User's Guide to lqa - LMU

REFERENCES 22
control
... further arguments
list of control parameters as returned by lqa.control(). See the
lqa manual (Ulbricht, 2010a) for details.
This function plots the coefficient build-ups for a given dimension of your tuning parameter(s).
The argument lambdaseq can be omitted. In this case a default sequence
R> lambdaseq plot.lqa (y, x, family = binomial (), penalty.family = fused.lasso,
+ offset.values = c (NA, 0.01), add.MLE = FALSE)
The corresponding plot is given in Figure 1. Note that the grouping effect does not
cover x 3 . This is due to the small weight λ 2 = 0.01 on the second penalty term that is
responsible for the fusion of correlated regressors. Note that the grouping effect among
x 1 and x 2 is kept all the time. Furthermore, the relevance of the regressors is recognized
as x 4 is selected to join the active set at last.
References
Bondell, H. D. and B. J. Reich (2008). Simultaneous regression shrinkage, variable selection
and clustering of predictors with oscar. Biometrics 64, 115–123.
Chambers, J. (1998). Programming with Data. A Guide to the S Language. New York:
Springer.
Fahrmeir, L., T. Kneib, and S. Lang (2009). Regression - Modelle, Methoden und Anwendungen
(2nd ed.). Berlin: Springer.
Fahrmeir, L. and G. Tutz (2001). Multivariate Statistical Modelling based on Generalized
Linear Models (2nd ed.). New York: Springer.
Frank, I. E. and J. H. Friedman (1993). A statistical view of some chemometrics regression
tools (with discussion). Technometrics 35, 109–148.
Hastie, T., R. Tibshirani, and J. H. Friedman (2009). The Elements of Statistical Learning
(2nd ed.). New York: Springer.