# @Author: Thomas Firmin <tfirmin>
# @Date: 2022-05-24T14:52:56+02:00
# @Email: thomas.firmin@univ-lille.fr
# @Project: Zellij
# @Last modified by: tfirmin
# @Last modified time: 2022-10-03T22:38:54+02:00
# @License: CeCILL-C (http://www.cecill.info/index.fr.html)
from zellij.core.addons import Distance
from zellij.core.variables import FloatVar, IntVar, CatVar, Constant
from scipy.spatial import distance
import numpy as np
import logging
logger = logging.getLogger("zellij.distances")
[docs]class Euclidean(Distance):
"""Euclidean distance
Compute the Euclidean distance between two points.
More info on `SciPy <https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.distance.euclidean.html#scipy.spatial.distance.euclidean>`__
Example
-------
>>> from zellij.core.variables import ArrayVar, FloatVar
>>> from zellij.utils.distances import Euclidean
>>> from zellij.core.loss_func import Loss
>>> from zellij.utils.benchmark import himmelblau
>>> from zellij.core.search_space import ContinuousSearchspace
>>> lf = Loss()(himmelblau)
>>> a = ArrayVar(FloatVar("float_1", 0,1),
... FloatVar("float_2", 0,1))
>>> sp = ContinuousSearchspace(a,lf, distance=Euclidean())
>>> p1,p2 = sp.random_point(), sp.random_point()
>>> print(p1)
[0.8922761649920034, 0.12709277668616326]
>>> print(p2)
[0.7730279148456985, 0.14715728189857524]
>>> sp.distance(p1,p2)
0.12092447863180801
See also
--------
:ref:`dist`: Distance addons
"""
def __call__(self, point_a, point_b):
return distance.euclidean(point_a, point_b, self.weights)
[docs]class Manhattan(Distance):
"""Manhattan distance
Compute the Manhattan distance between two points.
More info on `SciPy <https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.distance.cityblock.html#scipy.spatial.distance.cityblock>`__
Example
-------
>>> from zellij.core.variables import ArrayVar, FloatVar
>>> from zellij.utils.distances import Manhattan
>>> from zellij.core.loss_func import Loss
>>> from zellij.utils.benchmark import himmelblau
>>> from zellij.core.search_space import ContinuousSearchspace
>>> lf = Loss()(himmelblau)
>>> a = ArrayVar(FloatVar("float_1", 0,1),
... FloatVar("float_2", 0,1))
>>> sp = ContinuousSearchspace(a,lf, distance=Manhattan())
>>> p1,p2 = sp.random_point(), sp.random_point()
>>> print(p1)
[0.12946481931952147, 0.31940702810480137]
>>> print(p2)
[0.32347527913737095, 0.9356077155539462]
>>> sp.distance(p1,p2)
0.8102111472669943
See also
--------
:ref:`dist`: Distance addons
"""
def __call__(self, point_a, point_b):
return distance.cityblock(point_a, point_b, self.weights)
[docs]class Mixed(Distance):
"""Mixed distance
Compute a distance between two mixed points, using following equations:
.. math::
\\smash{
\\begin{cases}
\\delta_{i,j}^{(n)}=\\frac{|x_{i,n}-x_{j,n}|}{max_h(x_{h,n})-min_h(x_{h,n})}, \\quad \\text{if: $x_{h,n}$ is continuous or discrete}\\\\
\\begin{cases}
\\delta_{i,j}^{(n)}=0, \\quad \\text{if, $x_{i,n}=x_{j,n}$}\\\\
\\delta_{i,j}^{(n)}=1, \\quad \\text{otherwise}
\\end{cases} , \\quad \\text{if: $x_{h,n}$ is categorical}\\\\
d(x_i,x_j)=\\frac{\\sum_{n=1}^{p}(\\delta_{i,j}^{(n)})^2}{\\sum_{n=1}^{p}\\delta_{i,j}^{(n)}}\\\\
\\end{cases}}
Example
-------
>>> from zellij.core.variables import ArrayVar, IntVar, FloatVar, CatVar
>>> from zellij.utils.distances import Mixed
>>> from zellij.core.loss_func import Loss
>>> from zellij.utils.benchmark import himmelblau
>>> from zellij.core.search_space import MixedSearchspace
>>> a = ArrayVar(IntVar("int_1", 0,8),
>>> IntVar("int_2", 4,45),
>>> FloatVar("float_1", 2,12),
>>> CatVar("cat_1", ["Hello", 87, 2.56]))
>>> lf = Loss()(himmelblau)
>>> sp = MixedSearchspace(a,lf, distance=Mixed())
>>> p1,p2 = sp.random_point(), sp.random_point()
>>> print(p1)
[5, 34, 4.8808143412719485, 87]
>>> print(p2)
[3, 42, 2.8196595134477738, 'Hello']
>>> sp.distance(p1,p2)
0.5990169287736146
See also
--------
:ref:`dist`: Distance addons
"""
def __init__(self, search_space=None, weights=None):
self.weights = weights
self.operations = []
super(Mixed, self).__init__(search_space)
def __call__(self, point_a, point_b):
num = 0
denum = 0
for x, y, op, w in zip(point_a, point_b, self.operations, self.weights):
res = op(x, y) * w
num += res**2
denum += res
if denum == 0:
if num == 0:
return 0
else:
return float("inf")
else:
return num / denum
@Distance.target.setter
def target(self, object):
from zellij.core.search_space import MixedSearchspace
self._target = object
assert (
isinstance(self._target, MixedSearchspace) or object == None
), logger.error(
f"Target must be of type `MixedSearchspace`, got {object}"
)
if self._target:
if self.weights:
assert len(self.weights) == self._target.size, logger.error(
f"len(weights) must be equal to len(values) in `ArrayVar` of :ref:`Searchspace`"
)
else:
self.weights = [1] * self._target.size
self.operations = []
for v in self._target.values:
if isinstance(v, FloatVar) or isinstance(v, IntVar):
up, lo = v.up_bound, v.low_bound
self.operations.append(
lambda x, y: np.abs(x - y) / (up - lo)
)
elif isinstance(v, CatVar):
self.operations.append(lambda x, y: 1 if x == y else 0)
elif isinstance(v, Constant):
self.operations.append(lambda x, y: 0)
