ridge.approach: Internal lassoenet function
In MokyoZhou/lassoenet: An Interactive Implementation of Penalised Regressions

Description Usage Arguments Value Details Author(s)

Internal lassoenet function

ridge.approach(x = x, y = y, err.curves = err.curves,
  type.lambda = type.lambda, gamma.seq = c(0.5, 1, 2), CI = CI,
  dataa = dataa, parallel = parallel, xx.indices = xx.indices,
  B.rep = B.rep, significance = significance)

`x`	A `model.matrix` of the predictors.
`y`	A vector of response values for the model fitting.
`err.curves`	The number of error curves to be fitted. Default is 0.
`type.lambda`	Either "lambda.min" or "lambda.1se".
`gamma.seq`	The gamma sequence to try.
`CI`	TRUE for residual bootstrapping. In this version is always TRUE.
`dataa`	Your full `data.frame`
`parallel`	Parallelisation
`xx.indices`	Locations of the predictors within dataa.
`B.rep`	The number of residual bootstrappings to do.
`significance`	The significance level of the confidence intervals e.g. 100(1-α)%.

A vector of outputs of the best Adpative Lasso model and the 100(1-α)% confidence intervals for the point estimates from using residual bootstrapping.

These are not intended for use by users. This function is one of the main engines for the Adaptive Lasso computation. This function is used when the weighting method "ridge" is selected. This together with the function ridge.robust forms the computation operator for the Adaptive Lasso when using the weighting method "ridge". The return from this function will enter Adlasso for futher wrapping.

Mokyo Zhou

MokyoZhou/lassoenet documentation built on May 20, 2019, 11:38 a.m.