CalculateLocalSens: Local sensitivity analysis

In capm: Companion Animal Population Management

Description

Wraper for sensFun function, which estimates local effect of all model parameters on population size, applying the so-called sensitivity functions. The set of parameters used in any of the following functions can be assessed: SolveIASA, SolveSI or SolveTC.

Usage

CalculateLocalSens(model.out = NULL, sensv = "n")

Arguments

model.out	output from one of the previous functions or a list with equivalent structure.
sensv	string with the name of the output variables for which sensitivity are to be estimated.

Details

For further arguments of sensFun, defaults are used. See the help page of this function for details. Methods for class "sensFun" can be used.

Value

a data.frame of class sensFun containing the sensitivity functions. There is one row for each sensitivity variable at each independent time. The first column x, contains the time value; the second column var, the name of the observed variable; and remaining columns have the sensitivity parameters.

References

Soetaert K and Petzoldt T (2010). Inverse modelling, sensitivity and monte carlo analysis in R using package FME. Journal of Statistical Software, 33(3), pp. 1-28.

Reichert P and Kfinsch HR (2001). Practical identifiability analysis of large environmental simulation models. Water Resources Research, 37(4), pp.1015-1030.

Baquero, O. S., Marconcin, S., Rocha, A., & Garcia, R. D. C. M. (2018). Companion animal demography and population management in Pinhais, Brazil. Preventive Veterinary Medicine.

http://oswaldosantos.github.io/capm

See Also

sensRange.

Examples

## IASA model

## Parameters and intial conditions.
data(dogs)
dogs_iasa <- GetDataIASA(dogs,
                         destination.label = "Pinhais",
                         total.estimate = 50444)

# Solve for point estimates.
solve_iasa_pt <- SolveIASA(pars = dogs_iasa$pars,
                          init = dogs_iasa$init,
                          time = 0:15,
                          alpha.owned = TRUE,
                          method = 'rk4')

## Calculate local sensitivities to all parameters.
local_solve_iasa2 <- CalculateLocalSens(
  model.out = solve_iasa_pt, sensv = "n2")
local_solve_iasa1 <- CalculateLocalSens(
  model.out = solve_iasa_pt, sensv = "n1")

capm documentation built on Oct. 30, 2019, 9:52 a.m.