm")
The advanced user interface to the a4a fitting routine.
a4aInternal( stock, indices, fmodel = defaultFmod(stock), qmodel = defaultQmod(indices), srmodel = defaultSRmod(stock), n1model = defaultN1mod(stock), vmodel = defaultVmod(stock, indices), covar = missing, wkdir = missing, verbose = FALSE, fit = "assessment", center = TRUE, mcmc = missing )a4aInternal( stock, indices, fmodel = defaultFmod(stock), qmodel = defaultQmod(indices), srmodel = defaultSRmod(stock), n1model = defaultN1mod(stock), vmodel = defaultVmod(stock, indices), covar = missing, wkdir = missing, verbose = FALSE, fit = "assessment", center = TRUE, mcmc = missing )
stock |
an FLStock object containing catch and stock information |
indices |
an FLIndices object containing survey indices |
fmodel |
a formula object depicting the model for log fishing mortality at age |
qmodel |
a list of formula objects depicting the models for log survey catchability at age |
srmodel |
a formula object depicting the model for log recruitment |
n1model |
a formula object depicting the model for the first year of catch data |
vmodel |
a list of formula objects depicting the models for log survey and log fishing mortality variance |
covar |
a list with covariates |
wkdir |
used to set a working directory for the admb optimiser. If wkdir is set all admb files are saved to this folder otherwise they are deleted. |
verbose |
if true admb fitting information is printed to the screen |
fit |
character with type of fit: 'MP' or 'assessment'; the former doesn't require the hessian to be computed, while the latter does. |
center |
|
mcmc |
|
an a4aFit object if fit is "MP" or an a4aFitSA if fit is "assessment"
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
a4aM(object, ...) ## S4 method for signature 'missing' a4aM(object, ...) ## S4 method for signature 'a4aM' show(object) shape(object, ...) ## S4 method for signature 'a4aM' shape(object) shape(object) <- value ## S4 replacement method for signature 'a4aM' shape(object) <- value level(object, ...) ## S4 method for signature 'a4aM' level(object) level(object) <- value ## S4 replacement method for signature 'a4aM' level(object) <- value trend(object, ...) ## S4 method for signature 'a4aM' trend(object) trend(object) <- value ## S4 replacement method for signature 'a4aM' trend(object) <- valuea4aM(object, ...) ## S4 method for signature 'missing' a4aM(object, ...) ## S4 method for signature 'a4aM' show(object) shape(object, ...) ## S4 method for signature 'a4aM' shape(object) shape(object) <- value ## S4 replacement method for signature 'a4aM' shape(object) <- value level(object, ...) ## S4 method for signature 'a4aM' level(object) level(object) <- value ## S4 replacement method for signature 'a4aM' level(object) <- value trend(object, ...) ## S4 method for signature 'a4aM' trend(object) trend(object) <- value ## S4 replacement method for signature 'a4aM' trend(object) <- value
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
value |
the new object |
shapethe shape of M by age
levelthe mean level of M over a range of ages, which will be used to scale the shape
trendthe yearly trend in M
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~1.5*k, params=FLPar(k=0.4)) m1 <- a4aM(shape=mod1, level=mod2)mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~1.5*k, params=FLPar(k=0.4)) m1 <- a4aM(shape=mod1, level=mod2)
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
a4aStkParams(object, ...) ## S4 method for signature 'missing' a4aStkParams(object, ...) ## S4 method for signature 'a4aStkParams' m(object) ## S4 method for signature 'a4aStkParams' wt(object) ## S4 method for signature 'a4aStkParams' mat(object) fMod(object, ...) ## S4 method for signature 'a4aStkParams' fMod(object) fMod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' fMod(object) <- value n1Mod(object, ...) ## S4 method for signature 'a4aStkParams' n1Mod(object) n1Mod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' n1Mod(object) <- value srMod(object, ...) ## S4 method for signature 'a4aStkParams' srMod(object) srMod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' srMod(object) <- value ## S4 method for signature 'a4aStkParams' params(object) ## S4 replacement method for signature 'a4aStkParams,FLPar' params(object) <- value coefficients(object, ...) ## S4 method for signature 'a4aStkParams' coefficients(object) coefficients(object) <- value ## S4 replacement method for signature 'a4aStkParams,FLPar' coefficients(object) <- value ## S4 method for signature 'a4aStkParams' distr(object) ## S4 replacement method for signature 'a4aStkParams,character' distr(object) <- value ## S4 method for signature 'a4aStkParams' vcov(object) ## S4 replacement method for signature 'a4aStkParams,array' vcov(object) <- value ## S4 method for signature 'a4aStkParams' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'a4aStkParams' iter(obj, it)a4aStkParams(object, ...) ## S4 method for signature 'missing' a4aStkParams(object, ...) ## S4 method for signature 'a4aStkParams' m(object) ## S4 method for signature 'a4aStkParams' wt(object) ## S4 method for signature 'a4aStkParams' mat(object) fMod(object, ...) ## S4 method for signature 'a4aStkParams' fMod(object) fMod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' fMod(object) <- value n1Mod(object, ...) ## S4 method for signature 'a4aStkParams' n1Mod(object) n1Mod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' n1Mod(object) <- value srMod(object, ...) ## S4 method for signature 'a4aStkParams' srMod(object) srMod(object) <- value ## S4 replacement method for signature 'a4aStkParams,formula' srMod(object) <- value ## S4 method for signature 'a4aStkParams' params(object) ## S4 replacement method for signature 'a4aStkParams,FLPar' params(object) <- value coefficients(object, ...) ## S4 method for signature 'a4aStkParams' coefficients(object) coefficients(object) <- value ## S4 replacement method for signature 'a4aStkParams,FLPar' coefficients(object) <- value ## S4 method for signature 'a4aStkParams' distr(object) ## S4 replacement method for signature 'a4aStkParams,character' distr(object) <- value ## S4 method for signature 'a4aStkParams' vcov(object) ## S4 replacement method for signature 'a4aStkParams,array' vcov(object) <- value ## S4 method for signature 'a4aStkParams' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'a4aStkParams' iter(obj, it)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
value |
the new object |
iter |
the number of iterations to create |
fill.iter |
should the new iterations be filled with values (TRUE) or NAs (FALSE) |
obj |
the object to be subset |
it |
iteration to be extracted |
fModF submodel formula
n1Modfirst year N formula
srModstock-recruitment submodel formula
paramsFLPar with parameters
vcovarray with variance-covariance
centeringcentering values numeric
distrstatistical distribution character
mnatural mortality FLQuant
unitsdata units character
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
Update FLStock and FLIndex objects with stock assessment results.
## S4 method for signature 'FLStock,a4aFit' e1 + e2 ## S4 method for signature 'FLIndices,a4aFit' e1 + e2## S4 method for signature 'FLStock,a4aFit' e1 + e2 ## S4 method for signature 'FLIndices,a4aFit' e1 + e2
e1 |
the original |
e2 |
a |
If both objects have the same number of iterations, the FLStock slots will be replaced by the a4aFit slots, in the case of 1 iter, or a4aFitSA slots, in the case of n iters. If one of the objects has 1 iter and the other n, the method will simulate using the fit results from the a4aFitSA object to update the slots of the FLStock object.
Assorted methods needed by FLa4a
Assorted methods needed by FLa4a
getYidx(object, ...) ## S4 method for signature 'FLQuant' getYidx(object, year) is.empty(object) niters(object, ...) ## S4 method for signature 'FLModelSim' niters(object) ## S4 method for signature 'a4aGr' niters(object) ## S4 method for signature 'a4aStkParams' dims(obj) replaceZeros(object, ...) ## S4 method for signature 'FLQuant' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLStock' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLIndex' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLIndices' replaceZeros(object, fraction = 0.25)getYidx(object, ...) ## S4 method for signature 'FLQuant' getYidx(object, year) is.empty(object) niters(object, ...) ## S4 method for signature 'FLModelSim' niters(object) ## S4 method for signature 'a4aGr' niters(object) ## S4 method for signature 'a4aStkParams' dims(obj) replaceZeros(object, ...) ## S4 method for signature 'FLQuant' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLStock' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLIndex' replaceZeros(object, fraction = 0.25) ## S4 method for signature 'FLIndices' replaceZeros(object, fraction = 0.25)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
year |
|
obj |
an object |
fraction |
|
Gets an FLQuant's numeric id for a vector of "years". For internal use and not very interesting for users. It takes an FLQuant object and vector of years and returns a numeric vector that can be used to subset the FLQuant.
Method is.empty checks if an object is empty. It takes any object and returns a logical, TRUE, if the object is of length 0.
Compute number of iterations. Takes an object of any FLR class and returns a numeric.
Extracts the dims of the parameters.
