There is a direct correspondence between the inverse covariance matrix and multiple regression (Kwan 2014; Stephens 1998) . This readily allows for converting the GGM parameters to regression coefficients. All data types are supported.
Usage
# S3 method for class 'explore'
coef(object, iter = NULL, progress = TRUE, ...)
Value
An object of class coef
, containting two lists.
betas
A list of length p, each containing a p - 1 byiter
matrix of posterior samplesobject
An object of classexplore
(the fitted model).
References
Kwan CC (2014).
“A regression-based interpretation of the inverse of the sample covariance matrix.”
Spreadsheets in Education, 7(1), 4613.
Stephens G (1998).
“On the Inverse of the Covariance Matrix in Portfolio Analysis.”
The Journal of Finance, 53(5), 1821–1827.
Examples
# \donttest{
# note: iter = 250 for demonstrative purposes
# data
Y <- ptsd[,1:4]
##########################
### example 1: ordinal ###
##########################
# fit model (note + 1, due to zeros)
fit <- explore(Y + 1,
type = "ordinal",
iter = 250,
progress = FALSE)
# summarize the partial correlations
reg <- coef(fit, progress = FALSE)
# summary
summ <- summary(reg)
summ
#> BGGM: Bayesian Gaussian Graphical Models
#> ---
#> Type: ordinal
#> Formula: ~ 1
#> ---
#> Call:
#> explore(Y = Y + 1, type = "ordinal", iter = 250, progress = FALSE)
#> ---
#> Coefficients:
#>
#> B1:
#> Node Post.mean Post.sd Cred.lb Cred.ub
#> B2 0.242 0.103 0.009 0.428
#> B3 0.141 0.120 -0.104 0.375
#> B4 0.399 0.093 0.206 0.565
#>
#> B2:
#> Node Post.mean Post.sd Cred.lb Cred.ub
#> B1 0.223 0.096 0.007 0.400
#> B3 0.544 0.091 0.342 0.693
#> B4 0.013 0.091 -0.176 0.192
#>
#> B3:
#> Node Post.mean Post.sd Cred.lb Cred.ub
#> B1 0.112 0.094 -0.072 0.296
#> B2 0.483 0.083 0.313 0.626
#> B4 0.280 0.091 0.105 0.453
#>
#> B4:
#> Node Post.mean Post.sd Cred.lb Cred.ub
#> B1 0.399 0.091 0.202 0.566
#> B2 0.016 0.099 -0.174 0.213
#> B3 0.343 0.105 0.151 0.549
#>
# }