def __call__(self, rotation):
"""
Calculate probability of robot being in given position
Parameters
----------
rotation : float
Angle at which we see other robot.
Returns
-------
float
Probability in given rotation_angle
"""
d = math.pi - self.mean
p1 = gauss_distribution(rotation, self.mean, self.stdev)
p2 = gauss_distribution(rotation, -math.pi - d, self.stdev)
return max(p1, p2)
def __call__(self, angle):
"""
Calculate probability of robot being in given position
Parameters
----------
angle : float
Angle at which we see other robot.
Returns
-------
float
Probability in given rotation_angle
"""
return gauss_distribution(angle, self.mu, self.sigma)
def rotate(self, rotation_angle):
"""
Shift function accordingly to input angle and calculate probability
Parameters
----------
rotation_angle : float
Angle to shift function by.
Returns
-------
float
Probability after rotation.
"""
self.mu = (self.mu + rotation_angle) % 2 * math.pi
self.sigma = self.coef * self.sigma
return gauss_distribution(rotation_angle, self.mu, self.sigma)
def shift(self, rotation):
"""
Shift function accordingly to input angle and calculate probability
Parameters
----------
rotation : float
Angle to shift function by.
Returns
-------
float
Probability after rotation.
"""
mean = (self.mean + rotation) % (2 * math.pi) - math.pi
if self.mean > math.pi:
self.mean = mean - 2 * math.pi
else:
self.mean = mean
self.stdev = self.coef * self.stdev
return gauss_distribution(rotation, self.mean, self.stdev)
def test_max_value_location(self, mu, stdev):
xs = np.linspace(mu - 3 * stdev, mu + 3 * stdev, 250)
ys = [gauss_distribution(x, mu, stdev) for x in xs]
self.assertAlmostEqual(ys.index(max(ys)), len(ys) // 2, delta=3)
def _gaussian_normalized(x, mu, sigma):
k = 1 / gauss_distribution(mu, mu, sigma)
return k * gauss_distribution(x, mu, sigma)