r4ss (version 1.16)

addSSsummarize: Add a model to the list of models to compare

Description

Adds specified quantities from any model to the list of models returned from SSsummarize for further comparison.

Usage

addSSsummarize(origModels,newModels)

Arguments

origModels

A list of models created by SSsummarize.

newModels

A list of models to add to the originals models list. Each new model is an element of the list, and is a list itself with possible components described in the details below.

Value

• Returns list as is returned from SSsummarize, but contains additions for the new models.

Details

The models to add do not need to contain all of the entries. Missing entries are given NA values, so are not plotted. Each model is a list with the following elements: [object Object],[object Object],[object Object],[object Object],

• TOTAL
• Equil_catch
• Survey
• Length_comp
• Age_comp
• Recruitment
• recast_Recruitment
• Parm_priors
• Parm_softbounds
• Parm_devs
• Crash_Pen
• Size_at_age

,[object Object],[object Object],[object Object],[object Object],[object Object],

• 1st column is year
• 2nd column is spawning biomass in same units as original models (SS reports female spawning biomass)
• 3rd column is the standard deviation of estimated spawning biomass
• 4th column is a lower bound of the confidence interval to be plotted (say from an MCMC)
• 5th column is an upper bound of the confidence interval to be plotted (say from an MCMC)

,[object Object],

• 1st column is year
• 2nd column is depletion
• 3rd column is the standard deviation of depletion (optional)
• 4th column is a lower bound of the confidence interval to be plotted (say from an MCMC)
• 5th column is an upper bound of the confidence interval to be plotted (say from an MCMC)

,[object Object],

• 1st column is year
• 2nd column is depletion
• 3rd column is the standard deviation (optional)
• 4th column is a lower bound of the confidence interval to be plotted (say from an MCMC)
• 5th column is an upper bound of the confidence interval to be plotted (say from an MCMC)

,[object Object],

• 1st column is year
• 2nd column is recruitment as in original models (SS reports age-0 recruits)
• 3rd column is the standard deviation (optional)
• 4th column is a lower bound of the confidence interval to be plotted (say from an MCMC)
• 5th column is an upper bound of the confidence interval to be plotted (say from an MCMC)

,[object Object],

• 1st column is year
• 2nd column is deviate (matched with original models)

,[object Object],[object Object],

• 1st column is year
• 2nd column is observed index (data)
• 3rd column is expected index (prediction)
• 4th column is catchability coefficient (q)
• 5th column is standard error of index (total used in fitting)
• 6th column is a likelihood for this point, or enter any value to make sure it plots, or enter NA not to plot the estimate

,[object Object]

See Also

SSsummarize SSplotComparisons

Examples

######################################
  #DO NOT RUN
  tinss1 <- list(npars=A$fit$npar,maxgrad=A$fit$maxgrad,nsexes=1,
                SpawnBio=data.frame(c(1964,1965,A$yrs),c(A$sbo,A$sbo,A$sbt)*1e6,0,qnorm(0.025,c(A$so,A$so,A$sbt)*1e6,0),qnorm(0.975,c(A$so,A$so,A$sbt)*1e6,0)),     #there is an estimated parameter called sd_sbt, but it is a single value
                Bratio=data.frame(A$yrs,A$sbt/A$sbo,0,qnorm(0.025,A$sbt/A$sbo,0),qnorm(0.975,A$sbt/A$sbo,0)),
                SPRratio=data.frame(A$yr,A$spr,0,qnorm(0.025,A$spr,0),qnorm(0.975,A$spr,0)),
                recruits=data.frame(A$yr,A$nt[,1]*1e6,0,qnorm(0.025,A$nt[,1]*1e5,0),qnorm(0.975,A$nt[,1]*1e6,0)),
                recdevs=data.frame(A$recYrs,A$wt),  #I'm not sure exactly what wt are, but it is important to line them up correctly
                indices = data.frame(A$iyr,1e6*A$yt,1e6*A$qbt,rep(A$q,length(A$iyr)),rep(0.4,length(A$iyr)),rep(0,length(A$iyr)),rep(1,length(A$iyr)))
            )
  tinss <- list(tinss1,tinss1)   #can add more models here


  #add TINSS model to SS models already summarized                
  SSnTINSS <- addSSsummarize(models,tinss)
  mcmcInd <- seq(burnin+1,nrow(A$mc.sbt),thin)
  SSnTINSS$mcmc[[2]] <- data.frame(A$mc.sb0[mcmcInd],A$mc.sbt[mcmcInd,],A$mc.depl[mcmcInd,],A$mc.spr[mcmcInd,],A$mc.rt[mcmcInd,],log(A$mcmc[mcmcInd,"Ro"]*1e6),A$mcmc[mcmcInd,"msy"]*1e6)  
  names(SSnTINSS$mcmc[[2]]) <- c("SPB_Virgin",paste("SPB",A$yrs,sep="_"),paste("Bratio",A$yrs,sep="_"),paste("SPRratio",A$yr,sep="_"),paste("Recr",A$yr,sep="_"),"SR_R0","TotYield_MSY")
  modelnames <- c("SS", "TINSS","TINSS.MLE")
  
  SSplotComparisons(SSnTINSS, legendlabels=modelnames, subplot=2,endyr=2011,mcmcVec=c(T,T,F))
  title(main="MCMC")
  SSplotComparisons(SSnTINSS, legendlabels=modelnames, subplot=4,endyr=2011,mcmcVec=c(T,T,F))
  title(main="MCMC")
  ###############################################