Community Assembly Graphs

Community assembly graphs are an ecological tool that can be repurposed to understand possible evolutionary trajectories for populations under selection schemes that create complex ecology.

ec_ecology_toolbox.community_assembly_graph.CAG(start_node, user_func, num_nodes=10, normalize_w_self_edges=False, num_iterations=1500)

Main function that is used to explore the graph based on user function and start node Graph is explored based on continuous probability of reaching a node signified by edge weights Probability of reaching a node calculated from user given start node User defines the bumber of nodes to be discoered and returned

Parameters:

  • start_node (user-defined Node class) – starting node of exploration

  • user_func (function) – user defined function the returns all outgoing edges with weights from input node

  • num_nodes (int) – number of nodes to explore

  • normalize_w_self_edges (bool) – normalize nodes by including self edges if True. default: False

  • num_iterations (int) – number of iterations to run for random walk. default: 100

Returns:

explored graph until required number of nodes reached or full graph is explored

Return type:

dict {Node: dict {Node: float}}

class ec_ecology_toolbox.community_assembly_graph.Chr_Node(*args)

Sample class for user defined Node in graphs Nodes are characterized based on a single character e.g. integers Equality of nodes are based on having the same character Inequality is compared using number values if integers, ASCII values if string character Node class is hashable

__init__(*args)
class ec_ecology_toolbox.community_assembly_graph.Set_Node(*args)

Sample class for user defined Node in graphs Nodes are characterized based on a set of elements Equality of nodes are based on having the same set of elements Inequality is compared using length of the set i.e. number of elements in the set Node class is hashable and uses frozen sets

__init__(*args)
ec_ecology_toolbox.community_assembly_graph.random_walk_matrix_multiplication(graph, start_node, num_iterations)

Function used to calculate probabilities of ending up at a node i.e. node passing probability Implements iterative random walk theory using matrix multiplication Refer: https://chih-ling-hsu.github.io/2020/05/20/random-walks

Parameters:

  • graph (dict {Node: dict {Node: float}}) – manipulated graph with sink nodes on which the random walk algorithm is run

  • start_node (user-defined Node class) – the start node of the random walk

  • num_iterations (int) – number of iterations to run for the random walk. default: 100

Returns:

tuple including final passing probabilities after convergence and the node directory for index matching

Return type:

tuple (list [float], dict {Node: int})

ec_ecology_toolbox.community_assembly_graph.update_priorities(graph, leaves, start_node, node_passing_p, num_iterations)

Updates the probabilities of the discovered nodes in the priority queue Called when new paths to visited nodes or cycles are discovered in the explored graph

Parameters:

  • graph (dict {Node: dict {Node: float}}) – explored graph with all visited nodes and discovered paths

  • leaves (set {Node}) – all leaf nodes that have been discovered and are in the priority queue

  • start_node (user-defined Node class) – the starting node of exploration

  • node_passing_p (dict {Node: float}) – holds the updated passing probability of all discovered nodes

  • num_iterations (int) – number of iterations to run for the random walk. default: 100

Returns:

new priority queue with updated probabilities of all leaf nodes

Return type:

queue.PriorityQueue -> tuple (float, Node)