Process undirected graphs: nctx.undirected

Process undirected graphs: nctx.undirected

Process undirected graphs: nctx.undirected

This submodule provides functionality for undirected graphs. The examples on this page all build on the same graph:

>>> import nctx.undirected as nctx
>>> g = nctx.Graph()
>>> name_map = nctx.PropertyMapStr(g, 'name')
>>> context_map = PropertyMapVecDouble(g, 'context')
>>> nctx.read_graphml(g, 'bcde.ctx.undir.graphml', [name_map, context_map])

If you load this module, all operations refer to undirected graphs.

Up to now, there is no possibility to convert a undirected graph into a directed one or vice versa. This will be added in future versions.

Todo

Functionality to convert undirected to directed graphs.

The Graph Class

Summary
Graph This class represents an undirected graph.
__hash__  
__init__  
add_edge Add an edge that connects u and v.
add_vertex Add a vertex to the graph.
children Iterate over the children of a given vertex u.
edges Iterate over the edges of the graph.
in_edges Iterate over the incoming edges of a vertex.
numEdges Retrieve the number of edges in the graph.
numVertices Retrieve the number of vertices in the graph.
out_edges Iterate over the outgoing edges of a vertex.
parents Iterate over the parents of a given vertex u.
source Retrieve the source vertex of an edge.
target Retrieve the target vertex of an edge.
vertices Iterate over the vertices of the graph.
class nctx.undirected.Graph

This class represents an undirected graph.

Example:
>>> import nctx.undirected as nctx
>>> graph = nctx.Graph()
__hash__() → int

Not commented yet

__init__((object)arg1) → None

Not commented yet

add_edge((int)u, (int)v) → bool :

Add an edge that connects u and v.

Returns:

Status if adding the edge was successful.
Return type:

bool
Example:
>>> from collections import defaultdict
>>> vertices = defaultdict(lambda: graph.add_vertex())
>>> for i in range(10):
...   _ = vertices[i]
>>> graph.add_edge(vertices[0], vertices[1])
True
add_vertex() → int :

Add a vertex to the graph.

Returns:

The ID of the added vertex.
Return type:

int
Example:
>>> from collections import defaultdict
>>> name_map = nctx.PropertyMapStr(graph, 'name')
>>> vertices = defaultdict(lambda: graph.add_vertex())
>>> for i in range(10):
...   v = vertices[i]
...   name_map.set_elem(v, f"vertex {i}")
children((int)u) → a_iter_ids :

Iterate over the children of a given vertex u. Since there is no difference between children and parents in an undirected graph, children yields the same list of vertices as parents

Parameters:

u (int) – ID of the vertex whose children should be iterated.
Returns:

A vertex iterator
Return type:

v_iter_ids
Example:
>>> some_vertex = 0
>>> for c in g.children(some_vertex):
...   pass
edges() → e_iter_ids :

Iterate over the edges of the graph. Use source to access the source vertex of an edge and target to access its target. Note that incoming and outgoing edges are not distinguished in an undirected graph, therefore in_edges, out_edges, and this function all yield the same results.

Returns:

The edge iterator
Return type:

e_iter_ids
Example:
>>> for e in g.edges():
...   source_vertex = g.source(e)
...   target_vertex = g.target(e)
in_edges((int)arg2) → e_iter_ids_in :

Iterate over the incoming edges of a vertex. Use source to access the source vertex of an edge.

Parameters:

v (int) – ID of the vertex whose in-edges should be iterated.
Returns:

The in-edge iterator
Return type:

e_iter_ids_in
Example:
>>> v=4
>>> for e in g.in_edges(v):
...   source_vertex = g.source(e)
numEdges() → int :

Retrieve the number of edges in the graph.

Returns:The number of edges
Return type:int
numVertices() → int :

Retrieve the number of vertices in the graph.

Returns:The number of vertices
Return type:int
out_edges((int)arg2) → e_iter_ids_out :

Iterate over the outgoing edges of a vertex. Use target to access the target vertex of an edge.

