I'm pretty sure that this is caused by the older version of Vega that has been shipped in Kibana so far. The future release of 7.9 will upgrade to the latest.
I am assuming you tried to recreate this using your own data? If so, it could be with something they don’t really explain here or in the tutorials. Basically, if you look at the data set they use, the Les Miserables characters, you will see that the characters have ids and that those ids are used in the source - target of the links. Now the documentation says that you can name the ID field in the path transform but I have yet to get that to work in or outside of that editor. Instead, I recommend that you call your relationships ONLY in a data call. You then use that data to create tons of new data calls:
pull all source nodes
pull all target nodes
merge and aggregate the source and target nodes so you have ONE list of node names. Make sure these are unique so if you have to make them composite, like name-source-type, then do it.
use the identifier transform to assign them all unique ids. This is your final nodes data set that you will use in the marks.
call your original Paths and perform a lookup with the final nodes call to add each source’s ID and each target’s ID fields. This will give you a final field in source and target as ID which the chart needs in the path transform. Make sure this field is called “id”.
Give that highly complex setup a whirl, I guarantee that it will work. If not, please respond here with some of the code you tried and I will try to help.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.