Source code for sksurgeryfred.algorithms.errors
# -*- coding: utf-8 -*-
"""
Functions for point based registration using Orthogonal Procrustes.
"""
import numpy as np
[docs]def expected_absolute_value(std_devs):
"""
Returns the expected absolute value of a normal
distribution with mean 0 and standard deviations std_dev
"""
onedsd = np.linalg.norm(std_devs)
variance = onedsd * onedsd
return variance