Use real arguments instead of **kwargs where possible

[huonw]

Describe the bug

Many of the layers have layer-specific required or optional arguments that are passed by manipulating **kwargs, e.g., in GCN:

stellargraph/stellargraph/layer/gcn.py

Lines 54 to 99 in 2c513c1

	Args:
	units (int): dimensionality of output feature vectors
	activation (str or func): nonlinear activation applied to layer's output to obtain output features
	use_bias (bool): toggles an optional bias
	final_layer (bool): If False the layer returns output for all nodes,
	if True it returns the subset specified by the indices passed to it.
	kernel_initializer (str or func): The initialiser to use for the weights;
	defaults to 'glorot_uniform'.
	kernel_regularizer (str or func): The regulariser to use for the weights;
	defaults to None.
	kernel_constraint (str or func): The constraint to use for the weights;
	defaults to None.
	bias_initializer (str or func): The initialiser to use for the bias;
	defaults to 'zeros'.
	bias_regularizer (str or func): The regulariser to use for the bias;
	defaults to None.
	bias_constraint (str or func): The constraint to use for the bias;
	defaults to None.
	"""
	def __init__(
	self, units, activation=None, use_bias=True, final_layer=False, **kwargs
	):
	if "input_shape" not in kwargs and "input_dim" in kwargs:
	kwargs["input_shape"] = (kwargs.get("input_dim"),)
	self.units = units
	self.activation = activations.get(activation)
	self.use_bias = use_bias
	self.final_layer = final_layer
	self._get_regularisers_from_keywords(kwargs)
	super().__init__(**kwargs)
	def _get_regularisers_from_keywords(self, kwargs):
	self.kernel_initializer = initializers.get(
	kwargs.pop("kernel_initializer", "glorot_uniform")
	)
	self.kernel_regularizer = regularizers.get(
	kwargs.pop("kernel_regularizer", None)
	)
	self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
	self.bias_initializer = initializers.get(
	kwargs.pop("bias_initializer", "zeros")
	)
	self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
	self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))

For a specific example, the kernel_initializer argument defaults to "glorot_uniform"; the current set up is unfortunate because:

• this default has to be specifically and manually documented
• users' mistakes are flagged less well (e.g. they have to be manually checked for, and IDEs/editors cannot easily autocomplete or validate)
• our mistakes are flagged less well (e.g. misspelling the argument in the documentation, or renaming it)

If the argument was specified in the __init__ itself similar to bias and dropout, the documentation would list the default automatically, and tooling would be able to reason about it:

https://stellargraph.readthedocs.io/en/latest/api.html#stellargraph.layer.gcn.GCN

This is similar to Keras, which has Layer-specific arguments explicitly listed, and arguments for the Layer super-class passed in **kwargs: e.g. https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense

A list of probable suspects

(A search for kwargs[, kwargs.get, kwargs.pop)

