[Doc] User guide section 5 and 6 (#2029)

* [Doc] User guide section 5 and 6

* oops

* fix

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

@@ -78,6 +78,21 @@ Using Node/edge features
     DGLHeteroGraph.local_var
     DGLHeteroGraph.local_scope
 
+Using Node/edge features for blocks
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Please refer to :ref:`guide-minibatch` for the definition of blocks.
+
+.. autosummary::
+    :toctree: ../../generated/
+
+    DGLHeteroGraph.number_of_src_nodes
+    DGLHeteroGraph.number_of_dst_nodes
+    DGLHeteroGraph.srcnodes
+    DGLHeteroGraph.srcdata
+    DGLHeteroGraph.dstnodes
+    DGLHeteroGraph.dstdata
+
 Transforming graph
 ------------------
 
@@ -88,6 +103,7 @@ Transforming graph
     DGLHeteroGraph.edge_subgraph
     DGLHeteroGraph.node_type_subgraph
     DGLHeteroGraph.edge_type_subgraph
+    DGLHeteroGraph.__getitem__
 
 Computing with DGLHeteroGraph
 -----------------------------

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

@@ -30,6 +30,8 @@ General collating functions
 .. autoclass:: EdgeCollator
     :members: dataset, collate
 
+.. _api-dataloading-neighbor-sampling:
+
 Neighborhood Sampling Classes
 -----------------------------
 
@@ -45,7 +47,7 @@ Uniform Node-wise Neighbor Sampling (GraphSAGE style)
 .. autoclass:: MultiLayerNeighborSampler
     :members: sample_frontier
 
-.. _negative-sampling:
+.. _api-dataloading-negative-sampling:
 
 Negative Samplers for Link Prediction
 -------------------------------------

docs/source/api/python/dgl.rst

@@ -22,6 +22,8 @@ Graph Create Ops
     knn_graph
     segmented_knn_graph
 
+.. _api-subgraph-extraction:
+
 Subgraph Extraction Routines
 -------------------------------------
 

docs/source/api/python/nn.mxnet.rst

@@ -186,6 +186,15 @@ Set2Set
     :members:
     :show-inheritance:
 
+Heterogeneous Graph Convolution Module
+----------------------------------------
+
+HeteroGraphConv
+~~~~~~~~~~~~~~~
+
+.. autoclass:: dgl.nn.mxnet.HeteroGraphConv
+   :members:
+   :show-inheritance:
 
 Utility Modules
 ----------------------------------------

docs/source/api/python/nn.pytorch.rst

@@ -222,6 +222,16 @@ SetTransformerDecoder
     :members:
     :show-inheritance:
 
+Heterogeneous Graph Convolution Module
+----------------------------------------
+
+HeteroGraphConv
+~~~~~~~~~~~~~~~
+
+.. autoclass:: dgl.nn.pytorch.HeteroGraphConv
+   :members:
+   :show-inheritance:
+
 Utility Modules
 ----------------------------------------
 

docs/source/api/python/nn.tensorflow.rst

@@ -106,3 +106,13 @@ GlobalAttentionPooling
 .. autoclass:: dgl.nn.tensorflow.glob.GlobalAttentionPooling
     :members:
     :show-inheritance:
+
+Heterogeneous Graph Convolution Module
+----------------------------------------
+
+HeteroGraphConv
+~~~~~~~~~~~~~~~
+
+.. autoclass:: dgl.nn.tensorflow.HeteroGraphConv
+   :members:
+   :show-inheritance:

docs/source/guide/data.rst

+.. _guide-data-pipeline:
 
 Graph data input pipeline in DGL
 ==================================

docs/source/guide/message.rst

+.. _guide-message-passing:
+
 Message Passing
 ===============
 
@@ -288,4 +290,4 @@ string represents the cross type reducer. The reducer can be one of
     # Trigger message passing of multiple types.
     G.multi_update_all(funcs, 'sum')
     # return the updated node feature dictionary
-    return {ntype : G.nodes[ntype].data['h'] for ntype in G.ntypes}
\ No newline at end of file
+    return {ntype : G.nodes[ntype].data['h'] for ntype in G.ntypes}

docs/source/guide/nn.rst

-Build DGL NN Module
-===================
+.. _guide-nn:
+
+Building DGL NN Module
+======================
 
 DGL NN module is the building block for your GNN model. It inherents
-from `Pytorch’s NN Module <https://pytorch.org/docs/1.2.0/_modules/torch/nn/modules/module.html>`__, `MXNet Gluon’s NN Blcok  <http://mxnet.incubator.apache.org/versions/1.6/api/python/docs/api/gluon/nn/index.html>`__ and `TensorFlow’s Keras
+from `Pytorch’s NN Module <https://pytorch.org/docs/1.2.0/_modules/torch/nn/modules/module.html>`__, `MXNet Gluon’s NN Block  <http://mxnet.incubator.apache.org/versions/1.6/api/python/docs/api/gluon/nn/index.html>`__ and `TensorFlow’s Keras
 Layer <https://www.tensorflow.org/api_docs/python/tf/keras/layers>`__, depending on the DNN framework backend we are using. In DGL NN
 module, the parameter registration in construction function and tensor
 operation in forward function are the same with the backend framework.
@@ -264,8 +266,7 @@ The code actually does message passing and reducing computing. This part
 of code varies module by module. Note that all the message passings in
 the above code are implemented using ``update_all()`` API and
 ``built-in`` message/reduce functions to fully utilize DGL’s performance
-optimization as described in the `Message Passing User Guide
-Section <https://docs.dgl.ai/guide/message.html>`__.
+optimization as described in :ref:`guide-message-passing`.
 
 Update feature after reducing for output
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

python/dgl/dataloading/dataloader.py

@@ -435,7 +435,7 @@ class EdgeCollator(Collator):
           or a dictionary of edge types and such pairs if the graph is heterogenenous.
 
         A set of builtin negative samplers are provided in
-        :ref:`the negative sampling module <negative-sampling>`.
+        :ref:`the negative sampling module <api-dataloading-negative-sampling>`.
 
     Examples
     --------