skip_bias zeroes the mask at the FIRST element of the row-major
flattening, which for a matrix theta is entry [1, 1].
Usage
morie_geron_l1_regularization(theta, alpha, skip_bias = FALSE)
Arguments
- theta
Coefficients.
- alpha
Non-negative penalty.
- skip_bias
Leave the first coefficient unpenalised.
Value
List with penalty, gradient, prox, n_zero,
estimate, n.
Examples
morie_geron_l1_regularization(theta = 0.5, alpha = 0.5)
#> $penalty
#> [1] 0.25
#>
#> $gradient
#> [1] 0.5
#>
#> $prox
#> [1] 0
#>
#> $n_zero
#> [1] 1
#>
#> $estimate
#> [1] 0.25
#>
#> $n
#> [1] 1
#>
#> $method
#> [1] "L1 (Lasso) regularization penalty"
#>