Parameters:

v (int) – ID of the vertex whose out-edges should be iterated.
Returns:

The out-edge iterator
Return type:

e_iter_ids_out
Example:
>>> v=4
>>> for e in g.in_edges(v):
...   target_vertex = g.target(e)
parents((int)u) → a_iter_ids_rev :

Iterate over the parents of a given vertex u. Since there is no difference between children and parents in an undirected graph, parents yields the same list of vertices as children

Parameters:

u (int) – ID of the vertex whose parents should be iterated.
Returns:

A vertex iterator
Return type:

v_iter_ids
Example:
>>> some_vertex = 0
>>> for p in g.parents(some_vertex):
...   pass
source((Edge)e) → int :

Retrieve the source vertex of an edge. See edges for an example.

Parameters:e (Edge) – An edge.
Returns:The ID of the source vertex
Return type:int
target((Edge)e) → int :

Retrieve the target vertex of an edge. See edges for an example.

Parameters:e (Edge) – An edge.
Returns:The ID of the target vertex
Return type:int
vertices() → v_iter_ids :

Iterate over the vertices of the graph.

Returns:

A vertex iterator
Return type:

v_iter_ids
Example:
>>> for v in g.vertices():
...   pass

AlgPaths: Shortest-path discovery

Summary
AlgPaths  
dijkstra Conventional single-source shortest paths discovery using Dijkstra’s algorithm.
dijkstra_apsp_ctx Obtain all-pairs-shortest-paths using Dijkstra’s algorithm taking into account contextual constraints.
dijkstra_ctx Obtain shortest paths with dynamic contextual constraints.
find_path Conventional shortest path discovery using Dijkstra’s algorithm.
find_path_ctx Obtain a shortest paths from a start vertex to a target vertex using Dijkstra’s algorithm taking into account contextual constraints.
class nctx.undirected.AlgPaths
dijkstra((Graph)g, (int)start, (PropertyMapULong)distances) → None :
dijkstra((Graph)g, (int)start) → PropertyMapULong :

Conventional single-source shortest paths discovery using Dijkstra’s algorithm.

Parameters:
  • g (Graph) – The Graph object
  • start (int) – Source vertex
  • distances (PropertyMapULong) – PropertyMap in which distances will be written.
Returns:

PropertyMap with distances.
Return type:

PropertyMapULong
dijkstra_apsp_ctx((Graph)g, (object)decision_fct, (PropertyMapVecULong)distances) → None :
dijkstra_apsp_ctx((Graph)g, (object)decision_fct) → PropertyMapVecULong :

Obtain all-pairs-shortest-paths using Dijkstra’s algorithm taking into account contextual constraints.

This is basically a wrapper around dijkstra_ctx. Please see the mentioned function for more information regarding the contextual constraints.

Parameters:
  • g (Graph) – The Graph object
  • decision_fct (function) – The decision function enforcing contextual constraints. The signature of the function is (vertex index, vertex index, vertex index) -> Bool
  • distances (PropertyMapVecULong) – matrix in which distances will be written
Returns:

matrix of distances named apsp_constraint
Return type:

PropertyMapVecULong
Example:
>>> distmap = PropertyMapVecULong(g, 'distances_ctx')
>>> AlgPaths.dijkstra_apsp_ctx(g,lambda _start,_current,_next: (True), distmap)
Example:
>>> distmap = AlgPaths.dijkstra_apsp_ctx(g,lambda _start,_current,_next: (True))
dijkstra_ctx((Graph)g, (int)s, (object)decision_fct, (PropertyMapULong)distances) → None :
dijkstra_ctx((Graph)g, (int)s, (object)decision_fct) → PropertyMapULong :

Obtain shortest paths with dynamic contextual constraints.

Single-source shortest paths using Dijkstra’s algorithm taking into account contextual constraints. Enforcement of constraints is the task of the given user-defined function.

The function enforcing contextual constraints is evaluated at each node during shortest path discovery. The function needs to evaluate to true or false allowing an edge to be visited or not. As parameters, the function needs to accept the starting node of path traversal, the current node, and the descending node in question. If the function returns False, the descending node is not being visited. Passing the start vertex is unnecessary here. However, the signature is the same as for the other functions for usability reasons.

The three nodes are passed as indices allowing for access of (external) attribute and other associated information.

Note that the decision function is evaluated more than once during path traversal. That means, there should not happen any resource-intense computation inside this function. Also, it does not allow to keep track of the status of calculation, e.g. by calculating the visited edges or something similar.

If the decision function simply returns True all the time, the set of shortest paths is the unaltered set of shortest paths expected by the classical Dijkstra-implementation.

Parameters:

distances (PropertyMapULong) – map in which distances will be written
Returns:

map containing distances named dijkstra_constraint
Return type:

PropertyMapULong
Example:
>>> start = 3
>>> distmap = PropertyMapULong(g, 'distances_ctx')
>>> AlgPaths.dijkstra_ctx(g, start, lambda _start,_current,_next: (True), distmap)
Example:
>>> start = 3
>>> distmap = AlgPaths.dijkstra_ctx(g, start, lambda _start,_current,_next: (True))
find_path((Graph)g, (int)start, (int)target) → list :

Conventional shortest path discovery using Dijkstra’s algorithm.

Parameters:
  • g (Graph) – The Graph object
  • start (int) – Source vertex
  • target (int) – Target vertex
Returns:

A list of vertex indices describing the shortest path from start to target.
Return type:

list
find_path_ctx((Graph)g, (int)start, (int)target, (object)decision_fct) → list :

Obtain a shortest paths from a start vertex to a target vertex using Dijkstra’s algorithm taking into account contextual constraints.

This is basically a wrapper around dijkstra_ctx. Please see the mentioned function for more information regarding the contextual constraints.

Parameters:
  • g (Graph) – The Graph object
  • decision_fct (function) – The decision function enforcing contextual constraints. The signature of the function is (vertex index, vertex index, vertex index) -> Bool
  • start (int) – Start vertex
  • target (int) – Target vertex
Returns:

A list of vertex indices describing the shortest path from start to target.
Return type:

list
Example:
>>> start = 3
>>> target = 10
>>> path = AlgPaths.find_path_ctx(g,start,target,lambda _start,_current,_next: (True))

AlgCentralities: Centrality measures

Summary
AlgCentralities  
betweenness Obtain betweenness centrality for a graph using Brandes’ efficient algorithm.
betweenness_ctx Betweenness centrality with dynamic contextual constraints.
closeness Obtain closeness centrality for a graph.
closeness_ctx Closeness centrality with dynamic contextual constraints.
class nctx.undirected.AlgCentralities
betweenness((Graph)g, (PropertyMapDouble)outmap) → None :
betweenness((Graph)g) → PropertyMapDouble :

Obtain betweenness centrality for a graph using Brandes’ efficient algorithm.

Parameters:
Returns:

list of centrality values named betweenness
Return type:

PropertyMapDouble
Example:
>>> outmap = PropertyMapDouble(g, 'betw')
>>> AlgCentralities.betweenness(g, outmap)
Example:
>>> outmap = AlgCentralities.betweenness(g)
betweenness_ctx((Graph)g, (object)betw_decision_fct, (PropertyMapDouble)outmap) → None :
betweenness_ctx((Graph)g, (object)betw_decision_fct) → PropertyMapDouble :

Betweenness centrality with dynamic contextual constraints.

Using this function allows obtaining betweenness centrality under dynamic contextual constraints. Enforcement of constraints is the task of the given user-defined function.

The function enforcing contextual constraints is evaluated at each node during shortest path traversal. The function needs to evaluate to True or False allowing an edge to be visited or not. As parameters, the current state of the betweenness calculation is passed to the function, i.e. the starting node for which a centrality value is being calculated, the current node, and the descending node in question. If the function returns False, the descending node is not being visited.

The three nodes are passed as indices allowing for access of (external) attribute and other associated information.

Note that the decision function is evaluated more than once during path traversal. That means, there should not happen any resource-intense computation inside this function. Also, it does not allow to keep track of the status of calculation, e.g. by calculating the visited edges or something similar.

If the decision function simply returns True all the time, this function results in the unaltered betweenness centrality values.

Obtaining betweenness centrality is based on Brandes’ efficient algorithm. At the current stage, the implementation allows for single-core execution only.

