[Docs] Fix docstrings and documentations (#1420)

* fix docs of to_block

* fix

* oops

* more fixes

docs/source/api/python/data.rst

@@ -246,7 +246,7 @@ Dataset Classes
 
 If your dataset is stored in a ``.csv`` file, you may find it helpful to use
 
-.. autoclass:: dgl.data.chem.CSVDataset
+.. autoclass:: dgl.data.chem.MoleculeCSVDataset
     :members: __getitem__, __len__
 
 Currently four datasets are supported:

docs/source/api/python/index.rst

@@ -15,6 +15,7 @@ API Reference
    propagate
    udf
    sampler
+   sampling
    data
    transform
    graph_store

docs/source/api/python/sampling.rst

+.. _api-sampling:
+
+dgl.sampling
+=================================
+
+.. automodule:: dgl.sampling
+
+Sampling algorithms on graphs.
+
+Random walk sampling functions
+------------------------------
+
+.. autosummary::
+    :toctree: ../../generated/
+
+    random_walk
+    pack_traces
+
+Neighbor sampling functions
+---------------------------
+
+.. autosummary::
+    :toctree: ../../generated/
+
+    sample_neighbors
+    select_topk
+
+
+Builtin sampler classes for more complicated sampling algorithms
+----------------------------------------------------------------
+.. autoclass:: RandomWalkNeighborSampler
+.. autoclass:: PinSAGESampler

python/dgl/sampling/pinsage.py

@@ -178,6 +178,7 @@ class PinSAGESampler(RandomWalkNeighborSampler):
     Examples
     --------
     Generate a random bidirectional bipartite graph with 3000 "A" nodes and 5000 "B" nodes.
+
     >>> g = scipy.sparse.random(3000, 5000, 0.003)
     >>> G = dgl.heterograph({
     ...     ('A', 'AB', 'B'): g,
@@ -185,10 +186,12 @@ class PinSAGESampler(RandomWalkNeighborSampler):
 
     Then we create a PinSAGE neighbor sampler that samples a graph of node type "A".  Each
     node would have (a maximum of) 10 neighbors.
+
     >>> sampler = dgl.sampling.PinSAGESampler(G, 'A', 'B', 3, 0.5, 200, 10)
 
     This is how we select the neighbors for node #0, #1 and #2 of type "A" according to
     PinSAGE algorithm:
+
     >>> seeds = torch.LongTensor([0, 1, 2])
     >>> frontier = sampler(seeds)
     >>> frontier.all_edges(form='uv')

python/dgl/transform.py

@@ -884,21 +884,21 @@ def to_block(g, dst_nodes=None, include_dst_in_src=True):
     >>> g = dgl.bipartite([(0, 1), (1, 2), (2, 3)], utype='A', vtype='B')
 
     If you don't specify any node of type A on the right hand side, the node type ``A``
-    in the block would have zero nodes.
+    in the block would have zero nodes on the DST side.
     >>> block = dgl.to_block(g, {'B': torch.LongTensor([3, 2])})
-    >>> block.number_of_nodes('A')
+    >>> block.number_of_dst_nodes('A')
     0
-    >>> block.number_of_nodes('B')
+    >>> block.number_of_dst_nodes('B')
     2
-    >>> block.nodes['B'].data[dgl.NID]
+    >>> block.dstnodes['B'].data[dgl.NID]
     tensor([3, 2])
 
     The left hand side would contain all the nodes on the right hand side:
-    >>> block.nodes['B'].data[dgl.NID]
+    >>> block.srcnodes['B'].data[dgl.NID]
     tensor([3, 2])
 
     As well as all the nodes that have connections to the nodes on the right hand side:
-    >>> block.nodes['A'].data[dgl.NID]
+    >>> block.srcnodes['A'].data[dgl.NID]
     tensor([2, 1])
     """
     if dst_nodes is None: