added fibonacci-rank-1 to watson

Author: Vlad-Kor

model/distributions/sphere/watson/fibonachi.py

@@ -22,7 +22,7 @@ def __init__(self):
 		]
 
 	def get_name(self):
-		return "Fibonacci Lattice"
+		return "Fibonacci-Kronecker Lattice"
 
 
 	def sample(self, sample_options, distribution_options):

model/distributions/sphere/watson/fibonachi_rank1.py

@@ -0,0 +1,58 @@
+from abc import ABC, abstractmethod
+from matplotlib.pyplot import grid
+import numpy as np
+import scipy
+import scipy.integrate
+import scipy.interpolate
+import sphstat
+from pyrecest.backend import array
+from pyrecest.distributions import WatsonDistribution as WatsonDistributionPyrecest
+from scipy.special import erf, erfi, erfinv
+
+from model.distributions.cylinder.uniform.fibonacci_rank_1 import CylinderFibRank1UniformSampling
+from model.distributions.sphere.sphere_sampling_schema import SphereSamplingSchema
+from util.selectors.slider_fib import SliderFib
+from model.sphere.sphere import Sphere
+from util.selectors.silder_manual_input_wrapper import SliderManualInputWrapper as MI
+from model.distributions.sphere.watson.fibonachi import WatsonFibonachiSampling as wf
+
+
+
+class WatsonFibonachiRank1Sampling(SphereSamplingSchema):
+	def __init__(self):
+		self.sample_options = [
+			MI(SliderFib("Number of Samples", 3, 33, 21, 9, minus_1=True)),
+		]
+		self.sampler = CylinderFibRank1UniformSampling()
+	
+		
+	def get_name(self):
+		return "Fibonacci-Rank-1 Lattice"
+
+
+	def sample(self, sample_options, distribution_options):
+
+		sample_count = sample_options[0].state + 1 # because minus_1 is true slider displays fib(n)-1
+		k = distribution_options[0].state # kappa
+
+		x, y = self.sampler.get_rank_1(sample_count, sample_options[0].idx, without_first_point=True)
+
+		x = 2*x -1  # map x from [0,1] to [-1, 1]
+		phi = 2 * np.pi * y  # azimuthal angle, [0, 2pi] uniform
+
+		if k > 0:
+			w = 1 / (np.sqrt(k)) * wf.erfi_inv( x * erfi(np.sqrt(k)) )
+		elif k < 0:
+			la = -k
+			w = 1 / (np.sqrt(la)) * erfinv( x * erf(np.sqrt(la)) )
+		elif k == 0:
+			w = x
+
+
+		w = np.clip(w, -1.0, 1.0) # clamp to avoid sqrt warnings due to numerical issues
+
+		x_i_f_0 = w
+		x_i_f_1 = np.sqrt(1-w**2) * np.cos( phi)
+		x_i_f_2 = np.sqrt(1-w**2) * np.sin( phi)
+		x_i_f = np.column_stack((x_i_f_1, x_i_f_2, x_i_f_0)) # order so that mu=[0, 0, 1]
+		return x_i_f

model/distributions/sphere/watson/watson.py

@@ -10,6 +10,7 @@
 from model.distributions.sphere.watson.random_sampling import WatsonRandomSampling
 from model.distributions.sphere.watson.fibonachi import WatsonFibonachiSampling
 from model.distributions.sphere.watson.cartesian import WatsonCartesianSampling
+from model.distributions.sphere.watson.fibonachi_rank1 import WatsonFibonachiRank1Sampling			
 
 
 class WatsonDistribution(SphereDistribution):
@@ -22,6 +23,7 @@ def __init__(self):
 			WatsonRandomSampling(),
 			WatsonFibonachiSampling(),
 			WatsonCartesianSampling(),
+			WatsonFibonachiRank1Sampling(),
 		]
 
 	def get_name(self):