reeb module

class reeb.Reeb(intervals, overlap=0.2, max_k=5)

Bases: object

A class to generate a Reeb graph for a given dataset and projection function

Parameters

intervals – The amount of intervals the Reeb-algorithm will divide the range of points in.
overlap – Optional, default 0.2 - The overlap of the intervals, in percentages (1 = 100%, .5 = 50%, etc.)
max_k – The maximum amount of clusters per interval that are tested. Limits nodes per interval to the range 1 to max_k.

fit(data, function=<function second_dimension>)

Fit the Reeb class to the data, and set a projection function.

Parameters

data – A list of data-points before they are passed through the projection function.
function – Optional, default second_dimension() - A function that maps a datapoint to a number.

Returns

This class instance.

rearrange_nodes()

Rearrange the nodes in the graph of this instance to make it a bit easier to look at when visualizing.

transform(data)

Generate a Reeb-graph for a given dataset using the pre-defined projection function.

Parameters: data – A list of data-points before they are passed through the projection function.
Returns: A Reeb-graph fitted to the given dataset with given parameters.

reeb.first_dimension(x)

Get the value of the first dimension of x. Can be used as projection function for reeb graphs.

reeb.second_dimension(x)

Get the value of the second dimension of x. Can be used as projection function for reeb graphs.