rowstats: Apply selected significance test across rows

In matR: Metagenomics Analysis Tools

Description

Apply selected significance test across rows to grouped columns, with optional q-value calculation.

Usage

rowstats(x, ...)

## S3 method for class 'matrix'
rowstats(x, groups, 
  test=c("Kruskal-Wallis", "t-test-paired", "Wilcoxon-paired", "t-test-unpaired",
    "Mann-Whitney-unpaired-Wilcoxon", "ANOVA-one-way"),
  qvalue=FALSE, fdr.level=NULL, ...)

## S3 method for class 'biom'
rowstats(x, groups, ...)

Arguments

x	a matrix-like object (matrix or biom)
groups	a grouping of columns (character or factor)
test	name of statistical test (character)
qvalue	perform qvalue calculation? (logical)
fdr.level	false-discovery rate parameter, passed to qvalue() (numeric)
...	unused

Details

Complete technical documentation is forthcoming. For the current preliminary release, please refer to the examples provided.

Value

Complete technical documentation is forthcoming. For the current preliminary release, please refer to the examples provided.

Author(s)

Kevin P. Keegan and Daniel T. Braithwaite

See Also

stats::t.test, stats::wilcox.test, stats::kruskal.test, stats::anova

Examples

####  Kruskal test applied, for the case of more than two metagenome groups
columns (xx1, "host_common_name")
str (rowstats (xx1, groups="$$host_common_name", test="Kruskal"))

####  force a desired grouping of metagenomes
gg <- columns (xx2, "material") [[1]]
gg
levels (gg) <- levels (gg) [c(1,2,2)]
str (rowstats (xx2, groups=gg, test="t-test-unpaired"))

matR documentation built on May 2, 2019, 6:53 a.m.