wepy.resampling.distances.randomwalk module

This module here is part of the RandomWalk object that implements the distance metric for the RandomWalk walk system. This distance metric is a scaled version of the Manhattan Norm.

class wepy.resampling.distances.randomwalk.RandomWalkDistance

Bases: Distance

A class to implement the RandomWalkDistance metric for measuring differences between walker states. This is a normalized Manhattan distance measured between the difference in positions of the walkers.

Construct a RandomWalkDistance metric.

image(state)

Transform a state into a random walk image.

A random walk image is just the position of a walker in the N-dimensional space.

Parameters:

state (object implementing WalkerState) – A walker state object with positions in a numpy array of shape (N), where N is the the dimension of the random walk system.

Returns:

randomwalk_image – The positions of a walker in the N-dimensional space.

Return type:

array of floats of shape (N)

image_distance(image_a, image_b)

Compute the distance between the image of the two walkers.

image_aarray of float of shape (1, N)

Position of the first walker’s state.

image_b: array of float of shape (1, N)

Position of the second walker’s state.

Returns

distance: float

The normalized Manhattan distance.

distance(state_a, state_b)

Compute the distance between two states.

Parameters:

• state_a (object implementing WalkerState)

• state_b (object implementing WalkerState)

Returns:

distance – The distance between the two walker states

Return type:

float