cugraph.experimental.PropertyGraph

class cugraph.experimental.PropertyGraph(*args, **kwargs)

Class which stores vertex and edge properties that can be used to construct Graphs from individual property selections and used later to annotate graph algorithm results with corresponding properties.

__init__(*args, **kwargs)

Methods

__init__(*args, **kwargs)
add_edge_data(dataframe, vertex_col_names[, ...])	Add a dataframe describing edge properties to the PropertyGraph.
add_vertex_data(dataframe, vertex_col_name)	Add a dataframe describing vertex properties to the PropertyGraph.
annotate_dataframe(df, G, edge_vertex_col_names)	Add properties to df that represent the vertices and edges in graph G.
edge_props_to_graph(edge_prop_df, create_using)	Create a Graph from the edges in edge_prop_df.
edge_types_from_numerals(nums)	Returns the string edge type names given the numeric category labels.
edge_vector_property_to_array(df, col_name)	Convert a known edge vector property in a DataFrame to an array.
extract_subgraph([create_using, selection, ...])	Return a subgraph of the overall PropertyGraph containing vertices and edges that match a selection.
fillna_edges([val])	Fills empty edge property values with the given value, zero by default.
fillna_vertices([val])	Fills empty vertex property values with the given value, zero by default.
get_edge_data([edge_ids, types, columns])	Return a dataframe containing edge properties for only the specified edge_ids, columns, and/or edge type, or all edge IDs if not specified.
get_num_edges([type])	Return the number of all edges or edges of a given type.
get_num_vertices([type, include_edge_data])	Return the number of all vertices or vertices of a given type.
get_vertex_data([vertex_ids, types, columns])	Gets a DataFrame containing vertex properties
get_vertices([selection])	Return a Series containing the unique vertex IDs contained in both the vertex and edge property data in ascending order.
has_duplicate_edges(df[, columns])	Return True if df has rows with the same src, dst, type, and columns
is_multi_gpu()	Return True if this is a multi-gpu graph.
is_multigraph(df)	Parameters:
renumber_edges_by_type([prev_id_column])	Renumber edge IDs to be contiguous by type.
renumber_vertices_by_type([prev_id_column])	Renumber vertex IDs to be contiguous by type.
select_edges(expr)	Evaluate expr and return a PropertySelection object representing the edges that match the expression selection criteria.
select_vertices(expr[, from_previous_selection])	Evaluate expr and return a PropertySelection object representing the vertices that match the expression.
vertex_types_from_numerals(nums)	Returns the string vertex type names given the numeric category labels.
vertex_vector_property_to_array(df, col_name)	Convert a known vertex vector property in a DataFrame to an array.
vertices_ids()	Alias for get_vertices()

Attributes

dst_col_name	Column containing the id of the edge destination
edge_id_col_name	Column containing the edge identifier
edge_property_names	List containing each edge property name in the PropertyGraph instance.
edge_types	Series containing the set of edge type names
edges	All the edges in the graph as a DataFrame containing sources and destinations.
src_col_name	Column containing the id of the edge source
type_col_name	Column containing the type of the edge or vertex
vertex_col_name	Column containing the vertex id.
vertex_property_names	Names of all the vertex properties excluding type.
vertex_types	The set of vertex type names
weight_col_name	Column containing the edge weight if the graph is weighted.