def edge_summary(filepath, input_type, max_rows, output):
"""
Loads and summarizes a knowledge graph edge set
"""
t = build_transformer(filepath, input_type)
t.parse(filepath)
g = t.graph
tuples = []
with click.progressbar(g.edges(data=True), label='Reading knowledge graph') as bar:
for s, o, edge_attr in bar:
subject_attr = g.node[s]
object_attr = g.node[o]
subject_prefix = stringify(get_prefix(s))
object_prefix = stringify(get_prefix(o))
subject_category = stringify(subject_attr.get('category'))
object_category = stringify(object_attr.get('category'))
edge_label = stringify(edge_attr.get('edge_label'))
relation = stringify(edge_attr.get('relation'))
tuples.append((subject_prefix, subject_category, edge_label, relation, object_prefix, object_category))
tuple_count = OrderedDict(Counter(tuples).most_common(max_rows))
headers = [['Subject Prefix', 'Subject Category', 'Edge Label', 'Relation', 'Object Prefix', 'Object Category', 'Frequency']]
rows = [[*k, v] for k, v in tuple_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table)
else:
click.echo(AsciiTable(headers + rows).table)
def edge_summary(config: dict, filepath: str, input_type: str, max_rows: int,
output: str):
"""
Loads and summarizes a knowledge graph edge set, where the input is a file.
\f
Parameters
----------
config: dict
A dictionary containing the configuration for kgx.cli
filepath: str
Input file
input_type: str
Input file type
max_rows: int
Max number of rows to display in the output
output: str
Where to write the output (stdout, by default)
"""
t = build_transformer(filepath, input_type)
t.parse(filepath)
g = t.graph
tuples = []
with click.progressbar(g.edges(data=True),
label='Reading knowledge graph') as bar:
for s, o, edge_attr in bar:
subject_attr = g.node[s]
object_attr = g.node[o]
subject_prefix = stringify(get_prefix(s))
object_prefix = stringify(get_prefix(o))
subject_category = stringify(subject_attr.get('category'))
object_category = stringify(object_attr.get('category'))
edge_label = stringify(edge_attr.get('edge_label'))
relation = stringify(edge_attr.get('relation'))
tuples.append((subject_prefix, subject_category, edge_label,
relation, object_prefix, object_category))
tuple_count = OrderedDict(Counter(tuples).most_common(max_rows))
headers = [[
'Subject Prefix', 'Subject Category', 'Edge Label', 'Relation',
'Object Prefix', 'Object Category', 'Frequency'
]]
rows = [[*k, v] for k, v in tuple_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table)
else:
click.echo(AsciiTable(headers + rows).table)
def node_summary(config: dict, filepath: str, input_type: str, max_rows: int,
output: str):
"""
Loads and summarizes a knowledge graph node set, where the input is a file.
\f
Parameters
----------
config: dict
A dictionary containing the configuration for kgx.cli
filepath: str
Input file
input_type: str
Input file type
max_rows: int
Max number of rows to display in the output
output: str
Where to write the output (stdout, by default)
"""
t = build_transformer(filepath, input_type)
t.parse(filepath)
g = t.graph
tuples = []
xrefs = set()
with click.progressbar(g.nodes(data=True),
label='Reading knowledge graph') as bar:
for n, data in bar:
if 'same_as' in data:
for xref in data['same_as']:
xrefs.add(get_prefix(xref))
category = data.get('category')
prefix = get_prefix(n)
if category is not None and len(
category) > 1 and 'named_thing' in category:
category.remove('named_thing')
if isinstance(category, (list, set)):
category = ", ".join("'{}'".format(c) for c in category)
if prefix is not None:
prefix = "'{}'".format(prefix)
tuples.append((prefix, category))
click.echo('|nodes|: {}'.format(len(g.nodes())))
click.echo('|edges|: {}'.format(len(g.edges())))
xrefs = [x for x in xrefs if x is not None]
if len(xrefs) != 0:
line = 'xref prefixes: {}'.format(', '.join(xrefs))
if output is not None:
file_write(output, '|nodes|: {}'.format(len(g.nodes())))
file_write(output, '|edges|: {}'.format(len(g.edges())))
file_write(output, line)
else:
click.echo('|nodes|: {}'.format(len(g.nodes())))
click.echo('|edges|: {}'.format(len(g.edges())))
click.echo(line)
tuple_count = OrderedDict(Counter(tuples).most_common(max_rows))
headers = [['Prefix', 'Category', 'Frequency']]
rows = [[*k, v] for k, v in tuple_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)
category_count = defaultdict(lambda: 0)
prefix_count = defaultdict(lambda: 0)
for (prefix, category), frequency in tuple_count.items():
category_count[category] += frequency
prefix_count[prefix] += frequency
headers = [['Category', 'Frequency']]
rows = [[k, v] for k, v in category_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)
headers = [['Prefixes', 'Frequency']]
rows = [[k, v] for k, v in prefix_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)
def node_summary(filepath, input_type, max_rows, output):
"""
Loads and summarizes a knowledge graph node set
"""
t = build_transformer(filepath, input_type)
t.parse(filepath)
g = t.graph
tuples = []
xrefs = set()
with click.progressbar(g.nodes(data=True),
label='Reading knowledge graph') as bar:
for n, data in bar:
if 'same_as' in data:
for xref in data['same_as']:
xrefs.add(get_prefix(xref))
category = data.get('category')
prefix = get_prefix(n)
if category is not None and len(
category) > 1 and 'named thing' in category:
category.remove('named thing')
if isinstance(category, (list, set)):
category = ", ".join("'{}'".format(c) for c in category)
if prefix is not None:
prefix = "'{}'".format(prefix)
tuples.append((prefix, category))
xrefs = [x for x in xrefs if x is not None]
if len(xrefs) != 0:
line = 'xref prefixes: {}'.format(', '.join(xrefs))
if output is not None:
file_write(output, line)
else:
click.echo(line)
tuple_count = OrderedDict(Counter(tuples).most_common(max_rows))
headers = [['Prefix', 'Category', 'Frequency']]
rows = [[*k, v] for k, v in tuple_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)
category_count = defaultdict(lambda: 0)
prefix_count = defaultdict(lambda: 0)
for (prefix, category), frequency in tuple_count.items():
category_count[category] += frequency
prefix_count[prefix] += frequency
headers = [['Category', 'Frequency']]
rows = [[k, v] for k, v in category_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)
headers = [['Prefixes', 'Frequency']]
rows = [[k, v] for k, v in prefix_count.items()]
if output is not None:
file_write(output, AsciiTable(headers + rows).table, mode='a')
else:
click.echo(AsciiTable(headers + rows).table)