the pyspark ml library does not provide any clustering methods for K-Medoids. So my question is, how can one apply K-Medoids in a pyspark context?
1 Answer
There is a k-medoids clustering for PySpark at spark-packages.org/package/tdebatty/spark-kmedoids and the source code is github.com/tdebatty/spark-kmedoids.
It can be installed with:
> $SPARK_HOME/bin/spark-shell --packages tdebatty:spark-kmedoids:0.1.2
-
$\begingroup$ Hmm. There is no documentation and outdated gitrepo. Sure that it can handle Python? $\endgroup$ Feb 2, 2020 at 21:00