| ODEsensitivity {ODEsensitivity} | R Documentation |
Performing Sensitivity Analysis in ODE Models
Description
ODEsensitivity provides methods to perform sensitivity analysis (SA)
in ordinary differential equation (ODE) models. Its functions are based on
the implementations of Morris and Sobol' SA in the
sensitivity package (Pujol et al., 2015). However, a modified
version of the sensitivity-package is required that enables
morris,
soboljansen and
sobolmartinez to handle three-dimensional
arrays as model outputs. Each element of the third dimension of the output
array is then used to contain the results for one state variable of the ODE
model. Each element of the second dimension of the output array is used for
one timepoint.
Details
The main functions are ODEmorris and ODEsobol,
which are generic functions. They have default methods for general ODE models
(ODEmorris.default, ODEsobol.default) as well as methods
for objects of class ODEnetwork (ODEmorris.ODEnetwork,
ODEsobol.ODEnetwork). For the latter two methods, the package
ODEnetwork is required.
See the sensitivity package and its
morris, soboljansen and
sobolmartinez implementations for further
information on sensitivity analysis in R.