logLik.bsrr {bestridge}

Extract the log-likelihood from a "bsrr.one" object.

Description

This function returns the log-likelihood for the fitted models.

Usage

## S3 method for class 'bsrr'
logLik(object, best.model = TRUE, ...)

Arguments

object

A "bsrr" object.

best.model

Whether only return the log-likelihood of the best model. Default is TRUE. If best.model = FALSE, the log-likelihood of the best models with model size and \lambda in the original s.list and lambda.list (for method = "sequential") or in the iteration path (for method = "gsection", method = "pgsection", and method = "psequential") is returned.

...

additional arguments

Details

The log-likelihood for the best model chosen by a certain information criterion or cross-validation corresponding to the call in bsrr or the best models with model size and \lambda in the original s.list and lambda.list (or the in the iteration path) can be returned. For "lm" fits it is assumed that the scale has been estimated (by maximum likelihood or REML), and all the constants in the log-likelihood are included.

Value

A matrix or vector containing the log-likelihood for each model is returned. For bsrr objects fitted by sequantial method, values in each row in the returned matrix corresponding to the model size in s.list, and each column the shrinkage parameters in lambda.list.

For bsrr objects fitted by gsection, pgsection and psequential, the returned vector contains log-likelihood for fitted models in each iteration. The coefficients of those model can be extracted from beta.all and coef0.all in the bsrr object.

Author(s)

Liyuan Hu, Kangkang Jiang, Yanhang Zhang, Jin Zhu, Canhong Wen and Xueqin Wang.

See Also

bsrr, summary.bsrr.

Examples

# Generate simulated data
n <- 200
p <- 20
k <- 5
rho <- 0.4
SNR <- 10
cortype <- 1
seed <- 10
Tbeta <- rep(0, p)
Tbeta[1:k*floor(p/k):floor(p/k)] <- rep(1, k)
Data <- gen.data(n, p, k, rho, family = "gaussian", cortype = cortype, snr = SNR, seed = seed)
lm.bsrr <- bsrr(Data$x, Data$y, method = "sequential")

logLik(lm.bsrr, best.model = FALSE)