calibrate: Sequential parameter estimation for the calibration of models
In calibrar: Automated Parameter Estimation for Complex (Ecological) Models

Description Usage Arguments Details Author(s) Examples

View source: R/calibraR-main.R

This function performs the optimization of a function, possibly in sequential phases of increasing complexity, and it is designed for the calibration of a model, by minimizing the error function fn associated to it.

1
2
3

calibrate(par, fn, gr = NULL, ..., method = "default", lower = NULL,
  upper = NULL, control = list(), hessian = FALSE, phases = NULL,
  replicates = 1)

`par`	A numeric vector. The length of the par argument defines the number of parameters to be estimated (i.e. the dimension of the problem).
`fn`	The function to be minimized.
`gr`	the gradient of `fn`. Ignored, added for portability with other optimization functions.
`method`	The optimization method to be used. The 'default' method is the AHR-ES (Oliveros & Shin, 2016). All the methods from stats::optim, optimx::optimx and cmaes::cma_es are available.
`lower`	Lower threshold value(s) for parameters. One value or a vector of the same length as par. If one value is provided, it is used for all parameters. `NA` means `-Inf`. By default `-Inf` is used (unconstrained).
`upper`	Upper threshold value(s) for parameters. One value or a vector of the same length as par. If one value is provided, it is used for all parameters. `NA` means `Inf`. By default `Inf` is used (unconstrained).
`control`	Parameter for the control of the algorithm itself, see details.
`hessian`	Logical. Should a numerically differentiated Hessian matrix be returned? Currently not implemented.
`phases`	An optional vector of the same length as `par`, indicating the phase at which each parameter becomes active. If omitted, default value is 1 for all parameters, performing a single optimization.
`replicates`	The number of replicates for the evaluation of `fn`. The default value is 1. A value greater than 1 is only useful for stochastic functions.
`...`	Additional parameters to be passed to `fn`.

In the control list, aggFn is a function to aggregate fn to a scalar value if the returned value is a vector. Some optimization algorithm can exploite the additional information provided by a vectorial output from fn.

Ricardo Oliveros-Ramos

calibrate(par=rep(NA, 5), fn=SphereN)
## Not run: 
calibrate(par=rep(NA, 5), fn=SphereN, replicates=3)
calibrate(par=rep(0.5, 5), fn=SphereN, replicates=3, lower=-5, upper=5)
calibrate(par=rep(0.5, 5), fn=SphereN, replicates=3, lower=-5, upper=5, phases=c(1,1,1,2,3))
calibrate(par=rep(0.5, 5), fn=SphereN, replicates=c(1,1,4), lower=-5, upper=5, phases=c(1,1,1,2,3))

## End(Not run)