• G. Mias Lab »

    • Visibility Graph Community Detection

    Submodule pyiomica.visibilityGraphCommunityDetection

    Visibility Graph Community detection functions.

    Functions:

    createVisibilityGraph(data, times[, ...])

    Calculate adjacency matrix of visibility graph, create the networkx.Graph network.

    communityDetectByPathLength(G[, ...])

    Calculate community structure by shortest path length algorithm.
    createVisibilityGraph(data, times, graph_type='natural', weight=None, withsign=False)[source]

    Calculate adjacency matrix of visibility graph, create the networkx.Graph network.

    Parameters:
    data: 2d numpy.array

    Data to process

    times: 1d numpy.array

    Times corresponding to provided data points

    weight: str, Default None
    Type of normalization:

    None: no weighted

    ‘time’: weight = abs(times[i] - times[j])

    ‘tan’: weight = abs((data[i] - data[j])/(times[i] - times[j])) + 10**(-8)

    ‘distance’: weight = A[i, j] = A[j, i] = ((data[i] - data[j])**2 + (times[i] - times[j])**2)**0.5

    graph_type: str, Default ‘natural’
    Type of the visibility graph:

    “horizontal”, Horizontal Visibility Graph

    “natural”, Natural Visibility Graph

    “dual_horizontal”, Dual Perspective Horizontal Visibility Graph

    “dual_natural”, Dual Perspective Natural Visibility Graph

    withsign: boolean, Default False

    Whether to return the sign of adjacency matrix, the link from normal perspective VG is positive, the link from reflected perspective VG is negative

    Returns: tuple
    Tuple of two objects:
    networkx.Graph

    Graph of networkx type

    2d numpy.array

    Adjacency matrix

    Usage:

    A = createVisibilityGraph(data, times)

    communityDetectByPathLength(G, outputTimePoints=False, setSourceTarget=None, direction=None, cutoff=None)[source]

    Calculate community structure by shortest path length algorithm.

    Parameters:
    G: networkx.Graph

    Graph of networkx type

    outputTimePoints:boolean, default False

    Whether to output timepoints for communities or indices

    setSourceTarget:list, default None

    List for nodes to consider to use when finding weighted shortest path. First element used for source (starting node), and last for target (ending node)

    direction:str, default None
    The direction that nodes aggregate to communities:

    None: no specific direction, e.g. both sides.

    ‘left’: nodes can only aggregate to the left side hubs, e.g. early hubs

    ‘right’: nodes can only aggregate to the right side hubs, e.g. later hubs

    cutoff: int, float or str, Default None

    Cutoff is used to combine initial communities, e.g. whenever the shortest path length of two adjacent hub nodes is smaller than cutoff, the communities with the two hub nodes will be combined:

    int or float: the percentile of all shortest path length distribution, between 0 ~ 100

    ‘auto’: use optimized cutoff

    None: no cutoff

    Returns:
    list of list

    Detected communities in the form of nested list.

    Usage:

    c = communityDetectByPathLength(G)