Parameters:

outmap (PropertyMapDouble) – map in which centrality values will be written
Returns:

list of centrality values named betweenness_constraint
Return type:

PropertyMapDouble
Example:
>>> outmap = PropertyMapDouble(g, 'betw_ctx')
>>> AlgCentralities.betweenness_ctx(g,lambda _start,_current,_next: (True),outmap)
Example:
>>> outmap = AlgCentralities.betweenness_ctx(g,lambda _start,_current,_next: (True))
closeness((Graph)g) → PropertyMapDouble :
closeness((Graph)g, (PropertyMapDouble)outmap) → None :

Obtain closeness centrality for a graph.

Parameters:
Returns:

list of centrality values named betweenness
Return type:

PropertyMapDouble
Example:
>>> outmap = AlgCentralities.closeness(g)
Example:
>>> outmap = PropertyMapDouble(g, 'clsn')
>>> AlgCentralities.closeness(g, outmap)
closeness_ctx((Graph)g, (object)clsn_decision_fct, (PropertyMapDouble)outmap) → None :
closeness_ctx((Graph)g, (object)clsn_decision_fct) → PropertyMapDouble :

Closeness centrality with dynamic contextual constraints.

Using this function allows obtaining closeness centrality under dynamic contextual constraints. Enforcement of constraints is the task of the given user-defined function.

The function enforcing contextual constraints is evaluated at each node during shortest path traversal. The function needs to evaluate to True or False allowing an edge to be visited or not. As parameters, the current state of the centrality calculation is passed to the function, i.e. the starting node for which a centrality value is being calculated, the current node, and the descending node in question. If the function returns False, the descending node is not being visited.

The three nodes are passed as indices allowing for access of (external) attribute and other associated information.

Note that the decision function is evaluated more than once during path traversal. That means, there should not happen any resource-intense computation inside this function. Also, it does not allow to keep track of the status of calculation, e.g. by calculating the visited edges or something similar.

If the decision function simply returns True all the time, this function results in the unaltered betweenness centrality values.

Parameters:

outmap (PropertyMapDouble) – map in which centrality values will be written
Returns:

list of centrality values named closeness_constraint
Return type:

PropertyMapDouble
Example:
>>> outmap = PropertyMapDouble(g, 'clsn_ctx')
>>> AlgCentralities.closeness_ctx(g,lambda _start,_current,_next: (True),outmap)
Example:
>>> outmap = AlgCentralities.closeness_ctx(g,lambda _start,_current,_next: (True))

Graph Property Maps

Property maps allow to store attribute information together with the vertices. These are mainly used for importing and exporting graphs: read_graphml and write_graphml

Todo

Support edge property maps

Property Maps containing atom elements

Summary
PropertyMapDouble PropertyMapDouble serves as a list of atom items of type float.
PropertyMapInt PropertyMapInt serves as a list of atom items of type int.
PropertyMapStr PropertyMapStr serves as a list of atom items of type str.
PropertyMapULong PropertyMapULong serves as a list of atom items of type unsigned long - a number > 0.
class nctx.undirected.PropertyMapDouble

PropertyMapDouble serves as a list of atom items of type float. The list contains values of a single data type only as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list, i.e. it allows iteration and supports built-in functions append and extend.

The contained data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (float) – initial value
Example:
>>> my_map = PropertyMapDouble(g, 'context1')
>>> my_map = PropertyMapDouble(g, 'context1', .0)
Example:
>>> import numpy as np
>>> np_arr = np.array(list(my_map))
__contains__((float)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> .0 in my_map
True
>>> .1 in my_map
False
__getitem__((int)arg2) → float :

This allows to access items using square brackets. See get_elem for a detailed description of parameters.

Example:
>>> _ = my_map[2]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (float)arg4) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → float :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (float)arg3) → None :

This allows to modify items using square brackets. See set_elem for a detailed description of parameters.

Example:
>>> my_map[2] = .1
get_elem((int)i) → float :

Retrieve the ith element of this list.n

Parameters:i (int) – Index of element.
Returns:ith element.
Return type:float
get_name() → str :

Retrieve name of the property map.

The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → float :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (float)newval) → None :

Set the ith element of this list to be newval. Note that the type of newval needs to be float or a type that can be implicitly converted to float.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (float) – The new element.
Example:

Note the type conversion (float to int) from the input list to the output of get_elem.

