suretype_model: Create model and perform classification with RF-GDA in one...

In Meiomap/SureTypeSCR: Library for (single-cell) SNP array processing

Description

Function is almost always used with in connection with group_by(invidiual) and nest() to create a classification model per group (typically one individual)

Usage

suretype_model(df,individual, clf,.sclist=NULL)))

Arguments

df	Basic dataframe with precalculated M and A features
individual	sample name to be classified
clf	path to the classifier - currently support for Random Forest
.sclist	samples to be processed stored in list(), default is NULL

meaning all samples in df will be classified

Value

returns data frame with two columns corresponding to the classification results: rfgda_score and rf_score, while rfgda_score is score calculated by cascade RF-GDA algorithm and rf_score is score calculated by RF.

Examples

  setwd(system.file(package='SureTypeSCR'))
  samplesheet=system.file('files/GSE19247_example.csv',package='SureTypeSCR')
  manifest=system.file('files/HumanCytoSNP-12v2_H.bpm',package='SureTypeSCR')
  cluster=system.file('files/HumanCytoSNP-12v2_H.egt',package='SureTypeSCR')
  clf=system.file("files/rf.clf",package="SureTypeSCR")
  
  #df_model = df %>%
  #calculate_ma() %>%
  #group_by(individual) %>%
  #nest() %>%
  #mutate(model=map(data , function(df) suretype_model(df,individual, clf,.sclist=list('gsm477563')))

Meiomap/SureTypeSCR documentation built on Dec. 17, 2021, 3:22 a.m.