Determine the first q Predictors in the Lasso Path
Determines the q predictors that enter the lasso path first.
lasso.firstq(x, y, q, ...)
x |
numeric design matrix (without intercept) of dimension n * p. |
y |
response vector of length n. |
q |
number of predictors that should be selected, a positive integer. |
... |
optional additional arguments to be passed to |
The lasso.firstq function calls
glmnet in a special way and simply postprocesses its
nonzero predictor list, see its source code.
Vector of selected predictors.
Lukas Meier
x <- matrix(rnorm(100*1000), nrow = 100, ncol = 1000)
y <- x[,1] * 2 + x[,2] * 2.5 + rnorm(100)
sel <- lasso.firstq(x, y, q = 5)
sel # 5 integers from {1,2, ..., 1000}, including '1' and '2', typicallyPlease choose more modern alternatives, such as Google Chrome or Mozilla Firefox.