You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've been attempting to implement PAM in Julia for the past few days and I have a successful implementation that you can view here, along with some cluster quality comparisons (via silhouette scores). To sum it up, my PAM.jl package achieves higher silhouette scores than kmedoids (at the expense of performance) but not as high as those achieved by the pam function implemented in R's cluster package.
I'm opening this issue to see if there's any interest in adapting the code to make it work with Clustering.jl. If so, I'll work on it and submit a PR (WARNING: I'm inexperienced at package development). Also, for reference, I've found some literature that discusses parallel implemenetations of PAM:
I probably don't have the experience/Julia knowledge to implement a better version than what I have here but I'm more than willing to invest time into it if there's interest in adapting for Clustering.jl and implementing a faster/parallel version.
The text was updated successfully, but these errors were encountered:
I've been attempting to implement PAM in Julia for the past few days and I have a successful implementation that you can view here, along with some cluster quality comparisons (via silhouette scores). To sum it up, my PAM.jl package achieves higher silhouette scores than
kmedoids
(at the expense of performance) but not as high as those achieved by thepam
function implemented in R'scluster
package.I'm opening this issue to see if there's any interest in adapting the code to make it work with
Clustering.jl
. If so, I'll work on it and submit a PR (WARNING: I'm inexperienced at package development). Also, for reference, I've found some literature that discusses parallel implemenetations of PAM:I probably don't have the experience/Julia knowledge to implement a better version than what I have here but I'm more than willing to invest time into it if there's interest in adapting for
Clustering.jl
and implementing a faster/parallel version.The text was updated successfully, but these errors were encountered: