Description Usage Arguments Details References Examples
Significance test while performing variable selection with the lasso
1 2 3 4 |
x |
a penalized latent variable model contain |
It is essentially a copy of the covTest
function of the covTest package.
R. Lockhart, J. Taylor, R. J. Tibshirani and R. Tibshirani. A significance test for the lasso. Ann Stat. 2014 42(2):413-468
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | m <- lvm(Y ~ X1+X2+X3+X4+X5)
## simulation
set.seed(10)
mSim <- m
regression(mSim, Y~X1+X2+X3+X4+X5) <- list(1,0,0,0.1,0.5)
df <- sim(mSim,50)
dfs <- as.data.frame(scale(df))
## lvm
pm <- penalize(m)
res <- estimate(pm, data = dfs, regularizationPath = TRUE)
plot(res, type = "path")
getPath(res)
lassoTest(res)
penalized.PathL1 <- penalized(Y ~ ., data = dfs, steps = "Park", trace = FALSE)
seq_lambda <- unlist(lapply(penalized.PathL1, function(x){x@lambda1}))
n.lambda <- length(seq_lambda)
if(require(covTest)){
m.lars <- lars(y = dfs[,1], x = as.matrix(dfs[,-1]), type = "lasso")
a <- covTest(fitobj = m.lars, x = as.matrix(dfs[,-1]), y = dfs[,1])
# covTestLasso(beta = m.lars$beta, lambda = m.lars$lambda, Y = dfs[,1], X = as.matrix(dfs[,-1]))
# one difference with the LVM version is that p doesn't count the intercept
}
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.