0

I am using K-means clustering on a dataset with shape (563, 207383) via Dask-K-means (CPU based), and am getting the following error:

"Dask K-means Exception heppen Too many indices for array"

But when I use RapidsAI dask_k-means (GPU Based) it works fine on the same dataset. Could there be a bug in Dask-Kmeans, or do I need to set some configurations before using it?

Thanks in advance.

Vivek kala
  • 23
  • 3

1 Answers1

1

I had the same problem. Upgrade to 2021.04.0 solved the issue.

Yacine Mohamed Tidadini
  • 51
  • 1