getWilcoxon & getTtest

[TuomasBorman]

Create getWilcoxon() and getTtest() functions. They should have the following structure:

# SummarizedExperiment method
getWilcoxon <- function(x, formula, ...){
    tse <- .support_altexp_hiddenly()
    .check_formula(tse, formula)
    df <- extract_data(tse, formula)
    res <- getWilcoxon(df, formula)
    return(res)
}

The function should

1. Hiddenly support altExp which might be useful sometimes. See: https://github.com/microbiome/miaTime/blob/3f90283afeed3475082b64726eed357f73e06a09/R/getBimodality.R#L92
2. Check that the input is in correct format, see: Check input #3
3. Extract data as a long table: Extract data #4
4. Call data.frame method

# data.frame method
getWilcoxon <- function(df, formula, ...){
    .check_formula_from_df(df, formula)
    res <- .run_wilcoxon(df)
    return(res)
}

We could also support data.frame as it does not cause any extra work. The function should

1. Check that the data includes all the variables in formula
2. Run Wilcoxon test
3. Return result.

We can decide the output format in the future when we have implemented more methods. The we know what would be the best way to harmonize these. Also adding add* functions is task for the future.

rstatix package provides useful functions.

[TuomasBorman]

For paired samples, there could be for instance paired argument. It could specify sample grouping.

[TuomasBorman]

For paired samples, there could be for instance paired argument. It could specify sample grouping.

As paired is used as logical value in stats, for instance, this argument could be pair.by instead.