train_NBC: NBC Training Function

In naomiwilson/gnominator: Naive Bayes Classifier for scoring human microbiomes

NBC Training Function

Description

NBC Training Function

Usage

train_NBC(
  asv_table,
  sample_data,
  minimum_detection,
  min_rel_abund,
  sigma_for_alpha = 10,
  sigma_for_beta = 30
)

Arguments

asv_table	double: rows= ASV names, cols=sample names
sample_data	character matrix: column 1 = sample names; column 2 = class labels (e.g. "asthmatic" and "healthy")
minimum_detection	integer is how many samples in which the ASV has to be present in to be included in the training
min_rel_abund	integer sequencing detection limit - the relative abundance value at which we no longer trust the ASV is truly present
sigma_for_alpha	sigma value for alpha MAP prior (default=10)
sigma_for_beta	sigma value for beta MAP prior (default=30)

Value

nested matrix containing 1) parameters for class 1, 2) class 2, 3) both classes, 4) the class names in order, and 5) probability of class 1

naomiwilson/gnominator documentation built on Feb. 27, 2023, 9:39 p.m.