>>> complex_map = PropertyMapDouble(g, 'context')
>>> complex_map.set_elem(0, 3)
>>> complex_map.get_elem(0)
3.0
class nctx.undirected.PropertyMapInt

PropertyMapInt serves as a list of atom items of type int. The list contains values of a single data type only as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list, i.e. it allows iteration and supports built-in functions append and extend.

The contained data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (int) – initial value
Example:
>>> my_map = PropertyMapInt(g, 'context1')
>>> my_map = PropertyMapInt(g, 'context1', 0)
Example:
>>> import numpy as np
>>> np_arr = np.array(list(my_map))
__contains__((int)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> 0 in my_map
True
>>> 1 in my_map
False
__getitem__((int)arg2) → int :

This allows to access items using square brackets. See get_elem for a detailed description of parameters.

Example:
>>> _ = my_map[2]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (int)arg4) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → int :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (int)arg3) → None :

This allows to modify items using square brackets. See set_elem for a detailed description of parameters.

Example:
>>> my_map[2] = 1
get_elem((int)i) → int :

Retrieve the ith element of this list.n

Parameters:i (int) – Index of element.
Returns:ith element.
Return type:int
get_name() → str :

Retrieve name of the property map.

The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → int :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)newval) → None :

Set the ith element of this list to be newval. Note that the type of newval needs to be int or a type that can be implicitly converted to int. Please see PropertyMapDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (int) – The new element.
class nctx.undirected.PropertyMapStr

PropertyMapStr serves as a list of atom items of type str. The list contains values of a single data type only as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list, i.e. it allows iteration and supports built-in functions append and extend.

The contained data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (str) – initial value
Example:
>>> my_map = PropertyMapStr(g, 'context1')
>>> my_map = PropertyMapStr(g, 'context1', '')
Example:
>>> import numpy as np
>>> np_arr = np.array(list(my_map))
__contains__((str)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> '' in my_map
True
>>> '1' in my_map
False
__getitem__((int)arg2) → str :

This allows to access items using square brackets. See get_elem for a detailed description of parameters.

Example:
>>> _ = my_map[2]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (str)arg4) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → str :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (str)arg3) → None :

This allows to modify items using square brackets. See set_elem for a detailed description of parameters.

Example:
>>> my_map[2] = '1'
get_elem((int)i) → str :

Retrieve the ith element of this list.n

Parameters:i (int) – Index of element.
Returns:ith element.
Return type:str
get_name() → str :

Retrieve name of the property map.

The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → str :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (str)newval) → None :

Set the ith element of this list to be newval. Note that the type of newval needs to be str or a type that can be implicitly converted to str. Please see PropertyMapDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (str) – The new element.
class nctx.undirected.PropertyMapULong

PropertyMapULong serves as a list of atom items of type unsigned long - a number > 0. The list contains values of a single data type only as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list, i.e. it allows iteration and supports built-in functions append and extend.

The contained data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (int) – initial value
Example:
>>> my_map = PropertyMapULong(g, 'context1')
>>> my_map = PropertyMapULong(g, 'context1', 0)
Example:
>>> import numpy as np
>>> np_arr = np.array(list(my_map))
__contains__((int)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> 0 in my_map
True
>>> 1 in my_map
False
__getitem__((int)arg2) → int :

This allows to access items using square brackets. See get_elem for a detailed description of parameters.

Example:
>>> _ = my_map[2]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (int)arg4) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → int :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (int)arg3) → None :

This allows to modify items using square brackets. See set_elem for a detailed description of parameters.

Example:
>>> my_map[2] = 1
get_elem((int)i) → int :

Retrieve the ith element of this list.n

Parameters:i (int) – Index of element.
Returns:ith element.
Return type:int
get_name() → str :

Retrieve name of the property map.

The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → int :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)newval) → None :

Set the ith element of this list to be newval. Note that the type of newval needs to be int or a type that can be implicitly converted to int. Please see PropertyMapDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (int) – The new element.

