ReferenceIntermediateReferencev3.10Node2VecSupported Graph Characteristics Directed edges Undirected edges Homogeneous vertex types Heterogeneous vertex types Algorithm link: Node2Vec Node2Vec is a legacy node embeddDiscover
ReferenceIntermediateReferencev1.8Data SplittersThese classes contain perform data splitting. Make sure to create the appropriate attributes in the graph before using these splitters. Functions RandomVertexSplitter The split results are storedDiscover
ReferenceReferencev1.8Model Trainer and CallbacksTrain Graph Machine Learning models (such as GraphSAGE and NodePiece) in a concise way. pyTigerGraph offers built-in models that can be used with the Trainer, consuming pyTigerGraph dataloaders. CallDiscover
ReferenceIntermediateReferencev1.8Data LoadersData loaders are classes in the pyTigerGraph Graph Data Science (GDS) module. You can define an instance of each data loader class through a factory function . Requires querywriters user permissions for full functiDiscover
ReferenceIntermediateReferencev3.10k-Means ClusteringSupported Graph Characteristics Unweighted edges Weighted edges Directed edges Undirected edges Homogeneous vertex types Algorithm link: k-Means Clustering IDiscover
ReferenceAdvancedReferencev3.10k-Nearest Neighbors (Cross-Validation Version)k-Nearest Neighbors (kNN) is often used for machine learning. You can choose the value for topK based on your experience, or using cross-validation to optimize the hyperparameters.Discover
ReferenceIntermediateReferencev3.10Fast Random ProjectionSupported Graph Characteristics Unweighted edges Directed edges Undirected edges Homogeneous vertex types Heterogeneous vertex types Algorithm link: Fast RandDiscover
ReferenceIntermediateReferencev1.8PyTorch Geometric TransformsTemporalPyGTransform The TemporalPyGTransform creates a sequence of subgraph batches out of a single batch of data produced by a NeighborLoader or HGTLoader. It assumes that there are datetime attrDiscover
ReferenceIntermediateReferencev1.8Data TransformsThere are currently two sub-modules, nodepiecetransforms , and pygtransforms . The NodePiece transforms help convert batches produced by NodePiece dataloaders into a format used by PyTorch models. The PyG transforms hDiscover
ReferenceReferencev1.8NodePiece ModelsNodePieceMLPForVertexClassification This model is for training an multi-layer perceptron (MLP) on batches produced by NodePiece dataloaders, and transformed by the NodePieceMLPTransform . The architecture is fDiscover
ReferenceReferencev1.8GraphSAGE ModelsGraphSAGEForVertexClassification Use a GraphSAGE model to classify vertices. By default, this model collects ClassficationMetrics , and uses cross entropy as its loss function. init () init (numDiscover