MVC_distance¶

class turbustat.statistics.MVC_distance(data1, data2, fiducial_model=None)[source] [edit on github]¶

Bases: object

Distance metric for MVC and wrapper for whole analysis

Parameters:	data1 : dict dictionary containing necessary property arrays data2 : dict dictionary containing necessary property arrays fiducial_model : MVC Computed MVC object. use to avoid recomputing.

Methods Summary

Methods Documentation

distance_metric(low_cut=2.0, high_cut=64.0, verbose=False)[source] [edit on github]¶

Implements the distance metric for 2 MVC transforms. We fit the linear portion of the transform to represent the powerlaw A linear model with an interaction term is fit to the two powerlaws. The distance is the t-statistic of the interaction.

Parameters:	low_cut : int or float, optional Set the cut-off for low spatial frequencies. Visually, below ~2 deviates from the power law (for the simulation set). high_cut : int or float, optional Set the cut-off for high spatial frequencies. Values beyond the size of the root grid are found to have no meaningful contribution verbose : bool, optional Enables plotting.