6.7 KiB
6.7 KiB
In [ ]:
class Walker:
# ...
def sample_next_step(self, current_i, current_j, random_state=np.random):
""" Sample a new position for the walker. """
# Combine the next-step proposal with the context map to get a
# next-step probability map
next_step_map = self._next_step_proposal(current_i, current_j)
selection_map = self._compute_next_step_probability(next_step_map)
# Draw a new position from the next-step probability map
r = random_state.rand()
cumulative_map = np.cumsum(selection_map)
cumulative_map = cumulative_map.reshape(selection_map.shape)
i_next, j_next = np.argwhere(cumulative_map >= r)[0]
return i_next, j_next
def _next_step_proposal(self, current_i, current_j):
""" Create the 2D proposal map for the next step of the walker. """
# 2D Gaussian distribution , centered at current position,
# and with different standard deviations for i and j
grid_ii, grid_jj = self._grid_ii, self._grid_jj
sigma_i, sigma_j = self.sigma_i, self.sigma_j
rad = (
(((grid_ii - current_i) ** 2) / (sigma_i ** 2))
+ (((grid_jj - current_j) ** 2) / (sigma_j ** 2))
)
p_next_step = np.exp(-(rad / 2.0)) / (2.0 * np.pi * sigma_i * sigma_j)
return p_next_step / p_next_step.sum()
In [1]:
class Walker:
# ...
def _next_step_proposal(self, current_i, current_j):
""" Create the 2D proposal map for the next step of the walker. """
raise NotImplementedError("`_next_step_proposal` not implemented")
In [ ]:
class GaussianWalker(Walker):
# ...
def _next_step_proposal(self, current_i, current_j):
class RectangularWalker(Walker):
# ...
def _next_step_proposal(self, current_i, current_j):
class JumpingWalker(Walker):
# ...
def _next_step_proposal(self, current_i, current_j):
In [ ]:
class Walker:
# ...
def _compute_next_step_probability(self, next_step_map):
""" Compute the next step probability map from next step proposal and
context map. """
next_step_probability = next_step_map * self.context_map
next_step_probability /= next_step_probability.sum()
return next_step_probability
In [ ]:
class GaussianWalkerWithProductInteraction(Walker):
def _next_step_proposal(self, current_i, current_j):
# ...
def _compute_next_step_probability(self, next_step_map):
# ...
class GaussianWalkerWithSumInteraction(Walker):
def _next_step_proposal(self, current_i, current_j):
# ...
def _compute_next_step_probability(self, next_step_map):
# ...
class RectangularWalkerWithProductInteraction(Walker):
def _next_step_proposal(self, current_i, current_j):
# ...
def _compute_next_step_probability(self, next_step_map):
# ...
class RectangularWalkerWithSumInteraction(Walker):
def _next_step_proposal(self, current_i, current_j):
# ...
def _compute_next_step_probability(self, next_step_map):
# ...
In [ ]:
class Walker:
def __init__(self, size, context_map, next_step_proposal, next_step_proposal_arguments):
self.next_step_proposal = next_step_proposal
# ...
def sample_next_step(self, current_i, current_j, random_state=np.random):
""" Sample a new position for the walker. """
# Combine the next-step proposal with the context map to get a
# next-step probability map
next_step_map = self.next_step_proposal(current_i, current_j, **next_step_proposal_arguments)
selection_map = self._compute_next_step_probability(next_step_map)
# Draw a new position from the next-step probability map
r = random_state.rand()
cumulative_map = np.cumsum(selection_map)
cumulative_map = cumulative_map.reshape(selection_map.shape)
i_next, j_next = np.argwhere(cumulative_map >= r)[0]
return i_next, j_next