Skip to contents

Bayesian information criterion, vectorised (Geron Ch 8, morie.fn hmbic)

Source: R/geron_ml_native.R
morie_geron_bic.Rd

BIC = -2 log L + k log n. best_index is 0-based.

Usage

morie_geron_bic(log_lik, k, n)

Arguments

log_lik

Scalar or vector of log-likelihoods.

k

Scalar or vector of parameter counts.

n

Sample size, >= 1.

Value

List with bic, delta, weights, best_index, penalty.

Examples

morie_geron_bic(log_lik = c(1, 2, 3, 4, 5, 6, 7, 8), k = 5L, n = 5L)
#> $bic
#> [1]  6.04718956  4.04718956  2.04718956  0.04718956 -1.95281044 -3.95281044
#> [7] -5.95281044 -7.95281044
#> 
#> $delta
#> [1] 14 12 10  8  6  4  2  0
#> 
#> $weights
#> [1] 0.0005766128 0.0015673960 0.0042606241 0.0115815771 0.0314819905
#> [6] 0.0855769227 0.2326221940 0.6323326828
#> 
#> $best_index
#> [1] 7
#> 
#> $penalty
#> [1] 1.609438
#> 
#> $estimate
#> [1] -7.95281
#> 
#> $n
#> [1] 5
#> 
#> $method
#> [1] "Bayesian information criterion (BIC = -2 log L + k log n)"
#>