cno_smc: Function to perform SMC on the CNO protein network

In mikegros/cnosmc: SMC sampling for structure discovery and calibration of protein networks with CNOPTR

Description

TO DO: Add more of a full description later

Usage

cno_smc(n_samples, data, model, init_links = 1, p_link = 0.9, n_mh = 5, split_inhib = FALSE, n_cores = 1, diagnostics = F)

Arguments

n_samples	scalar number of SMC samples generated
data	data object from CNOFuzzy network. Contains simulated data, input settings, etc.
model	model object from CNOFuzzy network. Contains main values for showing the network structure
init_links	scalar or vector of indices of active links at the start of the sampling. Must contain one link to a stimulus node and one to an observed node and be a completed path.
p_link	scalar Prior probability that a link is active
n_mh	scalar number of Metropolis-Hastings jittering steps for preventing particle degeneracy in the SMC.
split_inhib	logical indicator of whether Gstring parameters should be split for each setting combination for the inhibitors or not. Allows model structure to be identified conditional on the input settings of the inhibitors. This is valuable in dealing with identifiability of links.
n_cores	scalar number of parallel cores to use for sampling through the SMC particles. n_cores = 1 runs in serial. NOTE (MJG): PARALLEL HAS NOT YET BEEN TESTED!
diagnostics	logical indicator of whether diagnostic information should be saved or printed. NOTE(MJG): Not yet tested and will give a lot of junk if TRUE for now

Details

TO DO: Add details

Value

list of matrices containing the SMC samples for the transfer function parameters on the links as well as for the indicator of whether the link is active.

Author(s)

Mike Grosskopf

References

TO DO: Add references

mikegros/cnosmc documentation built on Aug. 18, 2019, 3:31 a.m.