Property Maps containing list elements

Summary
PropertyMapVecDouble PropertyMapVecDouble serves as a container of lists, i.
PropertyMapVecInt PropertyMapVecInt serves as a container of lists, i.
PropertyMapVecStr PropertyMapVecStr serves as a container of lists, i.
PropertyMapVecULong PropertyMapVecULong serves as a container of lists, i.
class nctx.undirected.PropertyMapVecDouble

PropertyMapVecDouble serves as a container of lists, i.e. it is a list of lists of float. The nested lists all contain values of a single data type as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list returning lists as items when iterating over it.

The allowed data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (float) – initial value for each inner list
  • init_size (int) – size of each inner list.
Example:
>>> my_map = PropertyMapVecDouble(g, 'context1')
>>> my_map = PropertyMapVecDouble(g, 'context1', 0.0, 10)
Example:
>>> import numpy as np
>>> np_arr = np.array([np.array(xi) for xi in my_map])
__contains__((object)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> [.1,.2,.3] in my_map
False
__getitem__((int)arg2) → object :

This allows to access items using square brackets. See get_list for a detailed description of parameters.

Example:
>>> _ = my_map[2]
>>> _ = my_map[2][0]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (float)arg4, (int)arg5) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (object)arg3) → None :

This allows to modify items using square brackets. See set_list for a detailed description of parameters.

Example:
>>> my_map[2] = [.1,.2,.3,.4,.5]
>>> my_map[2][0] = 42.0
get_elem((int)i, (int)j) → float :

Retrieve the jth element of ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

jth element in ith list.
Return type:

float
get_list((int)i) → object :

Retrieve the ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

The list saved at position i.
Return type:

PropertyMapVecDouble
get_name() → str :

Retrieve name of the property map. The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)j, (float)newval) → None :

Set the jth element of ith list. Note that the type of the new element needs to be float or a type that can be implicitly converted to float.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (float) – The new element.
Example:

Note the type conversion (float to int) from the input to the output of get_elem.

>>> complex_map = PropertyMapVecDouble(g, 'context')
>>> complex_map.set_elem(0, 1, 3)
>>> complex_map.get_elem(0,1)
3.0
set_list((int)i, (object)newlist) → None :

Set the ith list to be newlist.

Parameters:
  • i (int) – Index of List
  • newlist (list) – List to be saved at position i. Note that all items of this list will be converted to float internally.
Example:

Note the type conversion (float to int) from the input list to the output of get_elem.

>>> complex_map = PropertyMapVecDouble(g, 'context')
>>> complex_map.set_list(0, [1,2,3,4,5])
>>> complex_map.get_elem(0,2)
3.0
class nctx.undirected.PropertyMapVecInt

PropertyMapVecInt serves as a container of lists, i.e. it is a list of lists of int. The nested lists all contain values of a single data type as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list returning lists as items when iterating over it.

The allowed data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (int) – initial value for each inner list
  • init_size (int) – size of each inner list.
Example:
>>> my_map = PropertyMapVecInt(g, 'context1')
>>> my_map = PropertyMapVecInt(g, 'context1', 0, 10)
Example:
>>> import numpy as np
>>> np_arr = np.array([np.array(xi) for xi in my_map])
__contains__((object)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> [1,2,3] in my_map
False
__getitem__((int)arg2) → object :

This allows to access items using square brackets. See get_list for a detailed description of parameters.

Example:
>>> _ = my_map[2]
>>> _ = my_map[2][0]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (int)arg4, (int)arg5) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (object)arg3) → None :

This allows to modify items using square brackets. See set_list for a detailed description of parameters.

Example:
>>> my_map[2] = [1,2,3,4,5]
>>> my_map[2][0] = 42
get_elem((int)i, (int)j) → int :

Retrieve the jth element of ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

jth element in ith list.
Return type:

int
get_list((int)i) → object :

Retrieve the ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

The list saved at position i.
Return type:

PropertyMapVecInt
get_name() → str :

Retrieve name of the property map. The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)j, (int)newval) → None :

Set the jth element of ith list. Note that the type of the new element needs to be int or a type that can be implicitly converted to int. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (int) – The new element.
set_list((int)i, (object)newlist) → None :

Set the ith list to be newlist. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • newlist (list) – List to be saved at position i. Note that all items of this list will be converted to int internally.
class nctx.undirected.PropertyMapVecStr

PropertyMapVecStr serves as a container of lists, i.e. it is a list of lists of str. The nested lists all contain values of a single data type as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list returning lists as items when iterating over it.

