predict.grpnet: make predictions from a "grpnet" object.

In adelie: Group Lasso and Elastic Net Solver for Generalized Linear Models

make predictions from a "grpnet" object.

Description

Similar to other predict methods, this functions predicts linear predictors, coefficients and more from a fitted "grpnet" object.

Usage

## S3 method for class 'grpnet'
predict(
  object,
  newx,
  lambda = NULL,
  type = c("link", "response", "coefficients"),
  newoffsets = NULL,
  ...
)

## S3 method for class 'grpnet'
coef(object, lambda = NULL, ...)

Arguments

object	Fitted "grpnet" model.
newx	Matrix of new values for x at which predictions are to be made. Can be a matrix, a sparse matrix as in Matrix package, or else any of the matrix forms allowable in the adelie package. The number of columns must match that of the input matrix used in fitting object. If the model object was fit with standardize=TRUE, the saved centers and scaling will be applied to this matrix. This argument is not used for type="coefficients"
lambda	Value(s) of the penalty parameter lambda at which predictions are required. Default is the entire sequence used to create the model. If values of lambda are supplied, the function uses linear interpolation to make predictions for values of lambda that do not coincide with those used in the fitting algorithm.
type	Type of prediction required. Type "link" is the default, and gives the linear predictors. Type "response" applies the inverse link to these predictions. Type "coefficients" extracts the coefficients, intercepts and the active-set sizes.
newoffsets	If an offset is used in the fit, then one must be supplied for making predictions (except for type="coefficients" or type="nonzero")
...	Currently ignored.

Details

The shape of the objects returned are different for "multinomial" and "multigaussian" objects coef(...) is equivalent to predict(type="coefficients",...)

Value

The object returned depends on type.

Author(s)

James Yang, Trevor Hastie, and Balasubramanian Narasimhan
Maintainer: Trevor Hastie [e-mail]

References

Yang, James and Hastie, Trevor. (2024) A Fast and Scalable Pathwise-Solver for Group Lasso and Elastic Net Penalized Regression via Block-Coordinate Descent. arXiv \Sexpr[results=rd]{tools:::Rd_expr_doi("10.48550/arXiv.2405.08631")}.
Adelie Python user guide https://jamesyang007.github.io/adelie/

See Also

grpnet, and print, and coef methods, and cv.grpnet.

Examples

set.seed(0)
n <- 100
p <- 200
X <- matrix(rnorm(n * p), n, p)
y <- X[,1] * rnorm(1) + rnorm(n)
fit <- grpnet(X, glm.gaussian(y))
coef(fit)
predict(fit,newx = X[1:5,])

adelie documentation built on Sept. 11, 2024, 6:36 p.m.