pick.samples: Pick samples

In graeberlab/small.cell.project: small.cell.project

Description

Reads file with features(PC's) in columns and samples in rows. And an annotation file.You pick a max # of features. Using caret runs a prediction iteratively from 2 to max.features. Picks best predictor with least featurs. Returns samples correctly predicted leaves out those samples incorrectly predicted.

Usage

pick.samples(data.file, max.features = 5, anno.file,
  train.method = "LOOCV", technique = "lda")

Arguments

data.file	Filepath/filename of data matrix
max.features	the maximum number of features to use
anno.file	annotation file, has headers, first column is sample names, 2nd is type
train.method	the method used for training, default is leave one out cross validation, check caret for options
technique	the method used for model building, default is linear discriminant analysis, check caret for options (e.g. svmLinear,

graeberlab/small.cell.project documentation built on May 12, 2019, 5:16 p.m.