pastboon (version 0.1.4)

calc_convergence_time: Calculate convergence time-step for node activities

Description

Given a node activity time-series for a set of variables node_act, this function calculates the time-step from which the changes in all the curves are below threshold for window_size consecutive time-steps.

Usage

calc_convergence_time(node_act, threshold, window_size = 1)

Value

The time-step at which convergence occurs. If no convergence is detected, NA is returned.

Arguments

node_act

A matrix describing node activities over consecutive time-steps (i.e., time-series), where rows represent time-steps and columns represent nodes. It is the output of calc_node_activities.

threshold

A value determining the maximum allowable change in node activities to decide if they have converged.

window_size

The number of consecutive time-steps for which the node activity curves must remain stable (i.e., changes below threshold) to be considered converged. The default is 1.

Details

The function checks if the changes in all node activity curves are less than threshold for window_size consecutive time-steps. If this condition is met, the node activity curves are considered to have converged to their stable values, and the convergence time-step (the starting point of the window) is returned. Since node activities represent marginal probabilities of the nodes being active at each time-step, convergence indicates that the steady-state distribution of the corresponding Boolean network has been reached, meaning that the probability of being in each state of the network no longer changes significantly.

Examples

# Load the example network
data(lac_operon_net)

# Define parameters for the SDDS method
props <- rep(0.95, length(lac_operon_net$genes))
params <- list(p00 = props, p01 = props, p10 = props, p11 = props)

# Get node activities after simulation using the SDDS method
node_act <- calc_node_activities(lac_operon_net, method = "SDDS", params = params,
    steps = 100, repeats = 10000)

# Calculate the convergence time
convergence_time <- calc_convergence_time(node_act, threshold = 0.01)


# Print the convergence time
print(convergence_time)