lambdamax {lassogrp} | R Documentation |
Calculates the maximal value of the weight lambda of the L1 penalty term in a Lasso regression. For values >= this value, the null model will be obtained as the result of the penalized regression.
lambdamax(x, ...) ## S3 method for class 'formula' lambdamax(formula, nonpen = ~1, data, weights, subset, na.action, offset, coef.init, penscale = sqrt, model = LogReg(), center = NA, standardize = TRUE, contrasts = NULL, nlminb.opt = list(), ...) ## Default S3 method: lambdamax(x, y, index, weights = NULL, offset = rep(0, length(y)), coef.init = rep(0, ncol(x)), penscale = sqrt, model = LogReg(), center = NA, standardize = TRUE, nlminb.opt = list(), ...)
x |
design matrix (including intercept) |
y |
response vector |
formula |
|
nonpen |
|
data |
|
index |
vector which defines the grouping of the
variables. Components sharing the same
number build a group. Non-penalized coefficients are marked with
|
weights |
vector of observation weights. |
subset |
an optional vector specifying a subset of observations to be used in the fitting process. |
na.action |
a function which indicates what should happen when the data
contain |
offset |
vector of offset values. |
coef.init |
initial parameter vector. Penalized groups are discarded. |
penscale |
rescaling function to adjust the value of the penalty parameter to the degrees of freedom of the parameter group. See the reference below. |
model |
an object of class |
standardize, center |
logical; see |
contrasts |
an (optional) list with the contrasts for the factors in the model. |
nlminb.opt |
arguments to be supplied to |
... |
additional arguments to be passed to the functions defined
in |
Uses nlminb
to optimize the non-penalized parameters.
Numerical value of the maximal lambda
Lukas Meier, Seminar f. Statistik, ETH Zurich
data(splice) lambdamax(y ~ ., data = splice, model = LogReg(), center = TRUE, standardize = TRUE)