Lets test Perlin and PCA on Perlin

This commit is contained in:
Dominik Moritz Roth 2023-05-04 12:18:33 +02:00
parent d04f245e9b
commit c0913ba965

65
test.py
View File

@ -9,6 +9,7 @@ from columbus import env
from columbus.observables import Observable, CnnObservable from columbus.observables import Observable, CnnObservable
import colorednoise as cn import colorednoise as cn
from perlin_noise import PerlinNoise
from priorConditionedAnnealing import pca from priorConditionedAnnealing import pca
@ -33,7 +34,7 @@ def getAvaibleEnvs():
yield getattr(env, s) yield getattr(env, s)
def loadConfigDefinedEnv(EnvClass): def loadConfigDefinedEnv(EnvClass, alg_name):
p = input('[Path to config> ') p = input('[Path to config> ')
with open(p, 'r') as f: with open(p, 'r') as f:
docs = list([d for d in yaml.safe_load_all( docs = list([d for d in yaml.safe_load_all(
@ -57,7 +58,7 @@ def loadConfigDefinedEnv(EnvClass):
print('Unable to find key "'+key+'"') print('Unable to find key "'+key+'"')
path = input('[Path> ') path = input('[Path> ')
print(cur) print(cur)
return EnvClass(fps=30, **cur) return EnvClass(fps=30, title_appendix=' ['+alg_name+']', **cur)
def chooseEnv(alg_name): def chooseEnv(alg_name):
@ -74,7 +75,7 @@ def chooseEnv(alg_name):
print( print(
'[!] That is a number, but not one that makes sense in this context...') '[!] That is a number, but not one that makes sense in this context...')
if envs[i] in [env.ColumbusConfigDefined]: if envs[i] in [env.ColumbusConfigDefined]:
return loadConfigDefinedEnv(envs[i]) return loadConfigDefinedEnv(envs[i], alg_name)
Env = envs[i] Env = envs[i]
return Env(fps=30, agent_draw_path=True, path_decay=1/1024, title_appendix=' ['+alg_name+']', max_steps=30*10, clear_path_on_reset=False) return Env(fps=30, agent_draw_path=True, path_decay=1/1024, title_appendix=' ['+alg_name+']', max_steps=30*10, clear_path_on_reset=False)
@ -111,12 +112,62 @@ class Colored_Noise():
self.beta, (self.dim_a, self.samples), random_state=rand_seed()) self.beta, (self.dim_a, self.samples), random_state=rand_seed())
class PCA_Noise(): class Perlin_Noise():
def __init__(self, dim_a=2, kernel_func='SE_1.41_1', window=64, ssf=-1): def __init__(self, scale=0.05, octaves=1, dim_a=2):
self.scale = scale
self.octaves = octaves
self.dim_a = dim_a
self.magic = 3.14159 # Axis offset
# We want to genrate samples, that approx ~N(0,1)
self.normal_factor = 0.0471
self.reset()
def __call__(self, obs, env):
self.index += 1
return [self.noise([self.index*self.scale, self.magic*a]) / self.normal_factor
for a in range(self.dim_a)]
def reset(self):
self.index = 0
self.noise = PerlinNoise(octaves=self.octaves, seed=rand_seed())
class Perlin_PCA_Noise():
def __init__(self, dim_a=2, kernel_func='SE_1.41_1', window=64, ssf=-1, f_sigma=1):
self.dim_a = dim_a self.dim_a = dim_a
self.kernel_func = kernel_func self.kernel_func = kernel_func
self.window = window self.window = window
self.ssf = ssf self.ssf = ssf
self.f_sigma = f_sigma
self.index = 0
self.perlin = Perlin_Noise()
self.reset()
def __call__(self, obs, env):
if self.ssf != -1 and self.index % self.ssf == 0:
self.traj = [[0]*len(self.traj[0])]
traj = th.Tensor(self.traj).unsqueeze(0)
eps = th.Tensor(self.perlin(None, None)).unsqueeze(0)
sample = self.dist.sample(traj, self.f_sigma, epsilon=eps).squeeze(0)
self.traj.append(sample)
self.index += 1
return sample
def reset(self):
self.dist = pca.PCA_Distribution(
action_dim=self.dim_a, par_strength='CONT_DIAG', kernel_func=self.kernel_func, window=self.window)
self.dist.proba_distribution(th.Tensor([[0]*2]), th.Tensor([[1]*2]))
self.traj = [[0]*self.dim_a]
self.perlin.reset()
class PCA_Noise():
def __init__(self, dim_a=2, kernel_func='SE_1.41_1', window=64, ssf=-1, f_sigma=1):
self.dim_a = dim_a
self.kernel_func = kernel_func
self.window = window
self.ssf = ssf
self.f_sigma = f_sigma
self.index = 0 self.index = 0
self.reset() self.reset()
@ -124,7 +175,7 @@ class PCA_Noise():
if self.ssf != -1 and self.index % self.ssf == 0: if self.ssf != -1 and self.index % self.ssf == 0:
self.traj = [[0]*len(self.traj[0])] self.traj = [[0]*len(self.traj[0])]
traj = th.Tensor(self.traj).unsqueeze(0) traj = th.Tensor(self.traj).unsqueeze(0)
sample = self.dist.sample(traj).squeeze(0) sample = self.dist.sample(traj, self.f_sigma).squeeze(0)
self.traj.append(sample) self.traj.append(sample)
self.index += 1 self.index += 1
return sample return sample
@ -172,7 +223,7 @@ def rand_seed():
def choosePlayType(): def choosePlayType():
options = {'human': human_input, 'PCA': PCA_Noise(), options = {'human': human_input, 'PCA': PCA_Noise(),
'REX': Colored_Noise(beta=0), 'PINK': Colored_Noise(beta=1), 'BROWN': Colored_Noise(beta=2), 'BETA.5': Colored_Noise(beta=.5), 'PINK_PCA': Colored_PCA_Noise(beta=1)} 'REX': Colored_Noise(beta=0), 'PINK': Colored_Noise(beta=1), 'BROWN': Colored_Noise(beta=2), 'BETA.5': Colored_Noise(beta=.5), 'PINK_PCA': Colored_PCA_Noise(beta=1), 'Precise_PCA': PCA_Noise(f_sigma=0.33), 'Perlin': Perlin_Noise(scale=0.05, octaves=1), 'FastPerlin': Perlin_Noise(scale=0.2, octaves=1), 'SlowPerlin': Perlin_Noise(scale=0.0125, octaves=1), 'Perlin_3': Perlin_Noise(scale=0.05, octaves=3), 'Perlin_8': Perlin_Noise(scale=0.05, octaves=8), 'Perlin_PCA': Perlin_PCA_Noise()}
for i, name in enumerate(options): for i, name in enumerate(options):
print('['+str(i)+'] '+name) print('['+str(i)+'] '+name)
while True: while True: