Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Problem with catch scenarios for longer management periods #154

Open
alko989 opened this issue Mar 8, 2022 · 0 comments
Open

Problem with catch scenarios for longer management periods #154

alko989 opened this issue Mar 8, 2022 · 0 comments

Comments

@alko989
Copy link
Member

alko989 commented Mar 8, 2022

There are sometimes issues when the management period is longer than one year and Catch fractiles are used. The F/Fmsy at the end of the management period is higher than 1.

It seems to affect assessments with low biomass in the end of the time series.

A possible solution is to split Cp and dtcp to yearly chunks instead of a single long catch duration.

ReprEx

library(spict)
#> Loading required package: TMB
#> Welcome to spict_v1.3.5@8a4e1b
set.seed(1234)
nt <- 50
inp <- list(nseasons=4, splineorder=3)
s1 <- seq(0, 10, by=1)
s2 <- seq(11, nt-3-1/inp$nseasons, by=1/inp$nseasons)
s3 <- seq(nt-3, nt-1, by = 1)
inp$timeC <- c(s1,s2,s3)
inp$dtc <- c(rep(1, length(s1)),
             rep(0.25, length(s2)),
             rep(1, length(s3)))
inp$timeI <- seq(0.1, nt-1/inp$nseasons, by=0.5)
inp$ini <- list(logK=log(1000), logm=log(123), logq=log(1), logn=log(2),
                logbkfrac=log(0.7), logsdf=log(0.123), logF0=log(0.4321),
                logphi=log(c(0.3, 0.5, 1.8)))
inp$dteuler <- 1/8
inpsim <- sim.spict(inp)
#> The stabilise option was used for fitting / is specified in the input list (inp$stabilise). This activates six very vague priors. Acknolowdging these priors when simulating is not yet implemented. 
#> Additional priors were used for fitting / are specified in the input list (inp$priors): logn, logalpha, logbeta. Acknolowdging these priors when simulating is not yet implemented.
inpsim$dtc <- inp$dtc ## Very important, sim.spict drops it
inpsim$maneval <- nt+5
inpsim$maninterval <- c(nt,nt+5)

inpsim$priors$logbkfrac <- c(log(0.7), 0.001, 1)
inpsim$priors$logn <- c(log(2), 0.001, 1)
inpsim$dtc <- inp$dtc
inpsim <- check.inp(inpsim)

plotspict.data(inpsim)
#> Error in xy.coords(x, y, xlabel, ylabel, log): 'x' and 'y' lengths differ

fit <- fit.spict(inpsim)
fit <- add.man.scenario(fit, "Fmsy")
fit <- add.man.scenario(fit, "C35", fractiles = list(catch = 0.35))
fit <- add.man.scenario(fit, "All35", fractiles = list(catch = 0.35, ffmsy=0.35, bbmsy=0.35))
fit <- add.man.scenario(fit, "FB35_noC", fractiles = list( ffmsy=0.35, bbmsy=0.35))
plot2(fit) ## This does not, indxmax is not defined

sumspict.manage(fit)
#> SPiCT timeline:
#>                                                   
#>       Observations              Management        
#>      0.00 - 50.00            50.00 - 55.00        
#>  |-----------------------| ----------------------|
#> 
#> Management evaluation: 55.00
#> 
#> Predicted catch for management period and states at management evaluation time:
#> 
#>                     C B/Bmsy F/Fmsy
#> 1. Fish at Fmsy 166.1   0.27   1.00
#> 2. C35          148.1   0.02   2.53
#> 3. All35        145.3   0.02   2.66
#> 4. FB35_noC     163.3   0.30   0.92

Created on 2022-03-08 by the reprex package (v2.0.1)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant