cellcomplex.graph.property_graph module

This module provide a set of concepts to add properties to graph elements

class cellcomplex.graph.property_graph.PropertyGraph(graph=None, **kwds)

Bases: cellcomplex.graph.interface.property_graph.IPropertyGraph, cellcomplex.graph.graph.Graph

Simple implementation of IPropertyGraph using dict as properties and two dictionaries to maintain these properties

add_edge_property(property_name, values=None)

add a new map between eid and a data do not fill this property for any edge

add_graph_property(property_name, values=None)

todo

add_region_from_func(func, region_name)

Create a region region_name of vertices according to a function func.

Parameters

• func : the function to make the region (might return True or False)

• region_name : the name of the region

add_regions_from_dict(dict_regions, region_names)

If one already posses a dict indicating for a list of vertex which region they belong to, it can be given to the graph directly.

Parameters

• dict_regions (dict) - *keys = ids (SpatialImage); *values = intergers indicating the region(s)

• region_name (list) - a list containing the name of the region(s)

add_vertex_property(property_name, values=None)

add a new map between vid and a data do not fill this property for any vertex

add_vertex_to_region(vids, region_name)

Add a set of vertices vids to a region region_name. Save it in two places:

• self.graph_property[region_name] will return the list of all vertices belonging to ‘region_name’;

• self.vertex_property[“regions”][vid] will return the list of all regions vid belong to.

adjacency_matrix(edge_type=None, edge_dist=1, no_edge_val=0, oriented=True, reflexive=True, reflexive_value=0)

Return the adjacency matrix of the graph. :Parameters: - edge_type : type of edges we want to consider - edge_dist : cost ot cost function to apply between two edges, default : 1 - no_edge_val : cost to put if there is no edge between two vertices, default : 0 - oriented : if True, the graph is considered oriented and we always add an edge j -> i if i -> j exists - reflexive : if True, the graph is considered reflexive and we will put the cost or the cost_function reflexive_value on the diagonal of the adjacency_matrix, default : 0

Return

• numpy.array : a NxN matrix

clear()

remove all vertices and edges don’t change references to objects

clear_edges()

remove all the edges of the graph don’t change references to objects

edge_properties()

todo

edge_property(property_name)

return a map between eid and data for all edges where property_name is defined

Return type

dict of {eid:data}

edge_property_names()

iter on names of all property maps defined on edge return iter of keys

edges(vid=None, edge_type=None)

Return edges of the vertex vid

If vid=None, return all edges of the graph

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider

:Returns

• edge_list : the set of the edges of the vertex vid

extend(graph)

add the specified graph to self, create new vid and eid

Parameters

graph (Graph) – the graph to add

Returns

two dictionnary specifying correspondence between graph id and self id

Return type

({vid:vid},{eid:eid})

extend_graph_property(property_name, values)

todo

extend_vertex_property(property_name, values)

todo AND TO CHECK AND TEST !!

floyd_warshall(edge_type=None, edge_dist=1, oriented=False)

from_networkx(graph)

Return a Graph from a NetworkX Directed graph. :Parameters:

• graph : A NetworkX graph.

Returns

get_graph_property_key_type(propertyname)

Return meta info on graph property key type.

get_property_value_type(propertyname, property_type=0)

Return meta info on property value type.

graph_properties()

graph_property(property_name)

return a dict containing the graph properties/

Return type

dict of {property_name:data}

graph_property_names()

todo

in_edges(vid, edge_type=None)

Return in edges of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• edge_list : the set of the in edges of the vertex vid

in_neighbors(vid, edge_type=None)

Return the in vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• neighbors_list : the set of parent vertices of the vertex vid

is_connected_region(region_name, edge_type=None)

Return True if a region is connected.

iter_edges(vid, edge_type=None)

Return in edges of the vertex vid

If vid=None, return all edges of the graph

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider

:Returns

• iterator : an iterator on the set of the edges of the vertex vid

iter_in_edges(vid, edge_type=None)

Return in edges of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• iterator : an iterator on the set of the in edges of the vertex vid

iter_in_neighbors(vid, edge_type=None)

Return the in vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• iterator : an iterator on the set of parent vertices of the vertex vid

iter_neighborhood(vid, n, edge_type=None)

Return the neighborhood of the vertex vid at distance n (the disc, not the circle)

Parameters

• vids : a set of vertex id

:Returns

• iterator : an iterator on the set of the vertices at distance n of the vertex vid

iter_neighbors(vid, edge_type=None)

Return the neighbors vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• iterartor : iterator on the set of neighobrs vertices of the vertex vid

iter_out_edges(vid, edge_type=None)

Return in edges of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider

:Returns

• iterator : an iterator on the set of the in edges of the vertex vid

iter_out_neighbors(vid, edge_type=None)

Return the out vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• iterator : an iterator on the set of child vertices of the vertex vid

iter_region(region_name)

metaeidtypepropertyname = 'valueproperty_as_eid'

metavidtypepropertyname = 'valueproperty_as_vid'

neighborhood(vid, max_distance=1, edge_type=None)

Return the neighborhood of the vertex vid at distance max_distance (the disc, not the circle)

Parameters

• vid : vertex id

:Returns

• neighbors_list : the set of the vertices at distance below max_distance of the vertex vid (including vid)

neighbors(vid, edge_type=None)

Return the neighbors vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• neighbors_list : the set of neighobrs vertices of the vertex vid

out_edges(vid, edge_type=None)

Return out edges of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• edge_list : the set of the out edges of the vertex vid

out_neighbors(vid, edge_type=None)

Return the out vertices of the vertex vid

Parameters

• vid : a vertex id

• edges_type : type of edges we want to consider (can be a set)

:Returns

• neighbors_list : the set of child vertices of the vertex vid

remove_edge(eid)

remove a specified edge from the graph

Parameters

eid (eid) – id of the edge to remove

remove_edge_property(property_name)

remove the map called property_name from the graph

remove_graph_property(property_name)

todo

remove_region(region_name)

Remove a region region_name.

remove_vertex(vid)

remove a specified vertex of the graph remove all the edges attached to it

Parameters

vid (vid) – the id of the vertex to remove

remove_vertex_from_region(vids, region_name)

Remove a set of vertices vids from a region region_name.

remove_vertex_property(property_name)

remove the map called property_name from the graph

set_graph_property_key_to_eid_type(propertyname)

Give meta info on graph property key type. Associate it to Edge Id type

set_graph_property_key_to_vid_type(propertyname)

Give meta info on graph property key type. Associate it to Vertex Id type

set_graph_property_value_to_eid_type(propertyname, property_type=1)

Give meta info on property value type. Associate it to Edge Id type

set_graph_property_value_to_vid_type(propertyname, property_type=0)

Give meta info on property value type. Associate it to Vertex Id type

to_networkx()

topological_distance(vid, edge_type=None, edge_dist=<function PropertyGraph.<lambda>>, max_depth=inf, full_dict=True, return_inf=True)

Return the distances of each vertices from the vertex vid according a cost function

Parameters

• vid (int) - a vertex id

• edges_type (str) - type of edges we want to consider e.g. ‘s’ or ‘t’

• edge_dist (function) - the cost function

• max_depth (float) - the maximum depth that we want to reach

• full_dict (bool) - if True this function will return the entire dictionary (with inf values)

• return_inf (bool) - if True (default) return ‘inf’ values, else ‘nan’.

:Returns

• dist_dict : a dictionary of the distances, key : vid, value : distance

translate_graph_property(prop_name, trans_vid, trans_eid)

Translate a graph property according to meta info

vertex_properties()

todo

vertex_property(property_name, vids=None)

return a map between vid and data for all vertices where property_name is defined

Return type

dict of {vid:data}

vertex_property_names()

iter on names of all property maps defined on vertices return iter of keys