fit.and.evaluate: Fits the QC data to distributions and returns the KS test...

In OmicsQC: Nominating Quality Control Outliers in Genomic Profiling Studies

Fits the QC data to distributions and returns the KS test result and BIC score

Description

This function takes the accumulated QC scores, a vector of distributions and a trimming factor. It then returns the results for each distribution in a dataframe. This function supports the following distributions:

• 'weibull'

• 'norm'

• 'gamma'

• 'exp'

• 'lnorm'

• 'cauchy'

• 'logis'

Usage

fit.and.evaluate(
  quality.scores,
  distributions = c("weibull", "norm", "gamma", "exp", "lnorm", "cauchy", "logis"),
  trim.factor = 0.05
)

Arguments

quality.scores	The accumulated QC scores, the output of accumulate.zscores
distributions	A vector of distributions to fit and test
trim.factor	The fraction of extremes on each end to trim before fitting

Value

A dataframe of the results with the following columns

distribution

Name of the fitted distribution

KS.rejected

Whether the Kolmogorov-Smirnov test rejects the fit; see fitdistrplus::gofstat - kstest

BIC.value

Bayesian Information Criterion

OmicsQC documentation built on May 29, 2024, 11:14 a.m.