Moved from GitHub dgraph/2517
Posted by mbande:
As Neo4j has implemented graph algorithms such as PageRank, is there any plan to support these algorithms?
Or provide some documentation(Best Practices) on how to do that on top of dgraph in a distributed manner?
MichelDiz commented :
Can be implemented but need help for (PRs are welcome). Actually we have one algo already supported
https://docs.dgraph.io/query-language#k-shortest-path-queries
And most of the algorithms could be worked via query. It would certainly be more complex, but for anyone who already understands GraphQL + - (and RDF) and knows the algorithms from their head. You’ll do well doing a mix with GraphQL + - and application (yours) level.
We have to keep in mind that GraphQL + - is very different from that used in Neo4j. If we wanna give some support, you have to think about the lang as a whole.
See:
https://github.com/dgraph-io/dgraph/commit/c9253021e7d29751d3ba3bbdd45914c83c7d4a2d
https://github.com/dgraph-io/dgraph/commit/b52041fa9fd9a52f997f2ed0e3c4a254509c2327
https://github.com/dgraph-io/dgraph/commit/d471f6df98f7981fe6f6ca88300c557b3d485ccd
potatomasterrace commented :
is adding some reflexivity based syntatic sugar on classes (django orm style) in GitHub - dgraph-io/pydgraph: Official Dgraph Python client provide a better way to make complex algorithms application level only code ?
python seems more fitting for datascience/algorithms