Format trajectory tables

This function is used inside trajectoriesFromBayesian() to simulate trajectories from Bayesian model results. Assumes that rec_pred and surv_pred each include the same years and populations.

Usage

convertTrajectories(pars)

summarizeTrajectories(pars, returnSamples = T)

simulateTrajectoriesFromPosterior(
  popInfo = NA,
  rec_pred,
  surv_pred,
  initYear = NULL,
  correlateRates = FALSE,
  returnExpected = FALSE,
  c = formals(caribouPopGrowth)$c,
  K = FALSE,
  ...
)

Arguments

pars

returnSamples

popInfo

If NA (default) predictions are made without populations size, density dependence, or demographic stochasticity. See caribouPopGrowth() for details.

rec_pred

results returned by bboutools::bb_fit_recruitment() or bboutools::bb_predict_calf_cow_ratio() functions of bboutools R package, or recruit_fit returned by estimateBayesianRates().

surv_pred

bboufit object return by bboutools::bb_fit_survival() or bboutools::bb_predict_survival() functions of bboutools R package, or surv_fit returned by estimateBayesianRates().

initYear

numeric. Initial year.

correlateRates

logical. Set TRUE to force correlation between recruitment and survival. Ignored

returnExpected

logical. Default FALSE. Set TRUE to return expected values of R, S, and lambda (without interannual variation). Ignored if rec_pred/surv_pred are bboutools::bb_predict_calf_cow_ratio()/bboutools::bb_predict_survival() results.

c

Number. Bias correction term.

K

Number. Carrying capacity.

Value

convertTrajectories: formatted tables

summarizeTrajectories:

simulateTrajectoriesFromPosterior: a data frame with results from caribouPopGrowth() for each set of survival/recruitment predictions.

Interpretation of Years

Throughout the package a year is treated as the calendar year only if it is in a table that also contains a month column. In all other contexts year is treated as the caribou year which is the annual demographic cycle of caribou, measured from one calving season to the next. By default, April is set as the start of the caribou year.

See also

Caribou demography functions: addMissingYears(), addN0Variation(), bayesianScenariosWorkflow(), bayesianTrajectoryWorkflow(), betaNationalPriors(), caribouPopGrowth(), compareTrajectories(), compositionBiasCorrection(), dataFromSheets(), demographicProjectionApp(), estimateBayesianRates(), estimateNationalRate(), getCaribouYear(), getNationalCoefficients(), getScenarioDefaults(), plotCompareTrajectories(), plotSurvivalSeries(), plotTrajectories(), popGrowthTableJohnsonECCC, simulateObservations(), trajectoriesFromBayesian(), trajectoriesFromNational(), trajectoriesFromSummary(), trajectoriesFromSummaryForApp()

Caribou demography functions: addMissingYears(), addN0Variation(), bayesianScenariosWorkflow(), bayesianTrajectoryWorkflow(), betaNationalPriors(), caribouPopGrowth(), compareTrajectories(), compositionBiasCorrection(), dataFromSheets(), demographicProjectionApp(), estimateBayesianRates(), estimateNationalRate(), getCaribouYear(), getNationalCoefficients(), getScenarioDefaults(), plotCompareTrajectories(), plotSurvivalSeries(), plotTrajectories(), popGrowthTableJohnsonECCC, simulateObservations(), trajectoriesFromBayesian(), trajectoriesFromNational(), trajectoriesFromSummary(), trajectoriesFromSummaryForApp()

Examples

# \donttest{
  # Note these examples take a long time to run!
  mod <- estimateBayesianRates(
    surv_data = bboudata::bbousurv_a,
    recruit_data = bboudata::bbourecruit_a, 
    N0 = NA, return_mcmc = TRUE
  )
                           
  outmcmc = simulateTrajectoriesFromPosterior(popInfo=NA,mod$recruit_fit,mod$surv_fit)
  names(outmcmc)
#>  [1] "N0"                  "lambda"              "lambdaE"            
#>  [4] "N"                   "R_t"                 "X_t"                
#>  [7] "S_t"                 "n_recruits"          "surviving_adFemales"
#> [10] "lab"                 "Year"                "PopulationName"     
#> [13] "id"                 

  #get 95% prediction intervals from demographic trajectories
  PImcmc <- summarizeTrajectories(convertTrajectories(outmcmc))
  str(PImcmc, max.level = 3, give.attr = FALSE)
#> List of 2
#>  $ summary:'data.frame':	189 obs. of  8 variables:
#>   ..$ MetricTypeID  : chr [1:189] "N" "N" "N" "N" ...
#>   ..$ Year          : num [1:189] 1989 2012 1995 1997 2001 ...
#>   ..$ PopulationName: chr [1:189] "A" "A" "A" "A" ...
#>   ..$ Mean          : num [1:189] NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN ...
#>   ..$ lower         : Named num [1:189] NA NA NA NA NA NA NA NA NA NA ...
#>   ..$ upper         : Named num [1:189] NA NA NA NA NA NA NA NA NA NA ...
#>   ..$ probViable    : num [1:189] NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN ...
#>   ..$ Metric        : chr [1:189] "Female population size" "Female population size" "Female population size" "Female population size" ...
#>  $ samples:'data.frame':	567000 obs. of  8 variables:
#>   ..$ MetricTypeID    : chr [1:567000] "N" "N" "N" "N" ...
#>   ..$ Replicate       : chr [1:567000] "x2565" "x2567" "x2569" "x2568" ...
#>   ..$ LambdaPercentile: logi [1:567000] NA NA NA NA NA NA ...
#>   ..$ Year            : num [1:567000] 2003 2003 2003 2003 2003 ...
#>   ..$ PopulationName  : chr [1:567000] "A" "A" "A" "A" ...
#>   ..$ Timestep        : num [1:567000] 2003 2003 2003 2003 2003 ...
#>   ..$ Amount          : num [1:567000] NA NA NA NA NA NA NA NA NA NA ...
#>   ..$ Metric          : chr [1:567000] "Female population size" "Female population size" "Female population size" "Female population size" ...
# }