A function to check the sensitivity of cnp_model predictions based on the variation of input parameters

This function runs the cnp_model fixing all parameters SD's but one to test for sensitivity

sensitivity(
  TL,
  param,
  iter = 1000,
  par,
  out = c("Ic", "In", "Ip", "Gc", "Gn", "Gp", "Fc", "Fn", "Fp", "Wc", "Wn", "Wp"),
  ...
)

Arguments

TL

total length of a fish in cm

param

list of all parameter means ("_m") and standard deviations ("_sd") Default parameters are set with very low sd's. See cnp_model_mcmc for a list of all requested parameters

iter

A positive integer specifying the number of iterations. The default is 1000

par

Charachter vector specifying which input parameter sd's should be used for sensitivity.

out

Charachter vector specifying which output parameter sd's should be returned.

...

Other arguments that can be used from cnp_model_mcmc

Value

Returns a dataframe with sd's of model predictions. Row names indicate the variable, who's sd was used for the model run. Plots a heatplot with width of the 95

Examples

library(fishflux)
sensitivity(TL = 10, param = list(k_sd = 0.2, Dn_sd = 0.2, Dc_sd = 0.1),
            par = c("k_sd","Dn_sd","Dc_sd"), out = c("Ic", "In", "Ip", "Gc"))
#> Joining, by = "key"

#>            Ic_CI        In_CI        Ip_CI        Gc_CI
#> k_sd  0.03235292 3.882350e-03 1.294117e-03 1.242352e-02
#> Dn_sd 0.22027450 3.511069e-03 8.810980e-03 6.942732e-11
#> Dc_sd 0.01056740 2.525939e-10 8.406169e-11 7.486628e-11