The allowed data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (str) – initial value for each inner list
  • init_size (int) – size of each inner list.
Example:
>>> my_map = PropertyMapVecStr(g, 'context1')
>>> my_map = PropertyMapVecStr(g, 'context1', '', 10)
Example:
>>> import numpy as np
>>> np_arr = np.array([np.array(xi) for xi in my_map])
__contains__((object)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> ['1','2','3'] in my_map
False
__getitem__((int)arg2) → object :

This allows to access items using square brackets. See get_list for a detailed description of parameters.

Example:
>>> _ = my_map[2]
>>> _ = my_map[2][0]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (str)arg4, (int)arg5) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (object)arg3) → None :

This allows to modify items using square brackets. See set_list for a detailed description of parameters.

Example:
>>> my_map[2] = ['1','2','3','4','5']
>>> my_map[2][0] = '42'
get_elem((int)i, (int)j) → str :

Retrieve the jth element of ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

jth element in ith list.
Return type:

str
get_list((int)i) → object :

Retrieve the ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

The list saved at position i.
Return type:

PropertyMapVecStr
get_name() → str :

Retrieve name of the property map. The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)j, (str)newval) → None :

Set the jth element of ith list. Note that the type of the new element needs to be str or a type that can be implicitly converted to str. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (str) – The new element.
set_list((int)i, (object)newlist) → None :

Set the ith list to be newlist. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • newlist (list) – List to be saved at position i. Note that all items of this list will be converted to str internally.
class nctx.undirected.PropertyMapVecULong

PropertyMapVecULong serves as a container of lists, i.e. it is a list of lists of unsigned long - numbers > 0. The nested lists all contain values of a single data type as this is a link to the C++ code base (and C++ is a strongly typed language). This PropertyMap behaves like a python list returning lists as items when iterating over it.

The allowed data type can be recognized from the class name of the PropertyMap.

The PropertyMap can be converted to a Python list or a numpy array easily.

Parameters:
  • g (Graph) – The Graph object
  • name (str) – Name of the PropertyMap
  • init_val (int) – initial value for each inner list
  • init_size (int) – size of each inner list.
Example:
>>> my_map = PropertyMapVecULong(g, 'context1')
>>> my_map = PropertyMapVecULong(g, 'context1', 0, 10)
Example:
>>> import numpy as np
>>> np_arr = np.array([np.array(xi) for xi in my_map])
__contains__((object)arg2) → bool :

Allows to check for the existence of elements in the PropertyMap.

Example:
>>> [1,2,3] in my_map
False
__getitem__((int)arg2) → object :

This allows to access items using square brackets. See get_list for a detailed description of parameters.

Example:
>>> _ = my_map[2]
>>> _ = my_map[2][0]
__init__((object)arg1, (Graph)arg2, (str)arg3) → None
__init__((object)arg1, (Graph)arg2, (str)arg3, (int)arg4, (int)arg5) → None

Not commented yet

__iter__((object)arg1) → object

Not commented yet

__len__() → int :

Allows to retrieve the length of the PropertyMap.

Example:
>>> len(my_map)
10
__next__() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
__setitem__((int)arg2, (object)arg3) → None :

This allows to modify items using square brackets. See set_list for a detailed description of parameters.

Example:
>>> my_map[2] = [1,2,3,4,5]
>>> my_map[2][0] = 42
get_elem((int)i, (int)j) → int :

Retrieve the jth element of ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

jth element in ith list.
Return type:

int
get_list((int)i) → object :

Retrieve the ith list.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
Returns:

The list saved at position i.
Return type:

PropertyMapVecULong
get_name() → str :

Retrieve name of the property map. The name is mainly used when exporting the Graph to a file, e.g. using write_graphml.

Returns:The name of the list
Return type:string
next() → object :

Allows to iterater over the PropertyMap.

