Python math.tanh方法代碼示例

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def FollowBallManualPolicy():
  """An example of a minitaur following a ball."""
  env = minitaur_ball_gym_env.MinitaurBallGymEnv(render=True,
                                                 pd_control_enabled=True,
                                                 on_rack=False)
  observation = env.reset()
  sum_reward = 0
  steps = 100000
  for _ in range(steps):
    action = [math.tanh(observation[0] * 4)]
    observation, reward, done, _ = env.step(action)
    sum_reward += reward
    if done:
      tf.logging.info("Return is {}".format(sum_reward))
      observation = env.reset()
      sum_reward = 0 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def __init__(self):
        self.functions = {}
        self.add('sigmoid', sigmoid_activation)
        self.add('tanh', tanh_activation)
        self.add('sin', sin_activation)
        self.add('gauss', gauss_activation)
        self.add('relu', relu_activation)
        self.add('elu', elu_activation)
        self.add('lelu', lelu_activation)
        self.add('selu', selu_activation)
        self.add('softplus', softplus_activation)
        self.add('identity', identity_activation)
        self.add('clamped', clamped_activation)
        self.add('inv', inv_activation)
        self.add('log', log_activation)
        self.add('exp', exp_activation)
        self.add('abs', abs_activation)
        self.add('hat', hat_activation)
        self.add('square', square_activation)
        self.add('cube', cube_activation) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def test_tanh():

    from keras.activations import tanh as t
    test_values = get_standard_values()

    x = T.vector()
    exp = t(x)
    f = theano.function([x], exp)

    result = f(test_values)
    expected = [math.tanh(v) for v in test_values]

    print(result)
    print(expected)

    list_assert_equal(result, expected) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def trig(a, b=' '):
    if is_num(a) and isinstance(b, int):

        funcs = [math.sin, math.cos, math.tan,
                 math.asin, math.acos, math.atan,
                 math.degrees, math.radians,
                 math.sinh, math.cosh, math.tanh,
                 math.asinh, math.acosh, math.atanh]

        return funcs[b](a)

    if is_lst(a):
        width = max(len(row) for row in a)
        padded_matrix = [list(row) + (width - len(row)) * [b] for row in a]
        transpose = list(zip(*padded_matrix))
        if all(isinstance(row, str) for row in a) and isinstance(b, str):
            normalizer = ''.join
        else:
            normalizer = list
        norm_trans = [normalizer(padded_row) for padded_row in transpose]
        return norm_trans
    return unknown_types(trig, ".t", a, b) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def addToMemory(reward,averageReward):

    prob = 0.1
    if( reward > averageReward):
        prob = prob + 0.9 * math.tanh(reward - averageReward)
    else:
        prob = prob + 0.1 * math.tanh(reward - averageReward)
    print("average reward", averageReward, " reward ", reward, " prob", prob)
    #prob = prob / (rangeH - rangeL)
    #prob = reward / (1 + math.fabs(reward))
    #prob = (prob+1)/2
    if np.random.rand(1)<=prob :
        print("Adding reward",reward," based on prob ", prob)
        return True
    else:
        return False 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def get(self):
        self.x += self.config.get('dx', 0.1)

        val = eval(self.config.get('function', 'sin(x)'), {
            'sin': math.sin,
            'sinh': math.sinh,
            'cos': math.cos,
            'cosh': math.cosh,
            'tan': math.tan,
            'tanh': math.tanh,
            'asin': math.asin,
            'acos': math.acos,
            'atan': math.atan,
            'asinh': math.asinh,
            'acosh': math.acosh,
            'atanh': math.atanh,
            'log': math.log,
            'abs': abs,
            'e': math.e,
            'pi': math.pi,
            'x': self.x
        })

        return self.createEvent('ok', 'Sine wave', val) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def thetappp(self,z_in):
        T = self.T_k_in
        G = self.G_ksi
        J = self.J_in4
        l = self.l_in
        a = self.a
        z = z_in
        theta_tripleprime = (-(T*m.cosh(z/a)) + T*m.sinh(z/a)*m.tanh(l/(2*a)))/(G*J*a**2)

