parallelFit: Fit ternary network models using parallel tempering

In mccallm/ternarynet: Ternary Network Estimation

Fit ternary network models using parallel tempering

Description

Fit ternary network models using parallel tempering

Usage

    parallelFit(experiment_set, 
        max_parents, 
        n_cycles, 
        n_write, 
        T_lo, 
        T_hi,
        target_score,
        n_proc,
        logfile,
        n_thread,
        init_parents,
        init_outcomes,
        exchange_interval,
        adjust_move_size_interval,
        max_states,
        callback)

Arguments

experiment_set	data frame containing five columns: i_exp (experiment index), i_node (node index), outcome (-1/0/1), value (cost for that outcome), is_perturbation (0 or 1)
max_parents	maximum number of parents allowed for each node
n_cycles	maximum number of Monte Carlo cycles
n_write	number of times to write output during the run
T_lo	T for lowest-temperature replica
T_hi	T for highest-temperature replica
target_score	target_score - run will terminate if this is reached
n_proc	number of replicas
logfile	filename for log file
n_thread	number of openMP threads to run per process; default=1
init_parents	initial parents; randomized if null
init_outcomes	initial outcomes; set to '.' if null
exchange_interval	steps between exchanges; default=1000
adjust_move_size_interval	steps between move size adjustment; default=7001
max_states	max states to propagate when testing for repetition; default=10
callback	callback function, should take one integer argument (the replica number), used to call set.seed with different seed for each replica

Value

The return value is a list with an element for each replica. Each element is itself a list of the best unnormalized score, normalized score (unnormalized score divided by product of number of nodes and number of experiments), list of parents for each node, and array describing the transition rule, giving the outcome of a node for each possible configuration of parent nodes.

Author(s)

Harry A. Stern and Matthew N. McCall

Examples

i_exp <- as.integer(c(0,0,0, 0,0,0, 0,0,0, 0,0,0,
                        1,1,1, 1,1,1, 1,1,1, 1,1,1,
                        2,2,2, 2,2,2, 2,2,2, 2,2,2,
                        3,3,3, 3,3,3, 3,3,3, 3,3,3,
                        4,4,4, 4,4,4, 4,4,4, 4,4,4,
                        5,5,5, 5,5,5, 5,5,5, 5,5,5,
                        6,6,6, 6,6,6, 6,6,6, 6,6,6,
                        7,7,7, 7,7,7, 7,7,7, 7,7,7))

i_node <- as.integer(c(0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3,
                        0,0,0, 1,1,1, 2,2,2, 3,3,3))

outcome <- as.integer(c(-1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1,
                        -1,0,1, -1,0,1, -1,0,1, -1,0,1))

value <- c(0,1,2, 0,1,2, 0,1,2, 0,1,2,
            2,1,0, 0,1,2, 0,1,2, 0,1,2,
            2,1,0, 2,1,0, 0,1,2, 0,1,2,
            2,1,0, 2,1,0, 2,1,0, 0,1,2,
            2,1,0, 2,1,0, 2,1,0, 2,1,0,
            0,1,2, 2,1,0, 2,1,0, 2,1,0,
            0,1,2, 0,1,2, 2,1,0, 2,1,0,
            0,1,2, 0,1,2, 0,1,2, 2,1,0)

is_perturbation <- 
c(TRUE,TRUE,TRUE,  FALSE,FALSE,FALSE, FALSE,FALSE,FALSE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  TRUE,TRUE,TRUE, FALSE,FALSE,FALSE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  FALSE,FALSE,FALSE, TRUE,TRUE,TRUE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  FALSE,FALSE,FALSE, FALSE,FALSE,FALSE, TRUE,TRUE,TRUE,
    TRUE,TRUE,TRUE,  FALSE,FALSE,FALSE, FALSE,FALSE,FALSE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  TRUE,TRUE,TRUE, FALSE,FALSE,FALSE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  FALSE,FALSE,FALSE, TRUE,TRUE,TRUE, FALSE,FALSE,FALSE,
    FALSE,FALSE,FALSE,  FALSE,FALSE,FALSE, FALSE,FALSE,FALSE, TRUE,TRUE,TRUE)

indata <- data.frame(i_exp,i_node,outcome,value,is_perturbation)

results <- parallelFit(indata,
                        max_parents=1,
                        n_cycles=100000,
                        n_write=10,
                        T_lo=0.001,
                        T_hi=2.0,
                        target_score=0,
                        n_proc=1,
                        logfile='try.log')

lowest_temp_results <- results[[1]]

print('Unnormalized score:')
print(lowest_temp_results$unnormalized_score)

print('Normalized score:')
print(lowest_temp_results$normalized_score)

print('Parents:')
print(lowest_temp_results$parents)

print('Outcomes:')
print(lowest_temp_results$outcomes)

mccallm/ternarynet documentation built on Feb. 26, 2024, 3:51 a.m.