$ git grep -n 'kwargs\(\[\|\.get\|\.pop\)'
stellargraph/layer/appnp.py:60:            kwargs["input_shape"] = (kwargs.get("input_dim"),)
stellargraph/layer/attri2vec.py:124:            self.input_node_num = kwargs["node_num"]
stellargraph/layer/attri2vec.py:125:            self.input_feature_size = kwargs["input_dim"]
stellargraph/layer/attri2vec.py:126:            self.multiplicity = kwargs["multiplicity"]
stellargraph/layer/cluster_gcn.py:84:            kwargs["input_shape"] = (kwargs.get("input_dim"),)
stellargraph/layer/cluster_gcn.py:318:            param_value = kwargs.pop(param_name, None)
stellargraph/layer/gcn.py:78:            kwargs["input_shape"] = (kwargs.get("input_dim"),)
stellargraph/layer/gcn.py:89:            kwargs.pop("kernel_initializer", "glorot_uniform")
stellargraph/layer/gcn.py:92:            kwargs.pop("kernel_regularizer", None)
stellargraph/layer/gcn.py:94:        self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
stellargraph/layer/gcn.py:96:            kwargs.pop("bias_initializer", "zeros")
stellargraph/layer/gcn.py:98:        self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
stellargraph/layer/gcn.py:99:        self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))
stellargraph/layer/gcn.py:347:            param_value = kwargs.pop(param_name, None)
stellargraph/layer/graph_attention.py:140:            kwargs.pop("kernel_initializer", "glorot_uniform")
stellargraph/layer/graph_attention.py:143:            kwargs.pop("kernel_regularizer", None)
stellargraph/layer/graph_attention.py:145:        self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
stellargraph/layer/graph_attention.py:148:            kwargs.pop("bias_initializer", "zeros")
stellargraph/layer/graph_attention.py:150:        self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
stellargraph/layer/graph_attention.py:151:        self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))
stellargraph/layer/graph_attention.py:154:            kwargs.pop("attn_kernel_initializer", "glorot_uniform")
stellargraph/layer/graph_attention.py:157:            kwargs.pop("attn_kernel_regularizer", None)
stellargraph/layer/graph_attention.py:160:            kwargs.pop("attn_kernel_constraint", None)
stellargraph/layer/graph_attention.py:765:            self.multiplicity = kwargs.get("multiplicity", 1)
stellargraph/layer/graph_attention.py:766:            self.n_nodes = kwargs.get("num_nodes", None)
stellargraph/layer/graph_attention.py:767:            self.n_features = kwargs.get("num_features", None)
stellargraph/layer/graph_attention.py:845:            param_value = kwargs.pop(param_name, None)
stellargraph/layer/graphsage.py:94:            kwargs.pop("kernel_initializer", "glorot_uniform")
stellargraph/layer/graphsage.py:97:            kwargs.pop("kernel_regularizer", None)
stellargraph/layer/graphsage.py:99:        self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
stellargraph/layer/graphsage.py:101:            kwargs.pop("bias_initializer", "zeros")
stellargraph/layer/graphsage.py:103:        self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
stellargraph/layer/graphsage.py:104:        self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))
stellargraph/layer/graphsage.py:882:            self.n_samples = kwargs["n_samples"]
stellargraph/layer/graphsage.py:883:            self.input_feature_size = kwargs["input_dim"]
stellargraph/layer/graphsage.py:884:            self.multiplicity = kwargs["multiplicity"]
stellargraph/layer/graphsage.py:909:            param_value = kwargs.pop(param_name, None)
stellargraph/layer/graphsage.py:1160:            self.in_samples = kwargs["in_samples"]
stellargraph/layer/graphsage.py:1161:            self.out_samples = kwargs["out_samples"]
stellargraph/layer/graphsage.py:1162:            self.input_feature_size = kwargs["input_dim"]
stellargraph/layer/graphsage.py:1163:            self.multiplicity = kwargs["multiplicity"]
stellargraph/layer/hinsage.py:84:            kwargs.pop("kernel_initializer", "glorot_uniform")
stellargraph/layer/hinsage.py:87:            kwargs.pop("kernel_regularizer", None)
stellargraph/layer/hinsage.py:89:        self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
stellargraph/layer/hinsage.py:91:            kwargs.pop("bias_initializer", "zeros")
stellargraph/layer/hinsage.py:93:        self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
stellargraph/layer/hinsage.py:94:        self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))
stellargraph/layer/hinsage.py:416:            self.n_samples = kwargs["n_samples"]
stellargraph/layer/hinsage.py:417:            self.input_dims = kwargs["input_dim"]
stellargraph/layer/hinsage.py:418:            self.multiplicity = kwargs["multiplicity"]
stellargraph/layer/hinsage.py:419:            self.subtree_schema = kwargs["input_neighbor_tree"]
stellargraph/layer/hinsage.py:461:            param_value = kwargs.pop(param_name, None)
stellargraph/layer/ppnp.py:61:            kwargs["input_shape"] = (kwargs.get("input_dim"),)
stellargraph/layer/rgcn.py:94:            kwargs["input_shape"] = (kwargs.get("input_dim"),)
stellargraph/layer/rgcn.py:124:            kwargs.pop("kernel_initializer", "glorot_uniform")
stellargraph/layer/rgcn.py:127:            kwargs.pop("kernel_regularizer", None)
stellargraph/layer/rgcn.py:129:        self.kernel_constraint = constraints.get(kwargs.pop("kernel_constraint", None))
stellargraph/layer/rgcn.py:132:            kwargs.pop("basis_initializer", "glorot_uniform")
stellargraph/layer/rgcn.py:134:        self.basis_regularizer = regularizers.get(kwargs.pop("basis_regularizer", None))
stellargraph/layer/rgcn.py:135:        self.basis_constraint = constraints.get(kwargs.pop("basis_constraint", None))
stellargraph/layer/rgcn.py:138:            kwargs.pop("coefficient_initializer", "glorot_uniform")
stellargraph/layer/rgcn.py:141:            kwargs.pop("coefficient_regularizer", None)
stellargraph/layer/rgcn.py:144:            kwargs.pop("coefficient_constraint", None)
stellargraph/layer/rgcn.py:148:            kwargs.pop("bias_initializer", "zeros")
stellargraph/layer/rgcn.py:150:        self.bias_regularizer = regularizers.get(kwargs.pop("bias_regularizer", None))
stellargraph/layer/rgcn.py:151:        self.bias_constraint = constraints.get(kwargs.pop("bias_constraint", None))
stellargraph/layer/rgcn.py:479:            param_value = kwargs.pop(param_name, None)