Replaces observations of 0 by a fraction of the minimum observed. It takes an FLQuant object and numeric of min fraction and returns a FLQuant with zeros replaced to be added to the FLStock or FLIndex objects.
#Example use of getYidx: data(ple4) flq <- catch(ple4) getYidx(flq, 2000:2004) flq[, getYidx(flq, 2000:2004)] #Example use of is.empty: is.empty(list()) is.empty(list(a=2)) #Example use of niters: mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") # Generate 100 sample sets vbObj <- mvrnorm(100,vbObj) niters(vbObj) #Example use of dims: dims(FLPar()) #Example use of getYidx: data(ple4) flq <- catch(ple4) flq <- replaceZeros(flq, 0.25) catch(ple4) <- flq#Example use of getYidx: data(ple4) flq <- catch(ple4) getYidx(flq, 2000:2004) flq[, getYidx(flq, 2000:2004)] #Example use of is.empty: is.empty(list()) is.empty(list(a=2)) #Example use of niters: mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") # Generate 100 sample sets vbObj <- mvrnorm(100,vbObj) niters(vbObj) #Example use of dims: dims(FLPar()) #Example use of getYidx: data(ple4) flq <- catch(ple4) flq <- replaceZeros(flq, 0.25) catch(ple4) <- flq
Method to set breakpoints in submodels
breakpts(var, ...) ## S4 method for signature 'numeric' breakpts(var, breaks, ...)breakpts(var, ...) ## S4 method for signature 'numeric' breakpts(var, breaks, ...)
var |
a |
... |
additional argument list that might never be used |
breaks |
a |
a factor with levels according to the defined breaks
Method to produce bubble plots of standardized residuals
## S4 method for signature 'a4aFitResiduals,missing' bubbles(x, data = missing, ...)## S4 method for signature 'a4aFitResiduals,missing' bubbles(x, data = missing, ...)
x |
an |
data |
ignored |
... |
additional argument list that might never be used |
a bubbles plot with stardardized log residuals
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) bubbles(flqs)data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) bubbles(flqs)
Method to collapse seasons of FLStock objects. M and catch-at-age are summed while mean weights at age, maturity at age and mortalities before spawning are averaged.
collapseSeasons(stock)collapseSeasons(stock)
stock |
an FLStock object |
a FLStock object
Methods to create formulas for sub-models. The sub-models are set automagically using defaults.
defaultFmod(stock, dfm = c(0.5, 0.7)) defaultQmod(indices, dfm = 0.6) defaultN1mod(stock) defaultVmod(stock, indices) defaultSRmod(stock)defaultFmod(stock, dfm = c(0.5, 0.7)) defaultQmod(indices, dfm = 0.6) defaultN1mod(stock) defaultVmod(stock, indices) defaultSRmod(stock)
stock |
an FLStock object |
dfm |
numeric vector with the data points fraction to be used to set the spline ks. |
indices |
an FLIndices object |
a FLStock object
Deprecated methods.
a4aSCA(...)a4aSCA(...)
... |
additional argument list that might never be used |
Methods to extract and replace the model coefficients.
formula(object) <- value ## S4 replacement method for signature 'submodel,formula' formula(object) <- value coef(object, ...) ## S4 method for signature 'a4aFitSA' coef(object) ## S4 method for signature 'SCAPars' coef(object) ## S4 method for signature 'a4aStkParams' coef(object) ## S4 method for signature 'submodels' coef(object) ## S4 method for signature 'submodel' coef(object) coef(object, ...) <- value ## S4 replacement method for signature 'a4aFitSA,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'SCAPars,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodels,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodel,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodel,FLPar' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,FLPar' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,matrix' coef(object, ...) <- valueformula(object) <- value ## S4 replacement method for signature 'submodel,formula' formula(object) <- value coef(object, ...) ## S4 method for signature 'a4aFitSA' coef(object) ## S4 method for signature 'SCAPars' coef(object) ## S4 method for signature 'a4aStkParams' coef(object) ## S4 method for signature 'submodels' coef(object) ## S4 method for signature 'submodel' coef(object) coef(object, ...) <- value ## S4 replacement method for signature 'a4aFitSA,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'SCAPars,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodels,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodel,numeric' coef(object, ...) <- value ## S4 replacement method for signature 'submodel,FLPar' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,FLPar' coef(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,matrix' coef(object, ...) <- value
object |
object of relevant class (see signature of method) |
value |
the new object |
... |
additional argument list that might never be used |
This method produces an FLIndex object by using the genFLQuant method.
genFLIndex(object, ...) ## S4 method for signature 'FLQuant' genFLIndex(object, cv = 0.2, niter = 250)genFLIndex(object, ...) ## S4 method for signature 'FLQuant' genFLIndex(object, cv = 0.2, niter = 250)
object |
an |
... |
additional argument list that might not ever be used. |
cv |
the coefficient of variation |
niter |
the number of iterations to be generated |
an FLIndex
This method uses the quant log-correlation matrix of the FLQuant object and generates a new FLQuant using a lognormal multivariate distribution.
genFLQuant(object, ...) ## S4 method for signature 'FLQuant' genFLQuant(object, cv = 0.2, method = "ac", niter = 250) ## S4 method for signature 'submodel' genFLQuant(object, type = c("link", "response"), nsim = 0, seed = NULL) ## S4 method for signature 'submodels' genFLQuant(object, type = c("link", "response"), nsim = 0, seed = NULL) ## S4 method for signature 'a4aStkParams' genFLQuant( object, type = c("link", "response"), nsim = 0, seed = NULL, simulate.recruitment = FALSE )genFLQuant(object, ...) ## S4 method for signature 'FLQuant' genFLQuant(object, cv = 0.2, method = "ac", niter = 250) ## S4 method for signature 'submodel' genFLQuant(object, type = c("link", "response"), nsim = 0, seed = NULL) ## S4 method for signature 'submodels' genFLQuant(object, type = c("link", "response"), nsim = 0, seed = NULL) ## S4 method for signature 'a4aStkParams' genFLQuant( object, type = c("link", "response"), nsim = 0, seed = NULL, simulate.recruitment = FALSE )
object |
an FLQuant |
... |
additional argument list that might not ever be used. |
cv |
the coefficient of variation |
method |
the method used to compute the correlation matrix; for now only "ac" (autocorrelation) is implemented |
niter |
the number of iterations to be generated |
type |
the type of output required. The default is on the scale of the linear predictors (link); the alternative "response" is on the scale of the response variable. Thus for a model on the log scale the default predictions are of log F (for example) and type = "response" gives the predicted F. |
nsim |
the number of iterations to simulate, if nsim = 0, then deterministic values are returned based on the coefficients. If nsim > 0 then coefficients are simluated using the covariance slot and distribution slot. |
seed |
if supplied the random numbers are generate with a fixed seed for repeatablility |
simulate.recruitment |
if FALSE (default) recruitment is simulated from the recruitment estimates of recruitment, which may or may not be based on a stock-recruit model in the origional fit. If TRUE, then new recruitments are simulated based on the stock recruitment model and supplied CV used in the fit, rsulting in a completly different timeseries of N and Catches. |
an FLQuant
data(ple4) sim.F <- genFLQuant(harvest(ple4))data(ple4) sim.F <- genFLQuant(harvest(ple4))
This method computes the FLStock slots consistently with the information provided by the FLQuant. It requires two of the triplet R/C/F to compute the third consistent with Baranov and survival's equations.
genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,FLQuant,FLQuant,missing' genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,missing,FLQuant,FLQuant' genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,FLQuant,missing,FLQuant' genFLStock(object, R, C, F, ...)genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,FLQuant,FLQuant,missing' genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,missing,FLQuant,FLQuant' genFLStock(object, R, C, F, ...) ## S4 method for signature 'FLStock,FLQuant,missing,FLQuant' genFLStock(object, R, C, F, ...)
object |
an FLStock |
R |
an FLQuant with iterations or missing |
C |
an FLQuant with iterations or missing |
F |
an FLQuant with iterations or missing |
... |
additional argument list that might not ever be used. |
an FLStock
Method to compute the log-correlation matrix for the first dimension (quant) of the FLQuant object.
getAcor(object, ...) ## S4 method for signature 'FLQuant' getAcor(object, ...)getAcor(object, ...) ## S4 method for signature 'FLQuant' getAcor(object, ...)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
an FLQuant object with a quant log-correlation matrix
data(ple4) getAcor(harvest(ple4))data(ple4) getAcor(harvest(ple4))
Reads the hessian file from any ADMB fit. Used here with the a4a model.
getADMBHessian(wkdir) getADMBCovariance(wkdir)getADMBHessian(wkdir) getADMBCovariance(wkdir)
wkdir |
the location of the admb output |
a list with the following elements
getADMBHessian is intended to be used internally
# load some data data(ple4) data(ple4.indices) # choose a working directory wkdir <- tempfile() # do an 'assessment' fit with default settings (not recomended!) and keep results in wkdir fit <- sca(stock=ple4,indices=ple4.indices,wkdir=wkdir) hessInfo <- getADMBHessian(wkdir) str(hessInfo) # calculate covariance matrix Sigma <- solve(hessInfo$hes)# load some data data(ple4) data(ple4.indices) # choose a working directory wkdir <- tempfile() # do an 'assessment' fit with default settings (not recomended!) and keep results in wkdir fit <- sca(stock=ple4,indices=ple4.indices,wkdir=wkdir) hessInfo <- getADMBHessian(wkdir) str(hessInfo) # calculate covariance matrix Sigma <- solve(hessInfo$hes)
Returns the covariance matrix of the specified Gaussian markov random field model.
getCov(n, model, tau)getCov(n, model, tau)
n |
integer giving the size of the random feild |
model |
chatacter giving the name of the GMRF |
tau |
numeric giving the multiplier of the structure matrix for the model |
a covariance matrix
Method to get values of the growth parameter K
getK(object, ...) ## S4 method for signature 'a4aGr' getK(object)getK(object, ...) ## S4 method for signature 'a4aGr' getK(object)
object |
an |
... |
additional argument list that might never be used |
a vector with K values
mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") vbObj <- mvrnorm(100,vbObj) getK(vbObj)mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") vbObj <- mvrnorm(100,vbObj) getK(vbObj)
Function to get the a4a TPL file with ADMB code and copy into a specific folder.
getTPL(dir)getTPL(dir)
dir |
folder where the a4a.tpl file will be copied to. |
file a4a.tpl
getTPL("myfolder")getTPL("myfolder")
Uses the user-specified formula to build a model matrix.
getX(object, ...) ## S4 method for signature 'formula' getX(object, df, newdf = df)getX(object, ...) ## S4 method for signature 'formula' getX(object, df, newdf = df)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
df |
the data.frame to build the model matrix against. |
newdf |
the data.frame to create the model matrix for. |
a matrix.
getX is intended to be used internally
Catch number, stocks weights, etc. for Gulf of Lions Hake (1998 - 2011).
hakeGSA7hakeGSA7
an FLStock object
Chato Osio
GFCM - STECF
Survey index for the Gulf of Lions Hake stock.
hakeGSA7.idxhakeGSA7.idx
an FLIndices object
Chato Osio
GFCM - STECF - MEDITS
Survey abundance index for hake in the Gulf of Cadiz. Lenth frequencies.
index_cd_lenindex_cd_len
an FLIndex object
Santiago Cervi\~no
Survey abundance index for hake in Portuguese continental waters. Lenth frequencies.
index_pt_lenindex_pt_len
an FLIndex object
Santiago Cervi\~no
Survey abundance index for hake in Northwest Spanish waters. Lenth frequencies.
index_sp_lenindex_sp_len
an FLIndex object
Santiago Cervi\~no
Method to convert length-based data to age-based
l2a(object, model, ...) ## S4 method for signature 'FLQuant,a4aGr' l2a( object, model, halfwidth = c(diff(as.numeric(dimnames(object)[[1]])), tail(diff(as.numeric(dimnames(object)[[1]])), 1))/2, stat = "sum", max_age = NA ) ## S4 method for signature 'FLStockLen,a4aGr' l2a(object, model, plusgroup = NA, ...) ## S4 method for signature 'FLIndex,a4aGr' l2a(object, model, ...)l2a(object, model, ...) ## S4 method for signature 'FLQuant,a4aGr' l2a( object, model, halfwidth = c(diff(as.numeric(dimnames(object)[[1]])), tail(diff(as.numeric(dimnames(object)[[1]])), 1))/2, stat = "sum", max_age = NA ) ## S4 method for signature 'FLStockLen,a4aGr' l2a(object, model, plusgroup = NA, ...) ## S4 method for signature 'FLIndex,a4aGr' l2a(object, model, ...)
object |
an |
model |
an |
... |
additional argument list that might never be used |
halfwidth |
the halfwidths of the length classes; a single numeric or vector the size of the number of length classes; not used if object is an |
stat |
the aggregation statistic, which must be |
max_age |
the maximum age in the returned |
plusgroup |
the plusgroup of the stock; only used if the object is an |
A deterministic slicing method converts the length-based data to age-based data, using the supplied growth model (the a4aGr object).
Each length-based observation is allocated to a corresponding age, based on the growth model, and aggregated accordingly (either the sum or the mean).
There should be 1 or n iterations in both the object being sliced and the growth model.
This means that although the slicing method is deterministic, the length-based data is sliced by each iteration of the growth parameters, thereby propagating the uncertainty in the biological growth parameters (representing process uncertainty) through to the age-based data.
an age based FLQuant, FLStock
# Southern hake # Variance-covariance matrix for parameters mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(2310, 0.13, 0.84) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(-7.22,-6.28,0.08) # Make the von Bertalanffy growth model md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=130, k=0.164, t0=-0.092, units=c("cm","yr-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") # Make a triangle copula for simulating process error linf <- list(a=104.5, b=155.5, c=130) k <- list(a=0.132, b=0.196, c=0.164) t0 <- list(a=-0.184, b=0, c=-0.092) tri_pars <- list(linf = linf, k = k, t0 = t0) # Simulate 10 iterations from it vbObj_tri <- mvrtriangle(10, vbObj, paramMargins=tri_pars) data(southernHakeLen) # Extract the catch numbers at length from stock object cth <- catch.n(shake_len) # Slice the data using the unsimulated growth object # so the stock and the growth object have 1 iteration cthA1 <- l2a(cth, vbObj) # Slice with 1 iteration in stock and multiple in growth object cthA2 <- l2a(cth, vbObj_tri) # Result is age based catch with multiple iterations # mod: iter=1, data: iter=n cthA3 <- l2a(propagate(cth,10), vbObj) # both with iter=n cthA4 <- l2a(propagate(cth,10), vbObj_tri) # converting a stock object shake_age <- l2a(shake_len, vbObj) shake_age <- l2a(shake_len, vbObj_tri) shake_age <- l2a(propagate(shake_len, 10), vbObj) shake_age <- l2a(propagate(shake_len, 10), vbObj_tri) # converting a index object index_pt_age <- l2a(index_pt_len, vbObj) index_pt_age <- l2a(index_pt_len, mvrnorm(10, vbObj)) index_pt_age <- l2a(propagate(index_pt_len, 10), vbObj)# Southern hake # Variance-covariance matrix for parameters mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(2310, 0.13, 0.84) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(-7.22,-6.28,0.08) # Make the von Bertalanffy growth model md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=130, k=0.164, t0=-0.092, units=c("cm","yr-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") # Make a triangle copula for simulating process error linf <- list(a=104.5, b=155.5, c=130) k <- list(a=0.132, b=0.196, c=0.164) t0 <- list(a=-0.184, b=0, c=-0.092) tri_pars <- list(linf = linf, k = k, t0 = t0) # Simulate 10 iterations from it vbObj_tri <- mvrtriangle(10, vbObj, paramMargins=tri_pars) data(southernHakeLen) # Extract the catch numbers at length from stock object cth <- catch.n(shake_len) # Slice the data using the unsimulated growth object # so the stock and the growth object have 1 iteration cthA1 <- l2a(cth, vbObj) # Slice with 1 iteration in stock and multiple in growth object cthA2 <- l2a(cth, vbObj_tri) # Result is age based catch with multiple iterations # mod: iter=1, data: iter=n cthA3 <- l2a(propagate(cth,10), vbObj) # both with iter=n cthA4 <- l2a(propagate(cth,10), vbObj_tri) # converting a stock object shake_age <- l2a(shake_len, vbObj) shake_age <- l2a(shake_len, vbObj_tri) shake_age <- l2a(propagate(shake_len, 10), vbObj) shake_age <- l2a(propagate(shake_len, 10), vbObj_tri) # converting a index object index_pt_age <- l2a(index_pt_len, vbObj) index_pt_age <- l2a(index_pt_len, mvrnorm(10, vbObj)) index_pt_age <- l2a(propagate(index_pt_len, 10), vbObj)
Method to compute natural mortality.
## S4 method for signature 'a4aM' m(object, grMod = "missing", ...)## S4 method for signature 'a4aM' m(object, grMod = "missing", ...)
object |
a |
grMod |
a |
... |
placeholder for covariates of the models. The names must match formula variables (not parameters), with the exception of the |
The method uses the range slot to define the quant and year dimensions of the resulting M FLQuant. The name for the quant dimension is taken as the name of a variable that is present in the shape formula, but not in the params slot of the shape model. If more than one such variable exists, then there is a problem with the shape model definition.
an FLQuant object
age <- 0:15 k <- 0.4 shp <- eval(as.list(~exp(-age-0.5))[[2]], envir=list(age=age)) lvl <- eval(as.list(~1.5*k)[[2]], envir=list(k=k)) M <- shp*lvl/mean(shp) # Now set up an equivalent a4aM object mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~1.5*k, params=FLPar(k=0.4)) m1 <- a4aM(shape=mod1, level=mod2) # set up the age range for the object... range(m1, c("min", "max")) <- c(0,15) # ...and the age range for mbar range(m1, c("minmbar", "maxmbar")) <- c(0,15) m(m1) mean(m(m1)[ac(0:15)]) all.equal(M, c(m(m1))) # another example m range(m1, c("min", "max")) <- c(2,15) range(m1, c("minmbar", "maxmbar")) <- c(2,4) m(m1) mean(m(m1)[ac(2:4)]) # example with specified iters (i.e. not simulated from a statistical distribution)... mod2 <- FLModelSim(model=~k^0.66*t^0.57, params=FLPar(matrix(c(0.4,10,0.5,11), ncol=2, dimnames=list(params=c("k","t"), iter=1:2))), vcov=array(c(0.004,0.,0.,0.001,0.006,0.,0.,0.002), dim=c(2,2,2))) m2 <- a4aM(shape=mod1, level=mod2) range(m2, c("min", "max")) <- c(2,10) m(m2) # ...and with randomly generated iters (based on the medians for params(mod2) and vcov(mod2)) m3 <- a4aM(shape=mod1, level=mvrnorm(100, mod2)) range(m3, c("min", "max")) <- c(0,15) m(m3) # example with a trend mod3 <- FLModelSim(model=~1+b*v, params=FLPar(b=0.05)) mObj <- a4aM(shape=mod1, level=mvrnorm(100, mod2), trend=mod3, range=c(min=0,max=15,minyear=2000,maxyear=2003,minmbar=0,maxmbar=0)) m(mObj, v=1:4)age <- 0:15 k <- 0.4 shp <- eval(as.list(~exp(-age-0.5))[[2]], envir=list(age=age)) lvl <- eval(as.list(~1.5*k)[[2]], envir=list(k=k)) M <- shp*lvl/mean(shp) # Now set up an equivalent a4aM object mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~1.5*k, params=FLPar(k=0.4)) m1 <- a4aM(shape=mod1, level=mod2) # set up the age range for the object... range(m1, c("min", "max")) <- c(0,15) # ...and the age range for mbar range(m1, c("minmbar", "maxmbar")) <- c(0,15) m(m1) mean(m(m1)[ac(0:15)]) all.equal(M, c(m(m1))) # another example m range(m1, c("min", "max")) <- c(2,15) range(m1, c("minmbar", "maxmbar")) <- c(2,4) m(m1) mean(m(m1)[ac(2:4)]) # example with specified iters (i.e. not simulated from a statistical distribution)... mod2 <- FLModelSim(model=~k^0.66*t^0.57, params=FLPar(matrix(c(0.4,10,0.5,11), ncol=2, dimnames=list(params=c("k","t"), iter=1:2))), vcov=array(c(0.004,0.,0.,0.001,0.006,0.,0.,0.002), dim=c(2,2,2))) m2 <- a4aM(shape=mod1, level=mod2) range(m2, c("min", "max")) <- c(2,10) m(m2) # ...and with randomly generated iters (based on the medians for params(mod2) and vcov(mod2)) m3 <- a4aM(shape=mod1, level=mvrnorm(100, mod2)) range(m3, c("min", "max")) <- c(0,15) m(m3) # example with a trend mod3 <- FLModelSim(model=~1+b*v, params=FLPar(b=0.05)) mObj <- a4aM(shape=mod1, level=mvrnorm(100, mod2), trend=mod3, range=c(min=0,max=15,minyear=2000,maxyear=2003,minmbar=0,maxmbar=0)) m(mObj, v=1:4)
Method to average across a set of models. This is still experimental. Use with care.
ma(object, ...) ## S4 method for signature 'a4aFitSAs' ma(object, stock, FUN, nsim = 1000)ma(object, ...) ## S4 method for signature 'a4aFitSAs' ma(object, stock, FUN, nsim = 1000)
object |
an |
... |
additional argument list that might never be used |
stock |
a |
FUN |
a |
nsim |
a |
an FLStock object with iterations defined by nsim
data(ple4) data(ple4.indices) fmod <- ~ factor(age) + s(year, k=20) qmod <- c(list(~ s(age, k = 4)), rep(list(~s(age, k=4)), 5)) f1 <- sca(ple4, ple4.indices, fmodel=fmod, qmodel=qmod, fit = "assessment") qmod <- c(list(~ s(age, k = 4) + year), rep(list(~s(age, k=4)), 5)) f2 <- sca(ple4, ple4.indices, fmodel=fmod, qmodel=qmod, fit = "assessment") # AIC weighting aicwt <- function(object){ ICs <- -1 * sapply(object, AIC) exp( 0.5 * (ICs - max(ICs))) } stock.sim <- ma(a4aFitSAs(list(f1=f1, f2=f2)), ple4, aicwt, nsim = 100) # equal weighting eqwt <- function(object){ v <- rep(1, length(object)) names(v) <- names(object) v } stock.sim <- ma(a4aFitSAs(list(f1=f1, f2=f2)), ple4, eqwt, nsim = 100)data(ple4) data(ple4.indices) fmod <- ~ factor(age) + s(year, k=20) qmod <- c(list(~ s(age, k = 4)), rep(list(~s(age, k=4)), 5)) f1 <- sca(ple4, ple4.indices, fmodel=fmod, qmodel=qmod, fit = "assessment") qmod <- c(list(~ s(age, k = 4) + year), rep(list(~s(age, k=4)), 5)) f2 <- sca(ple4, ple4.indices, fmodel=fmod, qmodel=qmod, fit = "assessment") # AIC weighting aicwt <- function(object){ ICs <- -1 * sapply(object, AIC) exp( 0.5 * (ICs - max(ICs))) } stock.sim <- ma(a4aFitSAs(list(f1=f1, f2=f2)), ple4, aicwt, nsim = 100) # equal weighting eqwt <- function(object){ v <- rep(1, length(object)) names(v) <- names(object) v } stock.sim <- ma(a4aFitSAs(list(f1=f1, f2=f2)), ple4, eqwt, nsim = 100)
Internal method to run several stock assessment fits with different stocks, indices and submodels
multisca( stocks, indicess, fmodel = missing, qmodel = missing, srmodel = missing, n1model = missing, vmodel = missing, combination.all = FALSE, ... )multisca( stocks, indicess, fmodel = missing, qmodel = missing, srmodel = missing, n1model = missing, vmodel = missing, combination.all = FALSE, ... )
fmodel |
a list of |
qmodel |
a list of |
srmodel |
a list of |
n1model |
a list of |
vmodel |
a list of |
combination.all |
bolean parameter (default is FALSE) to define if a full factorial across all stocks, indices, and submodel is run or just a sequence of runs. |
... |
all other arguments to be passed to |
stock |
an |
an a4aFits or a4aFitSAs or a4aFitMCMCs depending on the argument fit
Method to simulate multivariate normal parameters for an a4aM object.
## S4 method for signature 'numeric,a4aM,missing,missing,missing,missing' mvrnorm(n = 1, mu)## S4 method for signature 'numeric,a4aM,missing,missing,missing,missing' mvrnorm(n = 1, mu)
n |
the number of iterations to be generated |
mu |
an |
an a4aM object with n iterations
mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~k^0.66*t^0.57, params=FLPar(matrix(c(0.4,10,0.5,11), ncol=2, dimnames=list(params=c("k","t"), iter=1:2))), vcov=array(c(0.004,0.,0.,0.001,0.006,0.,0.,0.003), dim=c(2,2,2))) mod3 <- FLModelSim(model=~1+b*v, params=FLPar(b=0.05)) mObj <- a4aM(shape=mod1, level=mod2, trend=mod3, range=c(min=0,max=15,minyear=2000,maxyear=2003,minmbar=0,maxmbar=0)) mObj <- mvrnorm(100, mObj) # Generate 100 iterations with no trend over time m(mObj, v=c(1,1,1,1)) # Generate replicates based on iteration-specific multivariate distributions # (as defined by params() and vcov()) params(mod2) vcov(mod2) m1<-mvrnorm(mod2) c(params(m1)) # Generate replicates based on a single multivariate distribution (here the # median of params() and vcov() is used) mvrnorm(2,mod2) m2<-mvrnorm(2,mod2) c(params(m2))mod1 <- FLModelSim(model=~exp(-age-0.5)) mod2 <- FLModelSim(model=~k^0.66*t^0.57, params=FLPar(matrix(c(0.4,10,0.5,11), ncol=2, dimnames=list(params=c("k","t"), iter=1:2))), vcov=array(c(0.004,0.,0.,0.001,0.006,0.,0.,0.003), dim=c(2,2,2))) mod3 <- FLModelSim(model=~1+b*v, params=FLPar(b=0.05)) mObj <- a4aM(shape=mod1, level=mod2, trend=mod3, range=c(min=0,max=15,minyear=2000,maxyear=2003,minmbar=0,maxmbar=0)) mObj <- mvrnorm(100, mObj) # Generate 100 iterations with no trend over time m(mObj, v=c(1,1,1,1)) # Generate replicates based on iteration-specific multivariate distributions # (as defined by params() and vcov()) params(mod2) vcov(mod2) m1<-mvrnorm(mod2) c(params(m1)) # Generate replicates based on a single multivariate distribution (here the # median of params() and vcov() is used) mvrnorm(2,mod2) m2<-mvrnorm(2,mod2) c(params(m2))
Simulates model parameters with user-defined copulas and marginals.
mvrcop(n, mvdc, ...) ## S4 method for signature 'numeric,FLModelSim' mvrcop(n, mvdc, copula, ...)mvrcop(n, mvdc, ...) ## S4 method for signature 'numeric,FLModelSim' mvrcop(n, mvdc, copula, ...)
n |
the number of iterations |
mvdc |
an |
... |
arguments to be passed to the copula methods |
copula |
the name of the copula to be used |
an FLModelSim object with n groups of parameters
mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(100, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.1,0.0003) md <- ~linf*(1-exp(-k*(t-t0))) prs <- FLPar(linf=120, k=0.3, t0=0.1, units=c("cm","yr^-1","yr")) vb <- FLModelSim(model=md, params=prs, vcov=mm, distr="norm") pars <- list(list(a=90, b=125, c=120), list(a=0.2, b=0.4), list(a=0, b=0.4, c=0.1)) vbSim <- mvrcop(10000, vb, copula="archmCopula", family="clayton", param=2, margins="triangle", paramMargins=pars) boxplot(t(predict(vbSim, t=0:20+0.5))) splom(data.frame(t(params(vbSim)@.Data)), pch=".")mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(100, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.1,0.0003) md <- ~linf*(1-exp(-k*(t-t0))) prs <- FLPar(linf=120, k=0.3, t0=0.1, units=c("cm","yr^-1","yr")) vb <- FLModelSim(model=md, params=prs, vcov=mm, distr="norm") pars <- list(list(a=90, b=125, c=120), list(a=0.2, b=0.4), list(a=0, b=0.4, c=0.1)) vbSim <- mvrcop(10000, vb, copula="archmCopula", family="clayton", param=2, margins="triangle", paramMargins=pars) boxplot(t(predict(vbSim, t=0:20+0.5))) splom(data.frame(t(params(vbSim)@.Data)), pch=".")
Method to generate multivariate parameters with user-defined copulas and marginals for a4aGr objects.
## S4 method for signature 'numeric,a4aGr' mvrcop(n = 1, mvdc, ...)## S4 method for signature 'numeric,a4aGr' mvrcop(n = 1, mvdc, ...)
n |
the number of iterations |
mvdc |
the |
... |
arguments to be passed to the rMvdc and copula methods |
an FLModelSim object with n groups of parameters
mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") pars <- list(list(a=50, b=100, c=58.5), list(a=0.06, b=0.2, c=0.086), list(a=0, b=0.005, c=0.001)) #In the following, the third, fourth and fifth arguments refer to the copula, # while the final two arguments refer to the marginal distributions: vbObj <- mvrcop(10000, vbObj, copula="archmCopula", family="clayton", param=2, margins="triangle", paramMargins=pars) splom(data.frame(t(params(vbObj)@.Data)), pch=".")mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") pars <- list(list(a=50, b=100, c=58.5), list(a=0.06, b=0.2, c=0.086), list(a=0, b=0.005, c=0.001)) #In the following, the third, fourth and fifth arguments refer to the copula, # while the final two arguments refer to the marginal distributions: vbObj <- mvrcop(10000, vbObj, copula="archmCopula", family="clayton", param=2, margins="triangle", paramMargins=pars) splom(data.frame(t(params(vbObj)@.Data)), pch=".")
Method to generate multivariate normal parameters for a4aGr objects.
## S4 method for signature 'numeric,a4aGr,ANY,ANY,ANY,ANY' mvrnorm(n = 1, mu)## S4 method for signature 'numeric,a4aGr,ANY,ANY,ANY,ANY' mvrnorm(n = 1, mu)
n |
the number of simulations to be generated |
mu |
an |
an a4aGr object with n iterations
mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") vbObj <- mvrnorm(100,vbObj)mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") vbObj <- mvrnorm(100,vbObj)
Simulates model parameters using elliptical copulas and triangular marginals.
mvrtriangle(n, object, ...) ## S4 method for signature 'numeric,FLModelSim' mvrtriangle(n = 1, object, ...)mvrtriangle(n, object, ...) ## S4 method for signature 'numeric,FLModelSim' mvrtriangle(n = 1, object, ...)
n |
the number of iterations |
object |
the |
... |
arguments to be passed to the rMvdc and copula methods |
an FLModelSim object with n sets of parameters
# Set up the FLModelSim object mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(100, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.1,0.0003) md <- ~linf*(1-exp(-k*(t-t0))) prs <- FLPar(linf=120, k=0.3, t0=0.1, units=c("cm","yr^-1","yr")) vb <- FLModelSim(model=md, params=prs, vcov=mm, distr="norm") # Simulate from a multivariate normal distribution... set.seed(1) vbSim <- mvrnorm(10000, vb) mm <- predict(vbSim, t=0:20+0.5) #...from a multivariate triangular distribution with default ranges (0.01 and # 0.99 quantiles for min and max using a normal distribution with mean from # params and sigma from vcov, and with the apex located at params)... set.seed(1) vbSim1 <- mvrtriangle(10000, vb) mm1 <- predict(vbSim1, t=0:20+0.5) #...and from a multivariate triangular distribution with specified ranges # (note if "c" is missing, it will take the average of "a" and "b") set.seed(1) pars <- list(list(a=90, b=125, c=120), list(a=0.2, b=0.4), list(a=0, b=0.4, c=0.1)) vbSim2 <- mvrtriangle(10000, vb, paramMargins=pars) mm2 <- predict(vbSim2, t=0:20+0.5) # Plot the results par(mfrow=c(3,1)) boxplot(t(mm), main="normal") boxplot(t(mm1), main="triangular") boxplot(t(mm2), main="triangular2") splom(data.frame(t(params(vbSim)@.Data)), pch=".") splom(data.frame(t(params(vbSim1)@.Data)), pch=".") splom(data.frame(t(params(vbSim2)@.Data)), pch=".")# Set up the FLModelSim object mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(100, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.1,0.0003) md <- ~linf*(1-exp(-k*(t-t0))) prs <- FLPar(linf=120, k=0.3, t0=0.1, units=c("cm","yr^-1","yr")) vb <- FLModelSim(model=md, params=prs, vcov=mm, distr="norm") # Simulate from a multivariate normal distribution... set.seed(1) vbSim <- mvrnorm(10000, vb) mm <- predict(vbSim, t=0:20+0.5) #...from a multivariate triangular distribution with default ranges (0.01 and # 0.99 quantiles for min and max using a normal distribution with mean from # params and sigma from vcov, and with the apex located at params)... set.seed(1) vbSim1 <- mvrtriangle(10000, vb) mm1 <- predict(vbSim1, t=0:20+0.5) #...and from a multivariate triangular distribution with specified ranges # (note if "c" is missing, it will take the average of "a" and "b") set.seed(1) pars <- list(list(a=90, b=125, c=120), list(a=0.2, b=0.4), list(a=0, b=0.4, c=0.1)) vbSim2 <- mvrtriangle(10000, vb, paramMargins=pars) mm2 <- predict(vbSim2, t=0:20+0.5) # Plot the results par(mfrow=c(3,1)) boxplot(t(mm), main="normal") boxplot(t(mm1), main="triangular") boxplot(t(mm2), main="triangular2") splom(data.frame(t(params(vbSim)@.Data)), pch=".") splom(data.frame(t(params(vbSim1)@.Data)), pch=".") splom(data.frame(t(params(vbSim2)@.Data)), pch=".")
Method to generate multivariate parameters with elliptical copulas and triangular marginals for a4aGr objects.
## S4 method for signature 'numeric,a4aGr' mvrtriangle(n = 1, object, ...)## S4 method for signature 'numeric,a4aGr' mvrtriangle(n = 1, object, ...)
n |
the number of iterations |
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
The method is essentially a special case of mvrcop, where the copula is of type "ellipCopula" and family "t", and where the marginals are triangular.
an a4aGr object with n iterations
# Set up the a4aGr object and parameters for the marginals mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") pars <- list(list(a=50, b=100, c=58.5), list(a=0.06, b=0.2, c=0.086), list(a=0, b=0.005, c=0.001)) # Note that mvrtriangle is a special case of mvrcop set.seed(1) vbObj1 <- mvrtriangle(10000, vbObj, paramMargins=pars, dispstr="ex", param=0) set.seed(1) vbObj2 <- mvrcop(10000, vbObj, copula="ellipCopula", family="t", param=0, margins="triangle", paramMargins=pars) all.equal(vbObj2, vbObj1)# Set up the a4aGr object and parameters for the marginals mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") pars <- list(list(a=50, b=100, c=58.5), list(a=0.06, b=0.2, c=0.086), list(a=0, b=0.005, c=0.001)) # Note that mvrtriangle is a special case of mvrcop set.seed(1) vbObj1 <- mvrtriangle(10000, vbObj, paramMargins=pars, dispstr="ex", param=0) set.seed(1) vbObj2 <- mvrcop(10000, vbObj, copula="ellipCopula", family="t", param=0, margins="triangle", paramMargins=pars) all.equal(vbObj2, vbObj1)
Checks that the name of the second dimension in params is "iter". For internal use, not very interesting for users. It takes a FLModelSim object and returns a logical.
pars2dim(object) ## S4 method for signature 'FLModelSim' pars2dim(object) ## S4 method for signature 'FLPar' pars2dim(object)pars2dim(object) ## S4 method for signature 'FLModelSim' pars2dim(object) ## S4 method for signature 'FLPar' pars2dim(object)
object |
object of relevant class (see signature of method) |
Checks that the name of the second dimension in params is "iter". For internal use and not very interesting for users. It takes an FLPar object and returns a logical.
pars2dim(FLModelSim()) #Example use of pars2dim: pars2dim(FLPar()) pars2dim(FLPar(array(dim=c(1,1,1))))pars2dim(FLModelSim()) #Example use of pars2dim: pars2dim(FLPar()) pars2dim(FLPar(array(dim=c(1,1,1))))
Method to plot fitted versus observed catch numbers-at-age. Note the yaxis doesn't has a scale. The visual is about the difference between the two lines, not about the value of each line, which in any case would be very difficult to assess visually.
## S4 method for signature 'a4aFit,FLStock' plot(x, y, ...)## S4 method for signature 'a4aFit,FLStock' plot(x, y, ...)
x |
an |
y |
an |
... |
additional argument list that might never be used |
a plot with fitted and observed catch numbers-at-age
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) plot(obj, ple4)data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) plot(obj, ple4)
Method to plot fitted versus observed indices-at-age. Note the yaxis doesn't has a scale. The visual is about the difference between the two lines, not about the value of each line, which in any case would be very difficult to assess visually.
## S4 method for signature 'a4aFit,FLIndices' plot(x, y, ...)## S4 method for signature 'a4aFit,FLIndices' plot(x, y, ...)
x |
an |
y |
an |
... |
additional argument list that might never be used |
a plot with fitted and observed indices-at-age
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) plot(obj, FLIndices(ple4.index))data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) plot(obj, FLIndices(ple4.index))
Method to produce scatterplots of standardized residuals
Method to produce scatterplots of standardized residuals
## S4 method for signature 'a4aFitResiduals,missing' plot(x, y = missing, auxline = "smooth", by = "year", ...) ## S4 method for signature 'a4aFitCatchDiagn,missing' plot(x, y = missing, ...)## S4 method for signature 'a4aFitResiduals,missing' plot(x, y = missing, auxline = "smooth", by = "year", ...) ## S4 method for signature 'a4aFitCatchDiagn,missing' plot(x, y = missing, ...)
x |
an |
y |
ignored |
auxline |
a string defining the type of line to be added, by default uses 'smooth', a common alternative is to use 'r', a regression, or leave it empty ” |
by |
a string defining the type of plot to be depicted. At the moment accepts only 'age', the default, or 'year'. |
... |
additional argument list that might never be used |
a plot with stardardized log residuals
a plot with stardardized log residuals
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) plot(flqs) data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) plot(flqs)data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) plot(flqs) data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) plot(flqs)
a4aGr
Predicts ages or lengths using a growth class
## S4 method for signature 'a4aGr' predict(object, ...)## S4 method for signature 'a4aGr' predict(object, ...)
object |
the |
... |
arguments to be passed to the rMvdc and copula methods |
a matrix object with lengths or ages
# Set up the a4aGr object and parameters for the marginals mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") predict(vbObj, len=1:50+0.5) predict(vbObj, t=1:20+0.5)# Set up the a4aGr object and parameters for the marginals mm <- matrix(NA, ncol=3, nrow=3) diag(mm) <- c(50, 0.001,0.001) mm[upper.tri(mm)] <- mm[lower.tri(mm)] <- c(0.1,0.01,0.00004) md <- ~linf*(1-exp(-k*(t-t0))) imd <- ~t0-1/k*log(1-len/linf) prs <- FLPar(linf=58.5, k=0.086, t0=0.001, units=c("cm","yr^-1","yr")) vbObj <- a4aGr(grMod=md, grInvMod=imd, params=prs, vcov=mm, distr="norm") predict(vbObj, len=1:50+0.5) predict(vbObj, t=1:20+0.5)
Predict methods for a4a stock assessment fits.
## S4 method for signature 'a4aFitSA' predict(object) ## S4 method for signature 'SCAPars' predict(object)## S4 method for signature 'a4aFitSA' predict(object) ## S4 method for signature 'SCAPars' predict(object)
object |
object of relevant class (see signature of method) |
data(ple4) data(ple4.index) fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) flqs <- predict(fit1)data(ple4) data(ple4.index) fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) flqs <- predict(fit1)
Method to produce qqplots of standardized residuals
## S4 method for signature 'a4aFitResiduals,missing' qqmath(x, data = missing, ...)## S4 method for signature 'a4aFitResiduals,missing' qqmath(x, data = missing, ...)
x |
an |
data |
ignored |
... |
additional argument list that might never be used |
a qqplot with stardardized log residuals
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) qqmath(flqs)data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) qqmath(flqs)
Range method for a4aM objects
## S4 replacement method for signature 'a4aM,ANY,numeric' range(x, i) <- value## S4 replacement method for signature 'a4aM,ANY,numeric' range(x, i) <- value
x |
an a4aM object |
i |
the elements of range to be changed in a character vector |
value |
a numeric vector with values |
Simulated length data for redfish. Simulations were done using GADGET.
data(rfLen)data(rfLen)
An FLStock.
Ernesto Jardim
Daniel Howell
Simulated stock based on red fish.
rfLen.stkrfLen.stk
an FLStock object
Daniel Howell
Trawl survey index for red fish.
rfTrawl.idxrfTrawl.idx
an FLIndex object
Daniel Howell
Trawl survey index for red fish, with a jump in catchability.
rfTrawlJmp.idxrfTrawlJmp.idx
an FLIndex object
Daniel Howell
Trawl survey index for red fish, with a trend in catchability.
rfTrawlTrd.idxrfTrawlTrd.idx
an FLIndex object
Daniel Howell
Statistical catch-at-age method of the a4a stock assessment framework.
sca(stock, indices, ...) ## S4 method for signature 'FLStock,FLIndex' sca(stock, indices, ...) ## S4 method for signature 'FLStock,FLIndices' sca( stock, indices, fmodel = missing, qmodel = missing, srmodel = missing, n1model = missing, vmodel = missing, covar = missing, wkdir = missing, verbose = FALSE, fit = "assessment", center = TRUE, mcmc = missing )sca(stock, indices, ...) ## S4 method for signature 'FLStock,FLIndex' sca(stock, indices, ...) ## S4 method for signature 'FLStock,FLIndices' sca( stock, indices, fmodel = missing, qmodel = missing, srmodel = missing, n1model = missing, vmodel = missing, covar = missing, wkdir = missing, verbose = FALSE, fit = "assessment", center = TRUE, mcmc = missing )
stock |
an |
indices |
an |
... |
additional argument list that might never be used |
fmodel |
a formula object depicting the model for log fishing mortality at age |
qmodel |
a list of formula objects depicting the models for log survey catchability at age |
srmodel |
a formula object depicting the model for log recruitment |
n1model |
a formula object depicting the model for the population in the first year of the time series |
vmodel |
a list of formula objects depicting the model for the variance of fishing mortality and the indices |
covar |
a list with covariates to be used by the submodels. The formula must have an element with the same name as the list element. |
wkdir |
used to set a working directory for the admb optimiser; if wkdir is set, all admb files are saved to this folder, otherwise they are deleted. |
verbose |
if true, admb fitting information is printed to the screen. |
fit |
character with type of fit: 'MP' or 'assessment'; the former does not require the hessian to be computed, while the latter does. |
center |
logical defining if the data should be centered before fitting. |
mcmc |
an |
[REQUIRES REVISION] This method is the advanced method for stock assessment, it gives the user access to a set of arguments that the sca method doesn't. In particular, the default for the fit argument is 'assessment'. For detailed information about using the sca read the vignette 'The a4a Stock Assessment Modelling Framework' (vignette('sca')).
