search

schuyler-smith / phylosmith

functions for analyzing phyloseq data
https://schuyler-smith.github.io/phylosmith/
GNU General Public License v3.0
26 stars 3 forks source link

Which distance is being used in nmds, pcoa and t-SNE? #7

Closed FilipeMatteoli closed 4 years ago

FilipeMatteoli commented 4 years ago

Hello, thanks for the package. I am struggling trying to find out which distance the package is using in clustering analysis? Usually, we need to specify the distance, but phylosmith has no option for this, so what is being calculated under the hood?

schuyler-smith commented 4 years ago

Hi! Sorry, I didn't get a notification for this message in my e-mail.. and I'm not as good as I should be about checking github notifications..

This is a very good question. I overlooked adding the option to change the distance method. I have it written to use the Bray-Curtis distance, I definitely should have at least wrote that in the vignette, but I did not, so thank you for pointing this out. I will add an argument to change the distance method for these functions this week!

Thanks so much!

schuyler-smith commented 4 years ago

@FilipeMatteoli I just pushed an updated version with a 'method' argument for each of them. By default it uses Bray-Curtis distance ('bray'), but can now take any of “euclidian”, “manhattan”, “canberra”, “bray”, “kulczynski”, “gower”, “morisita”, “horn”, “mountford”, “jaccard”, “raup”, “binomial”, “chao”, “altGower”, “cao”, “mahalanobis”, “clark”

Please let me know if there is any other feedback on this!

FilipeMatteoli commented 4 years ago

Hello @schuyler-smith great news I will try it soon! I always enjoy testing how different metrics affect the clustering as this can reveal hidden patterns. Thanks for this.