sfit_options(
normalize = "log2",
pseudo = 1,
sparsity = 0.2,
max.edges = 3,
max.regulators = 0,
l1 = 0.01,
stability = 0.2,
smooth = 5,
prior.penalty = 0.2,
prior.sign.penalty = 0.2,
prior.rank.boost = 1,
prior = NULL,
de.pop = 40,
de.iters = 1200,
seed = 1234567,
parallel = TRUE,
verbose = FALSE,
refine = FALSE,
final.time = 0,
precise = 0.1,
steps = 0);
normalize
the normalization method of the gene expression matrix, could be log2(default), none, rowzscore or quantile. [as string]
pseudo
the pseudo count that used by the log2 normalization. [as double]
sparsity
the sparsity threshold of the net elasticity, the interaction with a smaller |g - h| than this threshold will not be kept in the network topology. [as double]
max.edges
the max number of the regulators that could be kept for each target gene. [as integer]
max.regulators
the max number of the candidate regulators for each target gene, the prior supported regulators are always included before the correlation based preselection; zero means that all of the genes will be used as candidate. [as integer]
l1
the L1 penalty weight of the kinetic orders. [as double]
stability
the penalty weight of a positive self loop. [as double]
smooth
the moving average window that used in the numerical differentiation, it is required by the noisy pseudotime data; zero means no smoothing. [as integer]
prior.penalty
the soft penalty weight of the interactions that are not supported by the prior. [as double]
prior.sign.penalty
the penalty weight of the direction conflict against the prior. [as double]
prior.rank.boost
the ranking boost of the prior supported edges in the sparsification. [as double]
prior
the prior knowledge network object. [as PriorNetwork]
de.pop
the population size of the differential evolution solver. [as integer]
de.iters
the max iteration number of the differential evolution solver. [as integer]
seed
the random seed. [as integer]
parallel
run the per gene fitting in parallel or not?. [as boolean]
verbose
print the fitting progress or not?. [as boolean]
refine
refine the kinetic parameters by the trajectory fitting after the decoupled slope fitting?. [as boolean]
final.time
the final time of the simulation, zero means the end of the time axis. [as double]
precise
the integration step of the simulation. [as double]
steps
the number of the simulation sample points, zero means auto. [as integer]