an a4aFit object if fit is "MP" or an a4aFitSA object if fit is "assessment"
data(ple4) data(ple4.index) # fishing mortality by age and year (separable) AND catchability at age without year trend fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability at age without year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~factor(age)) fit2 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability as a smoother by age without year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~s(age, k=4)) fit3 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability as a smoother by age with year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~s(age, k=4) + year) fit4 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # It's a statistical model BIC(fit1, fit2, fit3, fit4) # fishing mortality as a smoother by age and year with interactions (i.e. non-separable) AND # catchability as a smoother by age without year trend fmodel <- ~ te(age, year, k=c(4, 10)) qmodel <- list(~s(age, k=4)) fit5 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fit3 + smoother in recruitment fmodel <- ~ s(age, k=4) + s(year, k=20) qmodel <- list(~s(age, k=4)) rmodel <- ~s(year, k=20) fit6 <- sca(fmodel=fmodel, qmodel=qmodel, srmodel=rmodel, ple4, FLIndices(ple4.index)) # fit3 + bevholt rmodel <- ~ bevholt(CV=0.05) fit7 <- sca(fmodel=fmodel, qmodel=qmodel, srmodel=rmodel, ple4, FLIndices(ple4.index))data(ple4) data(ple4.index) # fishing mortality by age and year (separable) AND catchability at age without year trend fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability at age without year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~factor(age)) fit2 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability as a smoother by age without year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~s(age, k=4)) fit3 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fishing mortality as a smoother by age and year (but still separable) AND # catchability as a smoother by age with year trend fmodel <- ~ s(age, k=4) + s(year, k=10) qmodel <- list(~s(age, k=4) + year) fit4 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # It's a statistical model BIC(fit1, fit2, fit3, fit4) # fishing mortality as a smoother by age and year with interactions (i.e. non-separable) AND # catchability as a smoother by age without year trend fmodel <- ~ te(age, year, k=c(4, 10)) qmodel <- list(~s(age, k=4)) fit5 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) # fit3 + smoother in recruitment fmodel <- ~ s(age, k=4) + s(year, k=20) qmodel <- list(~s(age, k=4)) rmodel <- ~s(year, k=20) fit6 <- sca(fmodel=fmodel, qmodel=qmodel, srmodel=rmodel, ple4, FLIndices(ple4.index)) # fit3 + bevholt rmodel <- ~ bevholt(CV=0.05) fit7 <- sca(fmodel=fmodel, qmodel=qmodel, srmodel=rmodel, ple4, FLIndices(ple4.index))
This function provides an interface to sca() to be used inside the mp() function of the mse package.
sca.sa(stk, idx, update = TRUE, dfm = c(0.75, 0.75), args, tracking, ...)sca.sa(stk, idx, update = TRUE, dfm = c(0.75, 0.75), args, tracking, ...)
stk |
The FLStock input object. |
idx |
The FLIndices input object. |
update |
Should the fmodel be updated with the default? |
dfm |
data points fraction to be used to set the spline ks. |
... |
Any other arguments to sca() |
genArgs |
The mse arguments used by mp(). |
A list containing the estimated stock (stk, of class FLStock), and the tracking FLQuant, including convergence flags.
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
SCAMCMC(object, ...) ## S4 method for signature 'missing' SCAMCMC(object, ...) getADMBCallArgs(object, ...) ## S4 method for signature 'SCAMCMC' getADMBCallArgs(object, ...) getN(object, ...) ## S4 method for signature 'SCAMCMC' getN(object, ...)SCAMCMC(object, ...) ## S4 method for signature 'missing' SCAMCMC(object, ...) getADMBCallArgs(object, ...) ## S4 method for signature 'SCAMCMC' getADMBCallArgs(object, ...) getN(object, ...) ## S4 method for signature 'SCAMCMC' getN(object, ...)
object |
a |
... |
extra arguments |
mcmc NRun N MCMC iterations
mcsave NSave every N th MCMC iteration
mcscale NRescale step size for first N iterations
mcmult NRescale the covariance matrix
mcrb NReduce high parameter correlations
mcprobe XUse a fat-tailed proposal distribution
mcdiagUse a diagonal covariance matrix
mcnoscaleDo not scale the algorithm during
mcuUse a uniform distribution as proposal distribution
hybridUse the hybrid method
hynstep NMean number of steps for the leapfrog method
hyeps XThe stepsize for the leapfrog method [X numeric and > 0]
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
SCAPars(object, ...) ## S4 method for signature 'missing' SCAPars(object, ...) stkmodel(object, ...) ## S4 method for signature 'SCAPars' stkmodel(object) n1model(object, ...) ## S4 method for signature 'SCAPars' n1model(object) srmodel(object, ...) ## S4 method for signature 'SCAPars' srmodel(object) fmodel(object, ...) ## S4 method for signature 'SCAPars' fmodel(object) qmodel(object, ...) ## S4 method for signature 'SCAPars' qmodel(object) qMod(object, ...) ## S4 method for signature 'SCAPars' qMod(object) vmodel(object, ...) ## S4 method for signature 'SCAPars' vmodel(object) vMod(object, ...) ## S4 method for signature 'SCAPars' vMod(object) srPars(object, ...) ## S4 method for signature 'SCAPars' srPars(object) srCovar(object, ...) ## S4 method for signature 'SCAPars' srCovar(object) srFrml(object, ...) ## S4 method for signature 'SCAPars' srFrml(object) fPars(object, ...) ## S4 method for signature 'SCAPars' fPars(object) fCovar(object, ...) ## S4 method for signature 'SCAPars' fCovar(object) fFrml(object, ...) ## S4 method for signature 'SCAPars' fFrml(object) qPars(object, ...) ## S4 method for signature 'SCAPars' qPars(object) qCovar(object, ...) ## S4 method for signature 'SCAPars' qCovar(object) qFrml(object, ...) ## S4 method for signature 'SCAPars' qFrml(object) vPars(object, ...) ## S4 method for signature 'SCAPars' vPars(object) vCovar(object, ...) ## S4 method for signature 'SCAPars' vCovar(object) vFrml(object, ...) ## S4 method for signature 'SCAPars' vFrml(object) ## S4 method for signature 'SCAPars' m(object) ## S4 method for signature 'SCAPars' wt(object) ## S4 method for signature 'SCAPars' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'SCAPars' iter(obj, it)SCAPars(object, ...) ## S4 method for signature 'missing' SCAPars(object, ...) stkmodel(object, ...) ## S4 method for signature 'SCAPars' stkmodel(object) n1model(object, ...) ## S4 method for signature 'SCAPars' n1model(object) srmodel(object, ...) ## S4 method for signature 'SCAPars' srmodel(object) fmodel(object, ...) ## S4 method for signature 'SCAPars' fmodel(object) qmodel(object, ...) ## S4 method for signature 'SCAPars' qmodel(object) qMod(object, ...) ## S4 method for signature 'SCAPars' qMod(object) vmodel(object, ...) ## S4 method for signature 'SCAPars' vmodel(object) vMod(object, ...) ## S4 method for signature 'SCAPars' vMod(object) srPars(object, ...) ## S4 method for signature 'SCAPars' srPars(object) srCovar(object, ...) ## S4 method for signature 'SCAPars' srCovar(object) srFrml(object, ...) ## S4 method for signature 'SCAPars' srFrml(object) fPars(object, ...) ## S4 method for signature 'SCAPars' fPars(object) fCovar(object, ...) ## S4 method for signature 'SCAPars' fCovar(object) fFrml(object, ...) ## S4 method for signature 'SCAPars' fFrml(object) qPars(object, ...) ## S4 method for signature 'SCAPars' qPars(object) qCovar(object, ...) ## S4 method for signature 'SCAPars' qCovar(object) qFrml(object, ...) ## S4 method for signature 'SCAPars' qFrml(object) vPars(object, ...) ## S4 method for signature 'SCAPars' vPars(object) vCovar(object, ...) ## S4 method for signature 'SCAPars' vCovar(object) vFrml(object, ...) ## S4 method for signature 'SCAPars' vFrml(object) ## S4 method for signature 'SCAPars' m(object) ## S4 method for signature 'SCAPars' wt(object) ## S4 method for signature 'SCAPars' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'SCAPars' iter(obj, it)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
iter |
the number of iterations to create |
fill.iter |
should the new iterations be filled with values (TRUE) or NAs (FALSE) |
obj |
the object to be subset |
it |
iteration to be extracted |
stkmodelparameters related to stock dynamics
qmodelparamaters related to catchability of tunning fleets
vmodelparamaters related to the variance model
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
This function provides an interface to a call to a separable model based on sca() to be used inside the mp() function of the mse package.
sep.sa(stk, idx, args, update = TRUE, dfm = c(0.75, 0.75), ...)sep.sa(stk, idx, args, update = TRUE, dfm = c(0.75, 0.75), ...)
stk |
The FLStock input object. |
idx |
The FLIndices input object. |
update |
Should the fmodel be updated with the default? |
dfm |
data points fraction to be used to set the spline ks. |
... |
Any other arguments to sca() |
genArgs |
The mse arguments used by mp(). |
A list containing the estimated stock (stk, of class FLStock), and the tracking FLQuant, including convergence flags.
