| lower {holiglm} | R Documentation |
Lower Bound
Description
Set a lower bound on the coefficients of specific covariates.
Usage
lower(kvars)
Arguments
kvars |
a named vector giving the lower bounds. The names should correspond to the names of the covariates in the model matrix. |
Value
A holistic generalized model constraint, object inheriting from class "hglmc".
References
McDonald, J. W., & Diamond, I. D. (1990). On the Fitting of Generalized Linear Models with Nonnegativity Parameter Constraints. Biometrics, 46 (1): 201–206. doi:10.2307/2531643
Slawski, M., & Hein, M. (2013). Non-negative least squares for high-dimensional linear models: Consistency and sparse recovery without regularization. Electronic Journal of Statistics, 7: 3004-3056. doi:10.1214/13-EJS868
See Also
Other Constraint-Constructors:
group_equal(),
group_inout(),
group_sparsity(),
include(),
k_max(),
linear(),
pairwise_sign_coherence(),
rho_max(),
sign_coherence(),
upper()
Examples
set.seed(0)
dat <- rhglm(100, c(1, 2, -3, 4, 5, -6))
constraints <- lower(c(x2 = 0, x5 = 1))
hglm(y ~ ., constraints = constraints, data = dat)
# non-negative least squares
dat <- rhglm(100, c(1, 2, -3, 4, 5, -6))
constraints <- lower(setNames(double(5), paste0("x", 1:5)))
hglm(y ~ ., constraints = constraints, data = dat)