solver_opts {geneTree} R Documentation

Override the fixed point solving options of a trained model at runtime, so that

Description

the solver can be tuned without re-training the whole network.

Usage

solver_opts(tree,
    iterations = -1,
    damping = -1,
    tolerance = -1,
    hard.cap = -1);

Arguments

tree

the trained XGBoostGRN model object. [as XGBoostGRN]

iterations

the soft iteration limit; a negative value keeps the current value. [as integer]

damping

the damping coefficient alpha in the range (0, 1]; a negative value keeps the current value. A larger alpha moves faster but is easier to oscillate. [as double]

tolerance

the relative convergence tolerance; a negative value keeps the current value. [as double]

hard.cap

the hard iteration limit which is used when the soft limit is not enough for convergence; a negative value keeps the current value, 0 restores the default (max(iterations * 5, 300)). [as integer]

env

the R# runtime environment object. [as Environment]

Details

the control steady state and the last perturbation result are invalidated, so the next perturbation experiment re-solves the control steady state with the new options.

Authors

biosystem

Value

the input model object that the solver options have been applied.

clr value class

Examples


[Package geneTree version 1.0.0.0 Index]