FLR stock object for southern hake.
shake_lenshake_len
an FLStock object
Santiago Cervi\~no
Simulation methods for a4a stock assessment fits.
simulate(object, nsim = 1, seed = NULL, ...) ## S4 method for signature 'a4aFitSA' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'SCAPars' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'a4aStkParams' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'submodels' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'submodel' simulate(object, nsim = 1, seed = NULL, empirical = TRUE)simulate(object, nsim = 1, seed = NULL, ...) ## S4 method for signature 'a4aFitSA' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'SCAPars' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'a4aStkParams' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'submodels' simulate(object, nsim = 1, seed = NULL, empirical = TRUE) ## S4 method for signature 'submodel' simulate(object, nsim = 1, seed = NULL, empirical = TRUE)
object |
object of relevant class (see signature of method) |
nsim |
number of iterations |
seed |
|
... |
additional argument list that might never be used |
empirical |
logical, shall the empirical method in MASS be used |
data(ple4) data(ple4.index) fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) fit1 summary(fit1) stock.n(fit1)data(ple4) data(ple4.index) fmodel <- ~factor(age) + factor(year) qmodel <- list(~factor(age)) fit1 <- sca(fmodel=fmodel, qmodel=qmodel, stock=ple4, indices=FLIndices(ple4.index)) fit1 summary(fit1) stock.n(fit1)
Length based stock and three indices data for Southern hake.
data(southernHakeLen)data(southernHakeLen)
an FLStockLen and three FLIndex objects.
Finlay Scott
Santiago Cervino
Method to compute the standardized residuals on the log scale for index- and catch-at-age residuals in the a4a stock assessment framework.
stdlogres(obs, fit, ...) ## S4 method for signature 'FLQuant,FLQuant' stdlogres(obs, fit, ...)stdlogres(obs, fit, ...) ## S4 method for signature 'FLQuant,FLQuant' stdlogres(obs, fit, ...)
obs |
an |
fit |
an |
... |
additional argument list that might never be used |
an FLQuant with stardardized log residuals
data(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) stdlogres(catch.n(ple4), catch.n(obj)) # which is the same as the following (because residuals() uses stdlogres): flqs$catch.n # check: stdlogres(catch.n(ple4),catch.n(obj)) - flqs$catch.ndata(ple4) data(ple4.index) obj <- sca(ple4, FLIndices(ple4.index)) flqs <- residuals(obj, ple4, FLIndices(idx=ple4.index)) stdlogres(catch.n(ple4), catch.n(obj)) # which is the same as the following (because residuals() uses stdlogres): flqs$catch.n # check: stdlogres(catch.n(ple4),catch.n(obj)) - flqs$catch.n
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
submodel(object, ...) ## S4 method for signature 'missing' submodel(object, ...) ## S4 method for signature 'submodel' params(object) sMod(object, ...) ## S4 method for signature 'submodel' sMod(object) ## S4 method for signature 'submodel' iter(obj, it) ## S4 method for signature 'submodel' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'submodel' formula(x)submodel(object, ...) ## S4 method for signature 'missing' submodel(object, ...) ## S4 method for signature 'submodel' params(object) sMod(object, ...) ## S4 method for signature 'submodel' sMod(object) ## S4 method for signature 'submodel' iter(obj, it) ## S4 method for signature 'submodel' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'submodel' formula(x)
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
obj |
the object to be subset |
it |
iteration to be extracted |
iter |
the number of iterations to create |
fill.iter |
should the new iterations be filled with values (TRUE) or NAs (FALSE) |
x |
the submodel object that is to be modified |
Modformula describing the model
paramsFLPar with model parameters
vcovarray with variance covariance paramaters related to the variance model
centeringnumeric value used for centering the data
distra character with the parameters' statistical distribution; it must match a known distribution for R (e.g. "norm" for gaussian) so that rnorm can be called
All slots in the class have accessor and replacement methods defined that allow retrieving and substituting individual slots.
The values passed for replacement need to be of the class of that slot. A numeric vector can also be used when replacing FLQuant slots, and the vector will be used to substitute the values in the slot, but not its other attributes.
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
Class definition (slots), constructors, accessors, replacement (when relevant) and common methods.
submodels(...) submodels(...) corBlocks(object, ...) ## S4 method for signature 'submodels' corBlocks(object) ## S4 method for signature 'submodels' params(object) ## S4 method for signature 'submodels' sMod(object) ## S4 method for signature 'submodels' formula(x) corBlocks(object, ...) <- value ## S4 replacement method for signature 'submodels,list' corBlocks(object, ...) <- value ## S4 replacement method for signature 'submodels,submodel' x$name <- value ## S4 replacement method for signature 'submodels,character,missing' x[[i, j, ...]] <- value ## S4 replacement method for signature 'submodels,numeric,missing' x[[i, j, ...]] <- value ## S4 method for signature 'submodels' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'submodels' iter(obj, it)submodels(...) submodels(...) corBlocks(object, ...) ## S4 method for signature 'submodels' corBlocks(object) ## S4 method for signature 'submodels' params(object) ## S4 method for signature 'submodels' sMod(object) ## S4 method for signature 'submodels' formula(x) corBlocks(object, ...) <- value ## S4 replacement method for signature 'submodels,list' corBlocks(object, ...) <- value ## S4 replacement method for signature 'submodels,submodel' x$name <- value ## S4 replacement method for signature 'submodels,character,missing' x[[i, j, ...]] <- value ## S4 replacement method for signature 'submodels,numeric,missing' x[[i, j, ...]] <- value ## S4 method for signature 'submodels' propagate(object, iter, fill.iter = TRUE) ## S4 method for signature 'submodels' iter(obj, it)
... |
additional argument list that might never be used |
object |
object of relevant class (see signature of method) |
x |
object to be modified |
value |
value the new object |
name |
name(s) of entry to be extracted / modified |
i, j
|
indices specifying elements to extract or replace. |
iter |
the number of iterations to create |
fill.iter |
should the new iterations be filled with values (TRUE) or NAs (FALSE) |
obj |
the object to be subset |
it |
iteration to be extracted |
A construction method exists for this class that can take named arguments for
any of its slots. All slots are then created to match the requirements of the
class validity. If an unnamed FLQuant object is provided, this is used
for sizing, but not for populating any slot.
This class is similar to other 'plural' calsses in FLR. It is a list constrained to having all elements of the same class, in this case submodel. Otherwise it works exacly as any other list.
Methods to extract and replace the variance-covariance matrix.
## S4 method for signature 'a4aFitSA' vcov(object) ## S4 method for signature 'SCAPars' vcov(object) ## S4 method for signature 'submodels' vcov(object) ## S4 method for signature 'submodel' vcov(object) ## S4 replacement method for signature 'a4aFitSA,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'SCAPars,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,matrix' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,array' vcov(object, ...) <- value## S4 method for signature 'a4aFitSA' vcov(object) ## S4 method for signature 'SCAPars' vcov(object) ## S4 method for signature 'submodels' vcov(object) ## S4 method for signature 'submodel' vcov(object) ## S4 replacement method for signature 'a4aFitSA,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'SCAPars,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'a4aStkParams,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,numeric' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,matrix' vcov(object, ...) <- value ## S4 replacement method for signature 'submodel,array' vcov(object, ...) <- value
object |
object of relevant class (see signature of method) |
... |
additional argument list that might never be used |
value |
the new object |
Method to 3D plot FLQuant objects.
## S4 method for signature 'FLQuant,missing' wireframe(x, y, screen = list(x = -90, y = -45), ...)## S4 method for signature 'FLQuant,missing' wireframe(x, y, screen = list(x = -90, y = -45), ...)
x |
a |
y |
missing |
screen |
list with numeric components 'x','y' and 'z' to change the 3D perspective |
... |
additional argument list for the lattice engine |
a 3D surface plot
data(ple4) wireframe(harvest(ple4))data(ple4) wireframe(harvest(ple4))