Clustering with PAM

Hello my dear friends,
i'm running a script for clustering using PAM method. but unfortunately the data is too large to be processed. Tried Clara to take a sample from my data and run the clustering script but this message appeared.

Code used :

fviz_nbclust(DF, clara, method = "silhouette") + theme_classic()

clara.res = clara(DF, 5, samples = 50, pamLike = TRUE)

Error appeared :
Error: cannot allocate vector of size 31.0 Gb
How can i solve this problem. Is a solution available at R on my laptop or i must go to a cloud ??? and if so , Which cloud do you recommend to use ??
I Much appreciate your help please
thanks :slight_smile:

I'm not familiar with either of these methods. However, it seems like you don't sample (with Clara) until after you use the full data and get your error message relating to too many resources required. Perhaps reorder your commands would address this issue ?

Apologies if I misunderstood

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.