PMCMRplus post-hoc Nemenyi's all pairs test ignores alternative?

Question

After a successful Friedman test, I would like to perform post-hoc analysis using the all pairs comparisons Nemenyi's test. Specifically, as I am interested in finding out if the values of some of my samples are greater than the others, I'm trying to apply a one-tailed Nemenyi's test using the PMCMRplus R package with the alternative="greater" parameter. However, it seems to to me that the alternative parameter is always ignored. Example below.

Using the example from the PMCMRplus documentation:

## Sachs, 1997, p. 675
## Six persons (block) received six different diuretics
## (A to F, treatment).
## The responses are the Na-concentration (mval)
## in the urine measured 2 hours after each treatment.
##
y <- matrix(c(
  3.88, 5.64, 5.76, 4.25, 5.91, 4.33, 30.58, 30.14, 16.92,
  23.19, 26.74, 10.91, 25.24, 33.52, 25.45, 18.85, 20.45,
  26.67, 4.44, 7.94, 4.04, 4.4, 4.23, 4.36, 29.41, 30.72,
  32.92, 28.23, 23.35, 12, 38.87, 33.12, 39.15, 28.06, 38.23, 26.65),nrow=6, ncol=6,
dimnames=list(1:6, LETTERS[1:6]))

friedmanTest(y)

I obtain a p-value less than 0.05

Friedman rank sum test

data:  y
Friedman chi-squared = 23.333, df = 5, p-value = 0.0002915

So my understanding is that I can apply post-hoc analysis. I would like to use the Nemenyi's test to perform all possible pairwise comparisons:

## Nemenyi's test
frdAllPairsNemenyiTest(y) 

Results:

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox all-pairs test for a two-way balanced complete block design

data: y

  A      B      C      D      E     
B 0.1880 -      -      -      -     
C 0.0917 0.9996 -      -      -     
D 0.9996 0.3388 0.1880 -      -     
E 0.0395 0.9898 0.9996 0.0917 -     
F 0.0016 0.6363 0.8200 0.0052 0.9400

P value adjustment method: single-step

However, in my case I have to perform a one-tailed test as I want to compare if some samples have greater values than others. Hence:

## Nemenyi's test
frdAllPairsNemenyiTest(y, alternative="greater")

But I get the exact same results as before:

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox all-pairs test for a two-way balanced complete block design

data: y

  A      B      C      D      E     
B 0.1880 -      -      -      -     
C 0.0917 0.9996 -      -      -     
D 0.9996 0.3388 0.1880 -      -     
E 0.0395 0.9898 0.9996 0.0917 -     
F 0.0016 0.6363 0.8200 0.0052 0.9400

P value adjustment method: single-step

And the same happens if I try :

frdAllPairsNemenyiTest(y, alternative="less")

Results:

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox all-pairs test for a two-way balanced complete block design

data: y

  A      B      C      D      E     
B 0.1880 -      -      -      -     
C 0.0917 0.9996 -      -      -     
D 0.9996 0.3388 0.1880 -      -     
E 0.0395 0.9898 0.9996 0.0917 -     
F 0.0016 0.6363 0.8200 0.0052 0.9400

P value adjustment method: single-step

This leads me to believe that the alternative parameter does not work as intended. Am I misinterpreting the results? or am I missing something about the test?

EDIT: the same package offers a frdManyOneNemenyiTest function for many-to-one comparisons using the same test. When using that function the alternative parameter is taken into consideration:

frdManyOneNemenyiTest(y, alternative='greater')
frdManyOneNemenyiTest(y, alternative='less')
frdManyOneNemenyiTest(y)

Results:

> frdManyOneNemenyiTest(y, alternative='greater')

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox-Miller many-to-one test for a two-way balanced complete block design

data: y

  A      
B 0.04119
C 0.01845
D 0.72552
E 0.00747
F 0.00027

P value adjustment method: single-step
alternative hypothesis: greater
> frdManyOneNemenyiTest(y, alternative='less')

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox-Miller many-to-one test for a two-way balanced complete block design

data: y

  A   
B 1.00
C 1.00
D 0.91
E 1.00
F 1.00

P value adjustment method: single-step
alternative hypothesis: less
> frdManyOneNemenyiTest(y)

    Pairwise comparisons using Nemenyi-Wilcoxon-Wilcox-Miller many-to-one test for a two-way balanced complete block design

data: y

  A      
B 0.08243
C 0.03711
D 0.99817
E 0.01475
F 0.00052

P value adjustment method: single-step
alternative hypothesis: two.sided
```

Answer 1

The function frdAllPairsNemenyiTest has no argument alternative. It is only intended to do non-directional pairwise hypotheses tests (i.e. two-sided per default). See also help(frdAllPairsNemenyiTest).