Quantile Regression w/ Lasso Penalty — rq.fit.lasso • quantspace

Quantile Regression w/ Lasso Penalty

rq.fit.lasso(
  X,
  y,
  tau,
  lambda,
  weights,
  scale_x = T,
  method = "two_pass",
  nfold = 10,
  nlambda = 50,
  parallel = F,
  ...
)

Arguments

X: Design matrix, X
y: outcome variable, y
tau: quantile to estimate
lambda: penalty parameter
weights: optional vector of weights
scale_x: whether to scale the design matrix before estimation
method: method to use when fitting underlying quantile regression algorithm
nfold: number of folds to use when cross-validating
nlambda: number of lambdas to search over.
parallel: whether to run cv search in parallel, if applicable
...: other arguments to pass to underlying fitting algorithm