R/df_lists.R

In nissandjac/PacifichakeMSE: Management strategy evaluation of Pacific hake

#' Title
#'
#' @param ls.save list of MSE data
#' @param nms names of runs
#'
#' @return
#' @export
#'
#' @examples
#'
df_lists <- function(ls.save, nms){

nyears <- dim(ls.save[[1]][1]$Catch)[2]

if(dim(ls.save[[1]][1]$Catch)[2] == 1){
  nyears <- dim(ls.save[[1]][1]$Catch)[1]
}

if(dim(ls.save[[1]]$Catch)[2] == 1){

simyears <- nyears-(length(1966:2018))+1
}else{
  simyears <- nyears-(length(1966:2018))+1

}

yr <- 1966:(2018+simyears-1)

nruns <- length(ls.save)
nfailed <- rep(1, nruns)

if(all(is.na(ls.save[[1]]))){
  ls.save[[-1]]
}

if(all(is.na(ls.save[[1]]))){
  ls.save[[-1]]
}


for(i in 1:nruns){
  if(length(ls.save[[i]]) > 1){

  if(i == 1){

    if(dim(ls.save[[1]]$Catch)[2] == 1){
      Catch.us <- ls.save[[i]]$Catch*0.76
      Catch.can <- ls.save[[i]]$Catch*0.24
      Catch = ls.save[[i]]$Catch

      # Allocated catch
      Catch.q.us <- rowSums(ls.save[[i]]$Catch.quota[,2,])
      Catch.q.can <- rowSums(ls.save[[i]]$Catch.quota[,1,])
      Catch.q = rowSums(ls.save[[i]]$Catch.quota)
    }else{

      catchtmp <- apply(ls.save[[i]]$Catch, MARGIN = c(2,3), FUN = sum)

      Catch = rowSums(catchtmp)
      Catch.us = catchtmp[,2]
      Catch.can = catchtmp[,1]

      Catch.q = rowSums(ls.save[[i]]$Catch.quota)
      Catch.q.us = rowSums(ls.save[[i]]$Catch.quota[,2,])
      Catch.q.can = rowSums(ls.save[[i]]$Catch.quota[,1,])

    }

    source('calcMeanAge.R')


    if(is.null(ls.save[[i]]$F0)){
      ls.save[[i]]$F0 <- matrix(NA, length(yr),2)
    }

    ls.df <- data.frame(year = yr,
                        SSB.can = ls.save[[i]]$SSB[,1],
                        SSB.US = ls.save[[i]]$SSB[,2],
                        SSBtot =  ls.save[[i]]$SSB[,1]+ls.save[[i]]$SSB[,2],
                        F0.can = ls.save[[i]]$F0[,1],
                        F0.us = ls.save[[i]]$F0[,2],
                        amc = ls.save[[i]]$amc$amc.tot,
                        ams = ls.save[[i]]$ams$ams.tot,
                        amc.can = ls.save[[i]]$amc$amc.can,
                        amc.us = ls.save[[i]]$amc$amc.US,
                        ams.can = ls.save[[i]]$ams$ams.can,
                        ams.us = ls.save[[i]]$ams$ams.US,
                        SSB.mid.can = ls.save[[i]]$SSB.mid[,1],
                        SSB.mid.us = ls.save[[i]]$SSB.mid[,2],
                        run = paste('run',i, sep = '-'),
                        Catch.q = Catch.q,
                        Catch.q.us = Catch.q.us,
                        Catch.q.can = Catch.q.can,
                        Catch = Catch,
                        Catch.us = Catch.us,
                        Catch.can = Catch.can,
                        F0.us = Catch.us/ls.save[[i]]$SSB.mid[,2],
                        F0.can = Catch.can/ls.save[[i]]$SSB.mid[,1])

  }else{

    if(dim(ls.save[[i]]$Catch)[2] == 1){
      Catch.q.us <- ls.save[[i]]$Catch*0.76
      Catch.q.can <- ls.save[[i]]$Catch*0.24
      Catch = ls.save[[i]]$Catch

    }else{

      catchtmp <- apply(ls.save[[i]]$Catch, MARGIN = c(2,3), FUN = sum)

      Catch = rowSums(catchtmp)
      Catch.us = catchtmp[,2]
      Catch.can = catchtmp[,1]

      Catch.q = rowSums(ls.save[[i]]$Catch.quota)
      Catch.q.us = rowSums(ls.save[[i]]$Catch.quota[,2,])
      Catch.q.can = rowSums(ls.save[[i]]$Catch.quota[,1,])


    }

      if(is.null(ls.save[[i]]$F0)){
        ls.save[[i]]$F0 <- matrix(NA, length(yr),2)
      }

      ls.tmp <- data.frame(year = yr, SSB.can = ls.save[[i]]$SSB[,1], SSB.US = ls.save[[i]]$SSB[,2],
                           SSBtot =  ls.save[[i]]$SSB[,1]+ls.save[[i]]$SSB[,2],
                           F0.can = ls.save[[i]]$F0[,1],
                           F0.us = ls.save[[i]]$F0[,2],
                           amc = ls.save[[i]]$amc$amc.tot,
                           ams = ls.save[[i]]$ams$ams.tot,
                           amc.can = ls.save[[i]]$amc$amc.can,
                           amc.us = ls.save[[i]]$amc$amc.US,
                           ams.can = ls.save[[i]]$ams$ams.can,
                           ams.us = ls.save[[i]]$ams$ams.US,
                           SSB.mid.can = ls.save[[i]]$SSB.mid[,1],
                           SSB.mid.us = ls.save[[i]]$SSB.mid[,2],
                           run = paste('run',i, sep = '-'),
                           Catch.q = Catch.q,
                           Catch.q.us = Catch.q.us,
                           Catch.q.can = Catch.q.can,
                           Catch = Catch,
                           Catch.us = Catch.us,
                           Catch.can = Catch.can,
                           F0.us = Catch.us/ls.save[[i]]$SSB.mid[,2],
                           F0.can = Catch.can/ls.save[[i]]$SSB.mid[,1])
      ls.df <- rbind(ls.df, ls.tmp)}

