summarizeCpGs: summarize function that choose cpgs based on different...

In lizhongliu1996/PredictMisc: package that use three methods to predict methylation regions

Description

summarize function that choose cpgs based on different methods

Usage

summarizeCpGs(clust_ls, train_df, test_df, selectMethod)

Arguments

clust_ls	list of cpgs, each slot contains few cpgs
train_df	data frame that each row is a train samples
test_df	data frame that each row is a test samples
selectMethod	what cpg selection method to use, use full cpgs fullCpGs within cluster or PC1 score getPC1 of cluster or maximum maxCpGs expression score, default set to fullcpgs, it is feasible to write new methods by adding a new function that take in train/test dataset and cpglist then return a list of train and test subset data.

Value

a list contains train data frame and test data frame and number of predictors

Examples

## Not run: 
data(aclust.listDemo)
data(ExampleMvalue_train)
data(ExampleMvalue_test)

test <- summarizeCpGs(
  clust_ls = aclust.listDemo,
  train_df = ExampleMvalue_train[ , -1],
  test_df = ExampleMvalue_test[ , -1],
  selectMethod = fullCpGs
 )

## End(Not run)

lizhongliu1996/PredictMisc documentation built on Aug. 23, 2019, 5:55 a.m.