Github user vasia commented on the issue:
https://github.com/apache/flink/pull/2564
Providing a flattened tuple is certainly better than having the user
implement the reduce, but I think we should provide separate methods for the
default and custom operations. A projection is a very well-defined operation:
create a graph where there is an edge between any pair of vertices with a
common neighbor in the bipartite graph. If the user wants to apply mappers or
other transformations on the vertices and edges, they can do so afterwards,
using the graph methods. The problem is that with a projections, some
information is lost, e.g. the edge values. For these cases, we can provide a
custom projection method where we give the labels in a flattened tuple as
@greghogan suggested, but I'm afraid the API will look ugly with a Tuple8
there. Another, maybe more friendly solution, would be attaching the labels on
the projection graph edges. What do you think?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
