cuGraph

https://docs.rapids.ai/api/cugraph/nightly/
Admin
cuGraph is a library of GPU-accelerated graph algorithms that integrates into the RAPIDS data...
Tokens:1,012
Snippets:8
Trust Score:10
Update:1 year ago
Show doc for...
Tokens:
Raw
### cugraph_pyg.data.graph_store.GraphStore Initialization

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

Initializes a cuGraph-backed PyG GraphStore implementation. This class is designed to distribute the graph across multiple workers for efficient processing.

```Python
cugraph_pyg.data.graph_store.GraphStore()
```

--------------------------------

### cugraph_pyg.data.feature_store.FeatureStore Initialization

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

Initializes a FeatureStore with feature data stored in WholeGraph WholeMemory. This is a basic implementation adhering to the PyG FeatureStore interface.

```Python
cugraph_pyg.data.feature_store.FeatureStore([...])
```

--------------------------------

### cugraph_pyg.loader.neighbor_loader.NeighborLoader Usage

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

Employs a duck-typed version of torch_geometric.loader.NeighborLoader for loading neighbor data. This facilitates efficient neighbor sampling in graph processing.

```Python
cugraph_pyg.loader.neighbor_loader.NeighborLoader(...)
```

--------------------------------

### cugraph_pyg.sampler.sampler.SampleIterator Batch Creation

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

An iterator that combines output graphs with their corresponding features. It produces final mini-batches suitable for input into GNN models.

```Python
cugraph_pyg.sampler.sampler.SampleIterator(...)
```

--------------------------------

### cugraph_pyg.loader.node_loader.NodeLoader Usage

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

Utilizes a duck-typed version of torch_geometric.loader.NodeLoader for loading node data. This loader is compatible with the cuGraph ecosystem.

```Python
cugraph_pyg.loader.node_loader.NodeLoader(...)
```

--------------------------------

### cugraph_pyg.sampler.sampler.BaseSampler Definition

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

Defines the base class for samplers within the cugraph-pyg library. It serves as a foundation for various sampling strategies.

```Python
cugraph_pyg.sampler.sampler.BaseSampler(...)
```

--------------------------------

### cugraph_pyg.sampler.sampler.HomogeneousSampleReader Usage

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

A subclass of SampleReader specifically designed for reading homogeneous output samples generated by the cuGraph distributed sampler.

```Python
cugraph_pyg.sampler.sampler.HomogeneousSampleReader(...)
```

--------------------------------

### cugraph_pyg.sampler.sampler.SampleReader Iteration

Source: https://docs.rapids.ai/api/cugraph/nightly/api/cugraph/stable/api_docs/cugraph-pyg/cugraph_pyg

An iterator that processes results from the cuGraph distributed sampler. It enables sequential access to sampled graph data.

```Python
cugraph_pyg.sampler.sampler.SampleReader(...)
```
Tokens:
Raw