search

pkimes / sigclust2

tests for statistical significance of clustering
32 stars 6 forks source link

summary(shc_result) reports "euclidean" when specifying custom matmet. #6

Closed mebbert closed 6 years ago

mebbert commented 6 years ago

Hi, I noticed that summary(shc_result) reports dissimilarity = euclidean when I specify a custom matmet. My understanding is that the matmet, vatmet, and metric are mutually exclusive.

I assume this is just a bug in summary.shc and doesn't affect the actual results.

Here are my calls and result:

> shc_result <- shc(t(sig.heat), matmet = function(x){as.dist(1-cor(t(x), method = sample.dist))}, linkage="complete", null_alg = "2means", ci = "2CI")
> summary(shc_result)

shc object created using shc(..)
--------------------------------
Clustering Parameters:
    dissimilarity = euclidean
    linkage = complete
Testing Parameters:
    n_sim = 100
    icovest = 1
    ci = 2CI
    null_alg = 2means
    n_min = 10
    FWER control = FALSE

Really appreciate your help. Let me know if you need additional information.

pkimes commented 6 years ago

@mebbert, thanks for catching this!

You're right - this is just a bug in the summary output. It should be fixed with the update I made in commit ce27658 (and therefore, the current release).

pkimes commented 6 years ago

Closing the issue as resolved. Feel free to re-open if the issue still persists.