SubLasso: Logistic model via Lasso penalty with a subset features

In SubLasso: Gene selection using Lasso for Microarray data with user-defined genes fixed in model.

Description

Fit a logistic model via Lasso penalty. A subset features can be fixed in the model.

Usage

SubLasso(X, y, subset, nfold)

Arguments

X	gene expression matrix, column is sample, row is gene(probe sets).
y	category vector, 1 (positive, illness) or 0 (negative, normal).
subset	gene (probe sets) names must be included in the model; Default is null set.
nfold	number of cross-validation; Default is 5.

Details

some details

Value

selname	features selected by the model.
w	the coefficient (weight) of feature in the model
valid	sensitivity (Sn), specificity (Sp), Accuracy (Acc), and Matthews correlation coefficient(Mcc)
description	the description statistics of selected features by group.
correlation	the correlations between all selected features.

Author(s)

Youxi Luo

References

Friedman, J., Hastie, T. and Tibshirani, R. (2008) Regularization Paths for Generalized Linear Modelsvia Coordinate Descent, http://www.stanford.edu/~hastie/Papers/glmnet.pdf Journal of Statistical Software, Vol. 33(1), 1-22 Feb 2010. http://www.jstatsoft.org/v33/i01/

See Also

glmnet

Examples

 
#screen device is not support in examples but SubLasso function use it to visualize results.
#when you want to test examples, please uncomment following code.

##### Example 1
#data(Golub_Merge)
#X <- Golub_Merge$X
#y <- Golub_Merge$y
#f1=SubLasso(X,y,nfold=10)

## predict.sublasso(f1,X[1:10,]) ##error predicted x
#predy=predict.sublasso(f1,X)
#predy=predict.sublasso(f1,X,type="class")
#predy=predict.sublasso(f1,X,type="link")
#predy=predict.sublasso(f1,X,type="response")
#predy=predict.sublasso(f1,X,type="response",s=0.05)
#subset=f1$selname
#f2=SubLasso(X,y,subset,nfold=10)

#subset=row.names(X)[1:10]
#f3=SubLasso(X,y,subset,nfold=10)

#predy=predict.sublasso(f3,X)
#predy=predict.sublasso(f3,X,type="class")
#predy=predict.sublasso(f3,X,type="link")
#predy=predict.sublasso(f3,X,type="response")
#predy=predict.sublasso(f3,X,type="response",s=0.05)

###Example 2

#data(Colon)
#X<-t(Colon$X)
#y_tmp<-Colon$Y
#y<-ifelse(y_tmp==1,1,0)
#f1=SubLasso(X,y,nfold=10)
#subset=f1$selname
#f2=SubLasso(X,y,subset,nfold=10)
#subset=row.names(X)[30:40]
#f3=SubLasso(X,y,subset,nfold=10)

SubLasso documentation built on May 29, 2017, 8:45 p.m.