        return theta_tripleprime  

#Case 3 - Concentrated Torque at alpha*l with Pinned Ends
#T = Applied Concentrated Torsional Moment, Kip-in
#G = Shear Modulus of Elasticity, Ksi, 11200 for steel
#J = Torsinal Constant of Cross Section, in^4
#l = Span Lenght, in
#a = Torsional Constant
#alpa = load application point/l 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def _get_distorted_indices(self, nb_images):
        inverse = random.randint(0, 1)

        if inverse:
            scale = random.random()
            scale *= 0.21
            scale += 0.6
        else:
            scale = random.random()
            scale *= 0.6
            scale += 0.8

        frames_per_clip = nb_images

        indices = np.linspace(-scale, scale, frames_per_clip).tolist()
        if inverse:
            values = [math.atanh(x) for x in indices]
        else:
            values = [math.tanh(x) for x in indices]

        values = [x / values[-1] for x in values]
        values = [int(round(((x + 1) / 2) * (frames_per_clip - 1), 0)) for x in values]
        return values 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def step(self, action):
        obs, rew, done, info = self.env.step(action)
        self.raw_episode_return += rew
        self.episode_length += info.get('num_frames', 1)

        # optimistic asymmetric clipping from IMPALA paper
        squeezed = tanh(rew / 5.0)
        clipped = 0.3 * squeezed if rew < 0.0 else squeezed
        rew = clipped * 5.0

        if done:
            score = self.raw_episode_return

            info['episode_extra_stats'] = dict()
            level_name = self.unwrapped.level_name

            # add extra 'z_' to the summary key to put them towards the end on tensorboard (just convenience)
            level_name_key = f'z_{self.unwrapped.task_id:02d}_{level_name}'
            info['episode_extra_stats'][f'{level_name_key}_{RAW_SCORE_SUMMARY_KEY_SUFFIX}'] = score
            info['episode_extra_stats'][f'{level_name_key}_len'] = self.episode_length

        return obs, rew, done, info 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def adjust_learning_rate(optimizer, epoch, config):
    lr = get_current_lr(optimizer)
    if config.lr_scheduler.type == 'STEP':
        if epoch in config.lr_scheduler.lr_epochs:
            lr *= config.lr_scheduler.lr_mults
    elif config.lr_scheduler.type == 'COSINE':
        ratio = epoch / config.epochs
        lr = config.lr_scheduler.min_lr + \
            (config.lr_scheduler.base_lr - config.lr_scheduler.min_lr) * \
            (1.0 + math.cos(math.pi * ratio)) / 2.0
    elif config.lr_scheduler.type == 'HTD':
        ratio = epoch / config.epochs
        lr = config.lr_scheduler.min_lr + \
            (config.lr_scheduler.base_lr - config.lr_scheduler.min_lr) * \
            (1.0 - math.tanh(
                config.lr_scheduler.lower_bound
                + (config.lr_scheduler.upper_bound
                   - config.lr_scheduler.lower_bound)
                * ratio)
             ) / 2.0
    for param_group in optimizer.param_groups:
        param_group['lr'] = lr
    return lr 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def color(p):
	p = math.tanh(3*p)*.5+.5
	q = 1.-p*1.3
	r = 1.-abs(0.5-p)*1.3+.3*q
	p=1.3*p-.3
	i = int(p*255)	
	j = int(q*255)
	k = int(r*255)
	if j<0:
		j=0	
	if k<0:
		k=0
	if k >255:
		k=255
	if i<0:
		i = 0
	return ('\033[38;2;%d;%d;%dm' % (j, k, i)).encode() 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def tanh_activation(z):
    z = max(-60.0, min(60.0, 2.5 * z))
    return math.tanh(z) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def __call__(self, state, scope, pos, paramTypes, x):
        return unwrapForNorm(x, lambda y: math.tanh(y)) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def __call__(self, state, scope, pos, paramTypes, x, y, gamma, intercept):
	return math.tanh(gamma * dot(x, y, self.errcodeBase + 0, self.name, pos) + intercept) 

# 需要導入模塊: import math [as 別名]
# 或者: from math import tanh [as 別名]
def genpy(self, paramTypes, args, pos):
        return "math.tanh({0})".format(*args)