本文整理汇总了Python中neuralnet.NeuralNet.put_weights方法的典型用法代码示例。如果您正苦于以下问题：Python NeuralNet.put_weights方法的具体用法？Python NeuralNet.put_weights怎么用？Python NeuralNet.put_weights使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类neuralnet.NeuralNet的用法示例。
在下文中一共展示了NeuralNet.put_weights方法的1个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: MineSweeper
# 需要导入模块: from neuralnet import NeuralNet [as 别名]
# 或者: from neuralnet.NeuralNet import put_weights [as 别名]
#.........这里部分代码省略.........
if l < closest_so_far:
closest_so_far = l
closest_object = self.position - mine.position
self.closest_mine = ctr
ctr += 1
return closest_object
def update(self, all_mines):
'''
This is the real brains function. Takes an iterable of mines. It first
takes sensor readings and feed these to the ANN of our minesweeper.
The inputs are:
1) A vector (Vector2D) to the closest mine,
2) The "look at" vector (also a Vector2D).
The brain(ANN) returns 2 outputs, ltrack and rtrack - which are forces
applied on left and right tracks, respectively. Depending on these, the
acceleration and/or the rotation is calculated and the position vector
is updated accordingly.
'''
# Inputs to the brain.
inputs = []
# First input: vector to the closest mine.
closest_mine = self.get_closest_mine(all_mines)
closest_mine.normalize()
# Place the inputs on the input list
inputs.append(closest_mine.x)
inputs.append(closest_mine.y)
inputs.append(self.look_at.x)
inputs.append(self.look_at.y)
# Now, excite the brain and get the feedback
output = self.brain.excite(inputs, settings.BIAS, filter_sigmoid=True)
# Make sure we get back the expected number of outputs
if len(output) != settings.NUM_OUTPUTS:
raise Exception( 'An error occurred: The number of outputs from '
+'the ANN is not what was expected.')
self.ltrack, self.rtrack = output
rot_force = self.ltrack - self.rtrack
rot_force = clamp(rot_force,
-settings.MAX_TURN_RATE,
settings.MAX_TURN_RATE)
# New rotation and speed:
self.rotation += rot_force
self.speed = self.ltrack + self.rtrack
# Get the new look at:
self.look_at.x = -math.sin(self.rotation)
self.look_at.y = math.cos(self.rotation)
# Get the new position:
self.position += (self.look_at * self.speed)
# Wrap around the screen
if self.position.x > settings.WINDOW_WIDTH:
self.position.x = 0
if self.position.x < 0:
self.position.x = settings.WINDOW_WIDTH
if self.position.y > settings.WINDOW_HEIGHT:
self.position.y = 0
if self.position.y < 0:
self.position.y = settings.WINDOW_HEIGHT
# End update()
def check_for_mine(self, mines, size):
'''
checks for a collision with the closest mine.
'''
d = self.position - mines[self.closest_mine].position
if d.length() < (size + 5):
return self.closest_mine
return -1
def inc_fitness(self):
self.fitness += 1
def put_weights(self, weights):
self.brain.put_weights(weights)
def get_num_weights(self):
return self.brain.get_num_weights()
def __repr__(self):
return '<MineSweeper: ({0}, {1})>'.format(self.position.x,
self.position.y)