Here's a checklist (if you choose to work on one, edit this to put your name next to the row, but only check it once the PR merges):

• GCN @huonw Replace GCN use of **kwargs with real arguments #802
• APPNP @huonw
• Attri2Vec @kjun9
• ClusterGCN @kjun9
• GAT @kjun9
• GraphSAGE @kjun9
• HinSAGE @kjun9
• PPNP @huonw
• RGCN @kjun9

To Reproduce

N/A

Observed behavior

N/A

Expected behavior

Use of **kwargs should be minimised.

Environment

Operating system: Darwin-18.6.0-x86_64-i386-64bit
Python version:

3.7.3 (default, Jul 10 2019, 09:35:06) 
[Clang 10.0.1 (clang-1001.0.46.4)]

Package versions:

absl-py==0.9.0
appdirs==1.4.3
appnope==0.1.0
astor==0.8.1
attrs==19.3.0
backcall==0.1.0
black==19.10b0
bleach==3.1.0
boto==2.49.0
boto3==1.11.10
botocore==1.14.10
cachetools==4.0.0
certifi==2019.11.28
chardet==3.0.4
Click==7.0
coverage==4.5.4
cycler==0.10.0
decorator==4.4.1
defusedxml==0.6.0
docopt==0.6.2
docutils==0.15.2
entrypoints==0.3
gast==0.2.2
gensim==3.8.1
google-auth==1.11.0
google-auth-oauthlib==0.4.1
google-pasta==0.1.8
grpcio==1.27.0
h5py==2.10.0
idna==2.8
importlib-metadata==1.5.0
ipykernel==5.1.4
ipython==7.12.0
ipython-genutils==0.2.0
ipywidgets==7.5.1
isodate==0.6.0
jedi==0.16.0
Jinja2==2.11.1
jmespath==0.9.4
joblib==0.14.1
jsonschema==3.2.0
jupyter==1.0.0
jupyter-client==5.3.4
jupyter-console==6.1.0
jupyter-core==4.6.1
Keras-Applications==1.0.8
Keras-Preprocessing==1.1.0
kiwisolver==1.1.0
llvmlite==0.31.0
Markdown==3.1.1
MarkupSafe==1.1.1
matplotlib==3.1.3
mistune==0.8.4
more-itertools==8.2.0
mplleaflet==0.0.5
nbconvert==5.6.1
nbformat==5.0.4
networkx==2.4
notebook==6.0.3
numba==0.48.0
numpy==1.18.1
oauthlib==3.1.0
opt-einsum==3.1.0
packaging==20.1
pandas==1.0.0
pandocfilters==1.4.2
parso==0.6.1
pathspec==0.7.0
pexpect==4.8.0
pickleshare==0.7.5
pluggy==0.13.1
prometheus-client==0.7.1
prompt-toolkit==3.0.3
protobuf==3.11.3
ptyprocess==0.6.0
py==1.8.1
py-cpuinfo==5.0.0
pyasn1==0.4.8
pyasn1-modules==0.2.8
Pygments==2.5.2
pyparsing==2.4.6
pyrsistent==0.15.7
pytest==5.3.1
pytest-benchmark==3.2.3
pytest-cov==2.8.1
python-dateutil==2.8.1
pytz==2019.3
pyzmq==18.1.1
qtconsole==4.6.0
rdflib==4.2.2
regex==2020.1.8
requests==2.22.0
requests-oauthlib==1.3.0
rsa==4.0
s3transfer==0.3.2
scikit-learn==0.22.1
scipy==1.4.1
seaborn==0.10.0
Send2Trash==1.5.0
six==1.14.0
smart-open==1.9.0
-e git+git@github.com:stellargraph/stellargraph.git@2c513c19c8b8ba931eda3dbf9bf25197821f1ddb#egg=stellargraph&subdirectory=../../../stellargraph-2
tensorboard==2.0.2
tensorflow==2.0.1
tensorflow-estimator==2.0.1
termcolor==1.1.0
terminado==0.8.3
testpath==0.4.4
toml==0.10.0
tornado==6.0.3
traitlets==4.3.3
treon==0.1.3
typed-ast==1.4.1
urllib3==1.25.8
wcwidth==0.1.8
webencodings==0.5.1
Werkzeug==0.16.1
widgetsnbextension==3.5.1
wrapt==1.11.2
zipp==2.1.0

Additional context

N/A