Example:
>>> for ctx in my_map:
...   pass
set_elem((int)i, (int)j, (int)newval) → None :

Set the jth element of ith list. Note that the type of the new element needs to be int or a type that can be implicitly converted to int. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • j (int) – Index of element.
  • newval (int) – The new element.
set_list((int)i, (object)newlist) → None :

Set the ith list to be newlist. Please see PropertyMapVecDouble.set_elem for an example of type conversion.

Parameters:
  • i (int) – Index of List
  • newlist (list) – List to be saved at position i. Note that all items of this list will be converted to int internally.

Graph I/O: Importing and Exporting to files

Summary
read_graphml Reads graph specification from a GraphML file and fills a graph object accordingly.
write_graphml Reads graph specification from a GraphML file and fills a graph object accordingly.
nctx.undirected.read_graphml((Graph)g, (str)filename, (list)pmaps) → None :

Reads graph specification from a GraphML file and fills a graph object accordingly. You can pass it a list of PropertyMaps to be filled. The names of the PropertyMaps must match the names in the graphml file. Graph properties that are present in the GraphML file but that do not have a matching PropertyMap in pmaps are ignored. See https://en.wikipedia.org/wiki/GraphML for more infos about the file format.

Parameters:
  • g (Graph) – The Graph object
  • filename (str) – Path to a GraphML file
  • pmaps (list) – A list of PropertyMaps that will be filled with the graph properties found in the file.
Example:
>>> import nctx.undirected as nctx
>>> g = nctx.Graph()
>>> name_map = nctx.PropertyMapStr(g, 'name')
>>> context_map = PropertyMapVecDouble(g, 'context')
>>> nctx.read_graphml(g, 'bcde.ctx.undir.graphml', [name_map, context_map])
nctx.undirected.write_graphml((Graph)g, (str)filename, (list)pmaps) → None :

Reads graph specification from a GraphML file and fills a graph object accordingly. You can pass it a list of PropertyMaps to be filled. The names of the PropertyMaps must match the names in the graphml file. Graph properties that are present in the GraphML file but that do not have a matching PropertyMap in pmaps are ignored. See https://en.wikipedia.org/wiki/GraphML for more infos about the file format.

Parameters:
  • g (Graph) – The Graph object
  • filename (str) – Path to a GraphML file
  • pmaps (list) – A list of PropertyMaps that will be filled with the graph properties found in the file.
Example:
>>> import nctx.undirected as nctx
>>> g = nctx.Graph()
>>> name_map = nctx.PropertyMapStr(g, 'name')
>>> context_map = PropertyMapVecDouble(g, 'context')
>>> nctx.read_graphml(g, 'bcde.ctx.undir.graphml', [name_map, context_map])
>>> name_map[3] = 'X'
>>> context_map[0][2] = .1337
>>> nctx.write_graphml(g, 'bcde.ctx.undir.modified.graphml', [name_map, context_map])

Helper Classes

class nctx.undirected.Edge

This is an internal description of an edge. It can’t be instantiated from Python. Just don’t bother with it.

class nctx.undirected.a_iter_ids

This is an internal description of the child-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → int

Not commented yet

next() → int :

See Graph.children how to use this to iterate over the children of a node.

class nctx.undirected.a_iter_ids_rev

This is an internal description of the parents-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → int

Not commented yet

next() → int :

See Graph.parents how to use this to iterate over the parents of a node.

class nctx.undirected.e_iter_ids

This is an internal description of the edges-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → Edge

Not commented yet

next() → Edge :

See Graph.edges how to use this to iterate over the edges of a graph.

class nctx.undirected.e_iter_ids_in

This is an internal description of the in-edges-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → Edge

Not commented yet

next() → Edge :

See Graph.edges how to use this to iterate over the edges of a graph.

class nctx.undirected.e_iter_ids_out

This is an internal description of the out-edges-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → Edge

Not commented yet

next() → Edge :

See Graph.edges how to use this to iterate over the edges of a graph.

class nctx.undirected.v_iter_ids

This is an internal description of the vertex-iterator.

__iter__((object)arg1) → object

Not commented yet

__len__() → int

Not commented yet

__next__() → int

Not commented yet

next() → int :

See Graph.vertices how to use this to iterate over the vertices of a graph.