[GraphBolt][Doc] update impl (#6418)

docs/source/api/python/dgl.graphbolt.rst

@@ -56,8 +56,6 @@ Standard Implementations
 
     OnDiskDataset
     BuiltinDataset
-    OnDiskTask
-    OnDiskMetaData
     CSCSamplingGraph
     UniformNegativeSampler
     NeighborSampler

python/dgl/graphbolt/impl/basic_feature_store.py

@@ -10,7 +10,7 @@ __all__ = ["BasicFeatureStore"]
 
 
 class BasicFeatureStore(FeatureStore):
-    r"""Basic feature store."""
+    r"""A basic feature store to manage multiple features for access."""
 
     def __init__(self, features: Dict[Tuple[str, str, str], Feature]):
         r"""Initiate a basic feature store.

python/dgl/graphbolt/impl/csc_sampling_graph.py

@@ -16,6 +16,17 @@ from ..sampling_graph import SamplingGraph
 from .sampled_subgraph_impl import SampledSubgraphImpl
 
 
+__all__ = [
+    "GraphMetadata",
+    "CSCSamplingGraph",
+    "from_csc",
+    "load_from_shared_memory",
+    "load_csc_sampling_graph",
+    "save_csc_sampling_graph",
+    "from_dglgraph",
+]
+
+
 class GraphMetadata:
     r"""Class for metadata of csc sampling graph."""
 
@@ -76,7 +87,7 @@ class GraphMetadata:
 
 
 class CSCSamplingGraph(SamplingGraph):
-    r"""Class for CSC sampling graph."""
+    r"""A sampling graph in CSC format."""
 
     def __repr__(self):
         return _csc_sampling_graph_str(self)

python/dgl/graphbolt/impl/neighbor_sampler.py

@@ -8,9 +8,13 @@ from ..utils import unique_and_compact_node_pairs
 from .sampled_subgraph_impl import SampledSubgraphImpl
 
 
+__all__ = ["NeighborSampler", "LayerNeighborSampler"]
+
+
 @functional_datapipe("sample_neighbor")
 class NeighborSampler(SubgraphSampler):
-    """
+    """Sample neighbor edges from a graph and return a subgraph.
+
     Neighbor sampler is responsible for sampling a subgraph from given data. It
     returns an induced subgraph along with compacted information. In the
     context of a node classification task, the neighbor sampler directly
@@ -121,7 +125,8 @@ class NeighborSampler(SubgraphSampler):
 
 @functional_datapipe("sample_layer_neighbor")
 class LayerNeighborSampler(NeighborSampler):
-    """
+    """Sample layer neighbor edges from a graph and return a subgraph.
+
     Sampler that builds computational dependency of node representations via
     labor sampling for multilayer GNN from the NeurIPS 2023 paper
     `Layer-Neighbor Sampling -- Defusing Neighborhood Explosion in GNNs

python/dgl/graphbolt/impl/ondisk_dataset.py

@@ -279,13 +279,13 @@ class OnDiskTask:
 
 
 class OnDiskDataset(Dataset):
-    """An on-disk dataset.
+    """An on-disk dataset which reads graph topology, feature data and
+    Train/Validation/Test set from disk.
 
-    An on-disk dataset is a dataset which reads graph topology, feature data
-    and TVT set from disk. Due to limited resources, the data which are too
-    large to fit into RAM will remain on disk while others reside in RAM once
-    ``OnDiskDataset`` is initialized. This behavior could be controled by user
-    via ``in_memory`` field in YAML file.
+    Due to limited resources, the data which are too large to fit into RAM will
+    remain on disk while others reside in RAM once ``OnDiskDataset`` is
+    initialized. This behavior could be controled by user via ``in_memory``
+    field in YAML file.
 
     A full example of YAML file is as follows:
 
@@ -477,12 +477,10 @@ class OnDiskDataset(Dataset):
 
 
 class BuiltinDataset(OnDiskDataset):
-    """GraphBolt builtin on-disk dataset.
+    """A utility class to download built-in dataset from AWS S3 and load it as
+    ``OnDiskDataset``.
 
-    This class is used to help download datasets from DGL S3 storage and load
-    them as ``OnDiskDataset``.
-
-    Available builtin datasets include:
+    Available built-in datasets include:
 
     **ogbn-mag**
         The ogbn-mag dataset is a heterogeneous network composed of a subset of

python/dgl/graphbolt/impl/sampled_subgraph_impl.py

@@ -8,10 +8,12 @@ import torch
 from ..base import etype_str_to_tuple
 from ..sampled_subgraph import SampledSubgraph
 
+__all__ = ["SampledSubgraphImpl"]
+
 
 @dataclass
 class SampledSubgraphImpl(SampledSubgraph):
-    r"""Class for sampled subgraph specific for CSCSamplingGraph.
+    r"""Sampled subgraph of CSCSamplingGraph.
 
     Examples
     --------

python/dgl/graphbolt/impl/torch_based_feature_store.py

@@ -12,7 +12,7 @@ __all__ = ["TorchBasedFeature", "TorchBasedFeatureStore"]
 
 
 class TorchBasedFeature(Feature):
-    r"""Torch based feature."""
+    r"""A wrapper of pytorch based feature."""
 
     def __init__(self, torch_feature: torch.Tensor):
         """Initialize a torch based feature store by a torch feature.
@@ -130,7 +130,7 @@ class TorchBasedFeature(Feature):
 
 
 class TorchBasedFeatureStore(BasicFeatureStore):
-    r"""Torch based feature store."""
+    r"""A store to manage multiple pytorch based feature for access."""
 
     def __init__(self, feat_data: List[OnDiskFeatureData]):
         r"""Load feature stores from disk.

python/dgl/graphbolt/impl/uniform_negative_sampler.py

@@ -4,14 +4,16 @@ from torch.utils.data import functional_datapipe
 
 from ..negative_sampler import NegativeSampler
 
+__all__ = ["UniformNegativeSampler"]
+
 
 @functional_datapipe("sample_uniform_negative")
 class UniformNegativeSampler(NegativeSampler):
-    """
-    Negative samplers randomly select negative destination nodes for each
-    source node based on a uniform distribution. It's important to note that
-    the term 'negative' refers to false negatives, indicating that the sampled
-    pairs are not ensured to be absent in the graph.
+    """Sample negative destination nodes for each source node based on a uniform
+    distribution.
+
+    It's important to note that the term 'negative' refers to false negatives,
+    indicating that the sampled pairs are not ensured to be absent in the graph.
     For each edge ``(u, v)``, it is supposed to generate `negative_ratio` pairs
     of negative edges ``(u, v')``, where ``v'`` is chosen uniformly from all
     the nodes in the graph.