"""
Measures for estimating the information density of a given sample.
"""
from typing import Callable, Union
import numpy as np
from scipy.spatial.distance import cosine, euclidean
from sklearn.metrics.pairwise import pairwise_distances
from modAL.utils.data import modALinput
def similarize_distance(distance_measure: Callable) -> Callable:
"""
Takes a distance measure and converts it into a information_density measure.
Args:
distance_measure: The distance measure to be converted into information_density measure.
Returns:
The information_density measure obtained from the given distance measure.
"""
def sim(*args, **kwargs):
return 1/(1 + distance_measure(*args, **kwargs))
return sim
cosine_similarity = similarize_distance(cosine)
euclidean_similarity = similarize_distance(euclidean)
def information_density(X: modALinput, metric: Union[str, Callable] = 'euclidean') -> np.ndarray:
"""
Calculates the information density metric of the given data using the given metric.
Args:
X: The data for which the information density is to be calculated.
metric: The metric to be used. Should take two 1d numpy.ndarrays for argument.
Todo:
Should work with all possible modALinput.
Perhaps refactor the module to use some stuff from sklearn.metrics.pairwise
Returns:
The information density for each sample.
"""
# inf_density = np.zeros(shape=(X.shape[0],))
# for X_idx, X_inst in enumerate(X):
# inf_density[X_idx] = sum(similarity_measure(X_inst, X_j) for X_j in X)
#
# return inf_density/X.shape[0]
similarity_mtx = 1/(1+pairwise_distances(X, X, metric=metric))
return similarity_mtx.mean(axis=1)
