added frolov to wn

Author: Vlad-Kor

model/distributions/torus/wrapped_normal/frolov.py

@@ -0,0 +1,42 @@
+import numpy as np
+from deterministic_gaussian_sampling_fibonacci import sample_gaussian_fibonacci
+
+from util.selectors.silder_log import LogSlider
+from model.distributions.torus.torus_sampling_schema import TorusSamplingSchema
+from util.selectors.silder_manual_input_wrapper import SliderManualInputWrapper as MI
+from util.gaus_util import GausUtil as gu
+
+
+class CFrolovWNSampling(TorusSamplingSchema):
+	def __init__(self):
+		def _check_input(val):
+			return val >= 1 and val <= 100003 and isinstance(val, int)
+
+		self.sample_options = [
+			MI(LogSlider("Number of Samples", 10, 100, 10000))
+		]
+
+	def get_name(self):
+		return "Classical Frolov"
+	
+	def sample(self, sample_options, distribution_options):
+		sample_count = sample_options[0].state
+
+		mean_x = distribution_options[0].state
+		mean_y = distribution_options[1].state
+		sigma_t = distribution_options[2].state
+		sigma_p = distribution_options[3].state
+		correlation = distribution_options[4].state
+
+		mu = np.array([mean_x, mean_y])
+
+		Cov = np.array([
+			[sigma_t**2, correlation * sigma_t * sigma_p],
+			[correlation * sigma_t * sigma_p, sigma_p**2]
+		])
+		
+		gaus_grid = sample_gaussian_fibonacci(mu, Cov, sample_count, "ClassicalFrolov")
+
+		# wrapp
+		gaus_grid[:,0] = gaus_grid[:,0] % (2 * np.pi)
+		return gaus_grid

model/distributions/torus/wrapped_normal/improved_frolov.py

@@ -0,0 +1,42 @@
+import numpy as np
+from deterministic_gaussian_sampling_fibonacci import sample_gaussian_fibonacci
+
+from util.selectors.silder_log import LogSlider
+from model.distributions.torus.torus_sampling_schema import TorusSamplingSchema
+from util.selectors.silder_manual_input_wrapper import SliderManualInputWrapper as MI
+from util.gaus_util import GausUtil as gu
+
+
+class IFrolovWNSampling(TorusSamplingSchema):
+	def __init__(self):
+		def _check_input(val):
+			return val >= 1 and val <= 100003 and isinstance(val, int)
+
+		self.sample_options = [
+			MI(LogSlider("Number of Samples", 10, 100, 10000))
+		]
+
+	def get_name(self):
+		return "Improved Frolov"
+	
+	def sample(self, sample_options, distribution_options):
+		sample_count = sample_options[0].state
+
+		mean_x = distribution_options[0].state
+		mean_y = distribution_options[1].state
+		sigma_t = distribution_options[2].state
+		sigma_p = distribution_options[3].state
+		correlation = distribution_options[4].state
+
+		mu = np.array([mean_x, mean_y])
+
+		Cov = np.array([
+			[sigma_t**2, correlation * sigma_t * sigma_p],
+			[correlation * sigma_t * sigma_p, sigma_p**2]
+		])
+		
+		gaus_grid = sample_gaussian_fibonacci(mu, Cov, sample_count, "ImprovedFrolov")
+
+		# wrapp
+		gaus_grid[:,0] = gaus_grid[:,0] % (2 * np.pi)
+		return gaus_grid

model/distributions/torus/wrapped_normal/wrapped_normal.py

@@ -8,6 +8,8 @@
 from model.distributions.torus.wrapped_normal.fibonacci import TorusFibRank1WNSampling
 from model.distributions.torus.wrapped_normal.fibonacci_kronecker import TorusFibKroneckerWNSampling
 from model.distributions.torus.wrapped_normal.cartesian import TorusFibCartWNSampling
+from model.distributions.torus.wrapped_normal.frolov import CFrolovWNSampling
+from model.distributions.torus.wrapped_normal.improved_frolov import IFrolovWNSampling
 from model.torus.torus import Torus
 class WrappedNormalTorusDistribution(TorusDistribution):
 	def __init__(self):
@@ -23,6 +25,8 @@ def __init__(self):
 			TorusFibRank1WNSampling(),
 			TorusFibKroneckerWNSampling(),
 			TorusFibCartWNSampling(),
+			CFrolovWNSampling(),
+			IFrolovWNSampling(),
 		]