Calculates the deviance.
Usage
# S3 method for class 'cv.corila'
deviance(object, ...)Arguments
- object
object of class
"cv.corila"- ...
(for compatibility with stats::deviance)
Details
Returns the deviance calculated by glmnet::deviance.glmnet()
for the model with the optimised mixing and regularisation hyperparameters.
See also
The internal function .deviance() calculates
the deviance from fitted and observed values.
Examples
# listing S3 methods
methods(class = "cv.corila")
#> [1] coef deviance fitted nobs plot predict print
#> [8] residuals summary
#> see '?methods' for accessing help and source code
# simulating data
n <- 10L; p <- 20L; q <- 5L
x <- matrix(rnorm(n * p), nrow = n , ncol = p)
y <- rnorm(n)
group <- rep(seq_len(q), length.out = p)
primary <- as.logical(rbinom(n = p, size = 1L, prob = 0.5))
# fitting the model
object <- cv.corila(x = x, y = y, group = group, primary = primary)
#> Warning: Option grouped=FALSE enforced in cv.glmnet, since < 3 observations per fold
# using S3 methods
coef(object)
#> (intercept) <NA> <NA> <NA> <NA> <NA>
#> 0.14865837 0.00000000 0.05360646 0.00000000 0.00000000 0.00000000
#> <NA> <NA> <NA> <NA> <NA> <NA>
#> 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000
#> <NA> <NA> <NA> <NA> <NA> <NA>
#> 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000
#> <NA> <NA> <NA>
#> 0.00000000 0.00000000 0.00000000
predict(object, newx = x)
#> [1] 0.14328750 0.18686188 0.14471484 0.14664093 0.11211696 0.13127539
#> [7] 0.15188335 0.11708982 0.17714999 0.06726264
fitted(object)
#> [1] 0.14328750 0.18686188 0.14471484 0.14664093 0.11211696 0.13127539
#> [7] 0.15188335 0.11708982 0.17714999 0.06726264
residuals(object)
#> [1] -0.1792099 0.8822996 -0.6286898 -0.2676510 -1.4062570 0.3630375
#> [7] 1.1560182 1.3799512 0.6375527 -1.9370514
plot(object)
print(object)
#> object of class ‘cv.corila’
#> (contains multiple objects of class ‘cv.glmnet’)
#> selected 1 from 20 predictors
summary(object)
#> --- object of class “cv.corila” ---
#> generalised linear model with gaussian family
#> 20 features (12 primary and 8 auxiliary features)
#> initial coefficients: ridge regression
#> final coefficients: adaptive lasso regression
#> optimised regularisation parameter: lambda.min = 5.061
#> selected weights: local = 1, global = 0
#> selected exponents: local = 0, global = Inf
#> 2 non-zero coefficients (including intercept)