compareStructures: Parameter estimation over a list of different network...

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

Description

Given a list of network structures, estimate gLV parameters constrained to maintain such structure, and compare the results from each fitted model.

Usage

compareStructures <- function(data = mydata, structures = myStructures, verbose=TRUE, iterations = 1)

Arguments

data	Matrix or table containing time series of measurements in longitudinal form where first column corresponds to the time points and subsequent columns correspond to each model variable
structures	List of interaction network structures as generated by function networkStructures()
verbose	Boolean flag. Set to TRUE to print live output on the number of steps performed
iterations	Number of times the regression algorithm should be performed on each structure. The default is 1 for only one parameter estimation run per structure. If set to integers x > 1 the algorithm runs x times and keeps the best solution according to the goodness of fit. The number of different solutions found and the number of times the optimal solution was achieved is kept in the regression attributes in order to evaluate how many local maximas were found, and how certain the user can be that the kept optimal could approximate a global optimum point.

Value

Returns a list containing two objects, a list displaying all networks run and their respective goodness of fit, and a list containing full results for each run

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