HMF_fit: System identification

In WarrenDavidAnderson/dynamicNetworkID:

Description

This function performs system identification to infer the connectivity and dynamics of a linear network model (Anderson et al., 2017).

Usage

HMF_fit(datHMF = NULL, N_mRanges = 10, alpha = NULL, nlambda = NULL,
  Mall = NULL, Msc = 2, n = 1, dt = NULL, epsilon = 10^(-10),
  Harg0 = 1e-06)

Arguments

datHMF	The data matrix with scaled measurements measurements. Scaled center measurements are provided for each measurement/annotation (e.g., gene/organ) combination. This format is produced by scale_zeroOne(). Defaults to datHMF.
N_mRanges	The number of m-value sets to evaluate.
alpha	The L1 regularization terms for constraining the regression. Defaults to seq(0,1,0.2).
nlambda	The number of L2 regularization terms for consideration. Defaults to 10.
Mall	A set of m-values. Defaults to NULL.
Msc	A scale factor that determines the M value (M = max(m)) corresponding to the largest range of m-values. Mmax <- ceiling(Msc * Np / 2) where Np = number of parameters.
n	Highest order of the system (n=1 for the first degree system that this analysis was designed for).
dt	Simulation time step
epsilon	Term for computing simulation error. Defaults to 10^(-10).
Harg0	If omega_0 is equal to zero (x=0), set omega_0 to Harg0 (w=Harg0).

Value

Error data for all simulations, simulation data for the best simulation, a parameter matrix for the best simulation.

References

http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005627

Examples

err0 = HMF_fit(datHMF,N_mRanges=2,dt=0.1,alpha=c(0.2,0.6))

WarrenDavidAnderson/dynamicNetworkID documentation built on May 23, 2019, 4:23 p.m.