Release Notes: Graph Data Science 1.6.0 Preview

• maxStreakCount configuration parameter is renamed to patience. It is used in the train modes of Node Classification and Link Prediction.
• maxIterations and minIterations configuration parameters are renamed to maxEpochs and minEpochs. It is used in the train modes of Node Classification and Link Prediction.
• windowSize configuration parameters is removed from the train modes of Node Classification and Link Prediction.

• The same effect can be achieved by using gds.degree.mutate and use the mutated property as feature for GraphSAGE training.
• Important: GraphSAGE models persisted with earlier versions of GDS are not compatible with this version.

New features

• New ScaleProperties procedures to transform and scale node properties. Available scalers: Min-max, Max, Mean, Log, Standard Score, L1 Norm, L2 Norm
  • gds.alpha.scaleProperties.stream
  • gds.alpha.scaleProperties.mutate
• Added ability to create new in-memory graphs by filtering existing named graphs based on node and relationship properties with new catalog procedure gds.beta.graph.create.subgraph
• Two new centrality algorithms for influence maximization were contributed by community member @xkitsios
  • gds.alpha.influenceMaximization.celf.stream
  • gds.alpha.influenceMaximization.greedy.stream
• Link Prediction:
  • Added support for storing, loading and publishing Link Prediction models.
  • Added progress logging for gds.alpha.ml.linkPrediction.train and gds.alpha.ml.linkPrediction.predict.
  • Added write and stream modes to gds.alpha.ml.linkPrediction.predict
    • gds.alpha.ml.linkPrediction.stream
    • gds.alpha.ml.linkPrediction.write
  • Added estimate mode for Link Prediction:
    • gds.alpha.ml.linkPrediction.train.estimate
    • gds.alpha.ml.linkPrediction.stream.estimate
    • gds.alpha.ml.linkPrediction.mutate.estimate
    • gds.alpha.ml.linkPrediction.write.estimate
• Node Classification:
  • Added support for storing, loading and publishing Node Classification models.
  • Added support for f1, precision, recall and accuracy per class for NodeClassification.
    • Used with metric(class=<number>) or syntactic sugar metric(class=*).
  • Added write and stream modes to gds.alpha.ml.nodeClassification.predict
    • gds.alpha.ml.nodeClassification.stream
    • gds.alpha.ml.nodeClassification.write
  • Added estimate mode for Node Classification:
    • gds.alpha.ml.nodeClassification.train.estimate
    • gds.alpha.ml.nodeClassification.mutate.estimate
    • gds.alpha.ml.nodeClassification.stream.estimate
    • gds.alpha.ml.nodeClassification.write.estimate
• Added configuration parameter negativeSamplingWeight to gds.alpha.ml.splitRelationships
• Pregel API:
  • Added progress logging to Pregel. This affects HITS, SLLPA, Article Rank, Eigenvector and Page Rank.
  • Added partioning configuration parameter to Pregel procedures, possible options are
    • range (default): each partition (thread) has the same number of nodes
    • degree: each partition (thread) has approximately the same number of relationships
• Added randomSeed configuration parameter to gds.fastRP.* and gds.beta.fastRPExtended.*.
• Node2Vec:
  • Added an optional parameter randomSeed to the Node2Vec procedures, which allows the generation of deterministic random walks.
  • Optimized implementation, decreasing execution time by up to 90%
  • Added support for weights with relationshipWeightProperty
  • Added mutate mode: gds.beta.node2vec.mutate
    • gds.alpha.ml.nodeClassification.mutate.estimate
    • gds.alpha.ml.nodeClassification.stream.estimate
    • gds.alpha.ml.nodeClassification.write.estimate
• Added an optional db parameter to gds.graph.drop which allows dropping graphs from other neo4j databases in a multi-database environment.
• Added scaler configuration parameter to gds.pagerank which allows normalizing the computed scores.
• Added support for orientation in gds.degree procedures.
• Users can now set a threshold for community sizes when writing back to the database:
  • Added minCommunitySize to gds.louvain.write to only write communities that are larger than or equal to a given value.
  • Added minComponentSize to gds.wcc.write to only write components that are larger than or equal to a given value.
• RBAC Support for GDS Enterprise: Users that have been granted the admin role are allowed to operate on other users’ graphs. Those operations are:
  • View other users graphs in gds.graph.list
  • Use other users’ graphs in most algorithm procedures. Some alpha tier procedures and gds.graph.export do not support this feature
  • Drop other users graphs via gds.grap.drop
  • Naming conflicts (between users) can be resolved by using a username parameter in the configuration of the procedure

Bug fixes

• Fixed memory estimation for gds.graph.create.estimate on GDS Enterprise Edition, now estimates take into account enterprise compression format.
• Fixed a bug which caused gds.graph.list and gds.graph.drop to throw an error when specifying a graph with duplicate property keys by failing early.
• Fixed potential ArrayIndexOutOfBoundsException when running gds.triangleCount on a relationship-filtered graph.
• Fixed a bug that can lead to inconsistencies when writing or mutating new relationships created from a label-filtered graph.
• Support concurrency in gds.alpha.ml.linkPrediction.train and gds.alpha.ml.nodeClassification.train.
• Fixed a bug where Alpha similarity algorithms in some cases could fail on division by 0 when writing results back.
• Fixed an issue where gds.graph.drop could take a long time when the graph contained node embeddings.
• Fixed a bug where loading array properties via cypher loading was not possible.
• Fixed an issue where gds.fastRP.* and gds.beta.fastRPExtended.* were not failing on missing relationship properties when executing with weighted relationships.
• Fixed a bug where gds.beta.graphSage.train was failing in the presence of array properties.
• Fixed a bug in Node2Vec where the learning rate was scaled incorrectly.
• Fixed a bug where the number of deleted node properties returned by gds.graph.removeNodeProperties would be wrong.
• Fixed a bug where gds.alpha.scc would sometimes fail with an ArrayIndexOutOfBoundsException.
• Fixed a security bug that allowed restricted users to run procedures that write back to the database.