BI_bootstrap.lasso: bulid a bootstrap lasso model via exprs and design objects

In shijianasdf/BasicBioinformaticsAnalysisFromZhongShan: Launch Usual Clinical tool for Kind Yield

Description

bulid a bootstrap lasso model for binary induced variable via exprs and design objects and output a ROCList object

Usage

bootstrap.lasso(dds,
                transformation = "normTransform",
                expr.matrix=NULL,
                design,
                select,
                contrast = "N.status",
                contrast.control = "N0",
                k=10,
                R=2000,
                seed = 2018,seed.range = 1:15000,
                each.size = 15,
                optimize.method = "1se",
                show.music = T)

View(uniqueModel(lassoModel$modeldata))

result.lassoModel(lassoModel,
                  position = 1,
                  names = "test",
                  dig = 5)

Arguments

dds	DESeq2 dds object.If it was NULL,the expr.matrix should not be null.
transformation	one of "vst" and "normTransform".
expr.matrix	a log scale expression matrix.
design	a design object.
select	select genes names.
contrast	the colnames of the contrast
contrast.control	the control of the contrast
k	nfolds
R	2000.bootstrap
seed	random seed
seed.range	the range of random seeds
each.size	the size of each group in one bootstrap
optimize.method	one of "min" and "1se".Default is "1se".
show.music	whether show music at the end of the function

Note

result.lassoModel() function also support the result from bootstrap.lasso

Author(s)

Weibin Huang<[e-mail]>

Examples

## a virtual not-running programes
ROCList <- bootstrap.lasso(expr.matrix=model1.fpkm,
                           design = design.model1,
                           select = LM.genes,
                           contrast = "N.status",
                           contrast.control = "N0",
                           k=10,
                           R=2000,
                           seed = 2018,seed.range = 1:15000,
                           each.size = 15,
                           optimize.method = "1se")

shijianasdf/BasicBioinformaticsAnalysisFromZhongShan documentation built on Jan. 3, 2020, 10:08 p.m.