Least absolute shrinkage and selection operator through internal cross validation

Build a LASSO classifier using internal cross validation to choose the turning parameter, with a 5-fold cross validation as default.

Usage

lasso.intcv(kfold = 5, X, y, seed, alp = 1)

Arguments

kfold: number of folds. By default, kfold = 5.
X: dataset to be trained. This dataset must have rows as probes and columns as samples.
y: a vector of sample group of each sample for the dataset to be trained. It must have an equal length to the number of samples in X.
seed: an integer used to initialize a pseudorandom number generator.
alp: alpha, the penalty type. It can be any numeric value from 0 to 1. By default, alp = 1 which is for LASSO. alp = 0 is for ridge and any value in between is for elastic net.

Value

a list of 4 elements:

mc: an internal misclassification error rate
time: the processing time of performing internal validation with LASSO
model: a LASSO classifier, resulted from cv.fit
cfs: estimated coefficients for the final classifier

References

Friedman, J., Hastie, T. and Tibshirani, R. (2008) Regularization Paths for Generalized Linear Mod- els via Coordinate Descent, http://www.stanford.edu/~hastie/Papers/glmnet.pdf Journal of Statistical Software, Vol. 33(1), 1-22 Feb 2010

Examples

set.seed(101)
biological.effect <- estimate.biological.effect(uhdata = uhdata.pl)
ctrl.genes <- unique(rownames(uhdata.pl))[grep("NC", unique(rownames(uhdata.pl)))]
biological.effect.nc <- biological.effect[!rownames(biological.effect)
  %in% ctrl.genes, ]
group.id <- substr(colnames(biological.effect.nc), 7, 7)

biological.effect.train.ind <- colnames(biological.effect.nc)[c(sample(which(
  group.id == "E"), size = 64),
  sample(which(group.id == "V"), size = 64))]
biological.effect.nc.tr <- biological.effect.nc[, biological.effect.train.ind]

lasso.int <- lasso.intcv(X = biological.effect.nc.tr,
                         y = substr(colnames(biological.effect.nc.tr), 7, 7),
                         kfold = 5, seed = 1, alp = 1)
#>