sensitivityAnalysis: Compute parameter correlations and multicollinearity for the...

In lkshrsch/gLVInterNetworks: Inference of interaction networks based on generalised Lotka Volterra dynamics

Description

This function is a wrapper function for sensFun and Collin from package FME. It calculates both a sensitivity matrix S_ij and multicollinearity index for all parameter combinations.

Usage

sensitivityAnalysis(Parms)

Arguments

Parms

Numeric vector or matrix with the parameter coefficients to test

Value

List containing:

sens	Matrix containing sensitivity output values for each parameter and each model variable. The sensitivity matrix S_ij contains elements dy_i/dpar_j*parscale_j/varscale_i. The scale used to change the value of each parameter can be seen using the summary function on the sens table, and it is set to be the same value of the parameter itself
coll	Table with collinearity index for each possible parameter subset.

Warning

The function needs the original data as a global variable called "data". Please make sure that when using this function to perform identifiability analysis on a set of parameters, that the observation matrix or table used to estimate the parameter set is accessible under the name "data"

Note

For details in output interpretation see package FME

lkshrsch/gLVInterNetworks documentation built on May 21, 2019, 7:33 a.m.