  }
  else{
    nfailed[i] <- 0
  }
}

ggplot(ls.df, aes(x = year, y = Catch, color = run))+geom_line()+theme(legend.position = 'none')

SSB.plotquant <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = median(SSB.can),
            p95.can = quantile(SSB.can, 0.95),
            p5.can = quantile(SSB.can,0.05),
            med.US = median(SSB.US),
            p95.US = quantile(SSB.US, 0.95),
            p5.US = quantile(SSB.US,0.05))
SSB.plotquant$run <- nms

SSB.plottot <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med = median(SSBtot),
            avg = mean(SSBtot),
            p95 = quantile(SSBtot, 0.95),
            p5 = quantile(SSBtot,0.05))
SSB.plottot$run <- nms



SSB.plotmid <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = median(SSB.mid.can),
            p95.can = quantile(SSB.mid.can, 0.95),
            p5.can = quantile(SSB.mid.can,0.05),
            med.US = median(SSB.mid.us),
            p95.US = quantile(SSB.mid.us, 0.95),
            p5.US = quantile(SSB.mid.us,0.05))
SSB.plotmid$run <- nms


Catch.plotquant <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med = median(Catch),
            p95 = quantile(Catch, 0.95),
            p5 = quantile(Catch,0.05)
            )
Catch.plotquant$run <- nms
#
ams.plotquant <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med = median(ams,na.rm = TRUE),
            p95 = quantile(ams, 0.95,na.rm = TRUE),
            p5 = quantile(ams,0.05,na.rm = TRUE)
  )
ams.plotquant$run <- nms

amc.plotquant <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med= median(amc,na.rm = TRUE),
            p95 = quantile(amc, 0.95,na.rm = TRUE),
            p5 = quantile(amc,0.05,na.rm = TRUE)
  )
amc.plotquant$run <- nms

ams.space <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = median(ams.can,na.rm = TRUE),
            p95.can = quantile(ams.can, 0.95,na.rm = TRUE),
            p5.can = quantile(ams.can,0.05,na.rm = TRUE),
            med.us = median(ams.us,na.rm = TRUE),
            p95.us = quantile(ams.us, 0.95,na.rm = TRUE),
            p5.us = quantile(ams.us,0.05,na.rm = TRUE)

  )
ams.space$run <- nms

amc.space <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = median(amc.can,na.rm = TRUE),
            p95.can = quantile(amc.can, 0.95,na.rm = TRUE),
            p5.can = quantile(amc.can,0.05,na.rm = TRUE),
            med.us = median(amc.us,na.rm = TRUE),
            p95.us = quantile(amc.us, 0.95,na.rm = TRUE),
            p5.us = quantile(amc.us,0.05,na.rm = TRUE)

  )
amc.space$run <- nms

F0.space <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = median(F0.can,na.rm = TRUE),
            p95.can = quantile(F0.can, 0.95,na.rm = TRUE),
            p5.can = quantile(F0.can,0.05,na.rm = TRUE),
            med.us = median(F0.us,na.rm = TRUE),
            p95.us = quantile(F0.us, 0.95,na.rm = TRUE),
            p5.us = quantile(F0.us,0.05,na.rm = TRUE)

  )
F0.space$run <- nms

Catch.q <- ls.df[ls.df$year > 2010,] %>%
  group_by(year) %>%
  summarise(med.can = mean(Catch.can/Catch.q.can,na.rm = TRUE),
            p95.can = quantile(Catch.can/Catch.q.can, 0.95,na.rm = TRUE),
            p5.can = quantile(Catch.can/Catch.q.can,0.05,na.rm = TRUE),
            med.us = mean(Catch.us/Catch.q.us,na.rm = TRUE),
            p95.us = quantile(Catch.us/Catch.q.us, 0.95,na.rm = TRUE),
            p5.us = quantile(Catch.us/Catch.q.us,0.05,na.rm = TRUE),
            med.tot = mean(Catch/Catch.q),
            p5.tot = quantile(Catch/Catch.q,0.05,na.rm = TRUE),
            p95.tot = quantile(Catch/Catch.q,0.95,na.rm = TRUE)

  )
Catch.q$run <- nms

ggplot(Catch.q, aes(x = year, y = med.can))+geom_line()+coord_cartesian(ylim = c(0.2,1.3))+
  geom_line(aes(y = med.us))+geom_line(aes(y = p5.can))+geom_line(aes(y = p95.can), col = 'red')+geom_line(aes(y=  p95.us), col ='blue')+
  geom_line(aes(y = p5.us), col = 'blue')

return(list(
  ls.df,
  nfailed,
  list(
              SSBplot = SSB.plotquant,
              SSBmid = SSB.plotmid,
              SSBtot = SSB.plottot,
              Catchplot = Catch.plotquant,
              amcplot = amc.plotquant,
              amsplot = ams.plotquant,
              amc.space = amc.space,
              ams.space = ams.space,
              F0 =F0.space,
              Catch.q = Catch.q

    )
  )
)
}