can't reproduce sample in readme

[JohnRenshaw]

We get the following error when we try to reproduce readme example, we are using scala 2.11 jar downloaded from maven. We are running from AWS EMR cluster, with the following version
Spark 2.1.1 (git revision 173acdf) built for Hadoop 2.7.3-amzn-2

val clusters: Dataset[Assignment] =
MCL.train(graph).assignments

:37: error: type mismatch;
found : org.apache.spark.rdd.RDD[org.apache.spark.mllib.clustering.Assignment]
required: org.apache.spark.sql.Dataset[org.apache.spark.mllib.clustering.Assignment]
Error occurred in an application involving default arguments.
MCL.train(graph).assignments

[joandre]

Hi,

On maven there is no jar for Spark upper than 1.6.1. Indeed until this issue is not solved, I don't want to publish a new version on spark package. I have just updated the code so you can use it under 2.1.1 if you compile it locally. Let me know if it is ok for you.