ttestP: Compute t test P value
In nickytong/GenAnalysis: Genomic Analysis

Compute t test P value

1
2
3

ttestP(data, class, switchXY = FALSE, minN, more = FALSE,
  var.equal = TRUE, levels = NULL, base = 2, paired = FALSE,
  FCbyRatio = FALSE)

`data`	a vector of response
`class`	a vector of membership, binary variable. Internally, class is converted to a vector during computation
`switchXY`	sometimes we have data as binary and class as continuous (for some reason, we do not want to change this in calling ttestP) and we need to swtich the role of x and y; switchXY when specified as TRUE, it will treat data as class and class as data.
`minN`	minimum sample size (in each group) to carry out calculation
`more`	whether to include additional info i.e. FC, mean difference
`var.equal`	same as in t.test
`levels`	the user can specify the required ordering c(C1, C2): mean diff is C2-C1. If not specified, (1) the class is a factor, levels would be extracted; (2) otherwise, use alphabetical order from class. Notice that levels should be a vector of character strings. This means if true level is 0 and 1, levels should be specified as levels=c('0', '1') rather than levels=c(0, 1). This is due to the bug in gdata::reorder.factor that does not recognize new.order=c(0, 1).
`base`	base used to calculate fold change using FC(); default is 2, which is typical for microarray and RNAseq data.
`FCbyRatio`	sometimes the values are in original scale all positive, it is possible to compute the ratio and corresponding FC. Notice, this will override the FC calculation already computed through specified base.
`whether`	use paired t test (assumes data is special being paired)