本文整理汇总了Python中xarray.DataArray.drop方法的典型用法代码示例。如果您正苦于以下问题：Python DataArray.drop方法的具体用法？Python DataArray.drop怎么用？Python DataArray.drop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xarray.DataArray的用法示例。
在下文中一共展示了DataArray.drop方法的3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: DoNotCooccur
# 需要导入模块: from xarray import DataArray [as 别名]
# 或者: from xarray.DataArray import drop [as 别名]
class DoNotCooccur(HACConstraint):
"""Do NOT merge co-occurring face tracks"""
def initialize(self, parent=None):
current_state = parent.current_state
clusters = [cluster for cluster in current_state.labels()]
n_clusters = len(clusters)
self._cooccur = DataArray(
np.zeros((n_clusters, n_clusters)),
[('i', clusters), ('j', clusters)])
for (segment1, track1), (segment2, track2) in current_state.co_iter(current_state):
i = current_state[segment1, track1]
j = current_state[segment2, track2]
if i == j:
continue
self._cooccur.loc[i, j] = 1
self._cooccur.loc[j, i] = 1
def mergeable(self, clusters, parent=None):
return self._cooccur.loc[clusters, clusters].sum().item() == 0.
def update(self, merged_clusters, new_cluster, parent=None):
# clusters that will be removed
_clusters = list(set(merged_clusters) - set([new_cluster]))
# update coooccurrence matrix
self._cooccur.loc[new_cluster, :] += self._cooccur.loc[_clusters, :].sum(dim='i')
self._cooccur.loc[:, new_cluster] += self._cooccur.loc[:, _clusters].sum(dim='j')
# remove clusters
self._cooccur = self._cooccur.drop(_clusters, dim='i').drop(_clusters, dim='j')示例2: CloseInTime
# 需要导入模块: from xarray import DataArray [as 别名]
# 或者: from xarray.DataArray import drop [as 别名]
class CloseInTime(HACConstraint):
def __init__(self, closer_than=30.0):
super(CloseInTime, self).__init__()
self.closer_than = closer_than
def initialize(self, parent=None):
current_state = parent.current_state
extended = current_state.empty()
for segment, track, cluster in current_state.itertracks(label=True):
extended_segment = Segment(segment.start - 0.5 * self.closer_than,
segment.end + 0.5 * self.closer_than)
extended[extended_segment, track] = cluster
clusters = [cluster for cluster in current_state.labels()]
n_clusters = len(clusters)
self._neighbours = DataArray(
np.zeros((n_clusters, n_clusters)),
[('i', clusters), ('j', clusters)])
for (segment1, track1), (segment2, track2) in extended.co_iter(extended):
i = extended[segment1, track1]
j = extended[segment2, track2]
self._neighbours.loc[i, j] += 1
self._neighbours.loc[j, i] += 1
def update(self, merged_clusters, new_cluster, parent=None):
# clusters that will be removed
_clusters = list(set(merged_clusters) - set([new_cluster]))
# update coooccurrence matrix
self._neighbours.loc[new_cluster, :] += self._neighbours.loc[_clusters, :].sum(dim='i')
self._neighbours.loc[:, new_cluster] += self._neighbours.loc[:, _clusters].sum(dim='j')
# remove clusters
self._neighbours = self._neighbours.drop(_clusters, dim='i').drop(_clusters, dim='j')
def mergeable(self, clusters, parent=None):
return connected_components(
self._neighbours.loc[clusters, clusters],
directed=False, return_labels=False) == 1示例3: HACModel
# 需要导入模块: from xarray import DataArray [as 别名]
# 或者: from xarray.DataArray import drop [as 别名]
#.........这里部分代码省略.........
# compute similarity if (and only if) clusters are mergeable
if not parent.constraint.mergeable([i, j], parent=parent):
continue
similarity = self.compute_similarity(i, j, parent=parent)
self._similarity.loc[i, j] = similarity
self._similarity.loc[j, i] = similarity
else:
for i, j in product(clusters, repeat=2):
# compute similarity if (and only if) clusters are mergeable
if not parent.constraint.mergeable([i, j], parent=parent):
continue
similarity = self.compute_similarity(i, j, parent=parent)
self._similarity.loc[i, j] = similarity
# NOTE - for now this (get_candidates / block) combination assumes
# that we merge clusters two-by-two...
def get_candidates(self, parent=None):
"""
Returns
-------
clusters : tuple
similarity : float
"""
_, n_j = self._similarity.shape
ij = np.argmax(self._similarity.data)
i = ij // n_j
j = ij % n_j
similarity = self._similarity[i, j].item()
clusters = [self._similarity.coords['i'][i].item(),
self._similarity.coords['j'][j].item()]
return clusters, similarity
def block(self, clusters, parent=None):
if len(clusters) > 2:
raise NotImplementedError(
'Constrained clustering merging 3+ clusters is not supported.'
)
i, j = clusters
self._similarity.loc[i, j] = -np.inf
self._similarity.loc[j, i] = -np.inf
def update(self, merged_clusters, into, parent=None):
# compute merged model
self._models[into] = self.compute_merged_model(merged_clusters,
parent=parent)
# remove old models and corresponding similarity
removed_clusters = list(set(merged_clusters) - set([into]))
for cluster in removed_clusters:
del self._models[cluster]
self._similarity = self._similarity.drop(removed_clusters, dim='i').drop(removed_clusters, dim='j')
# compute new similarities
# * all at once if model implements compute_similarities
# * one by one otherwise
remaining_clusters = list(set(self._models) - set([into]))
try:
if remaining_clusters:
# all at once (when available)
similarity = self.compute_similarities(
into, remaining_clusters, dim='j', parent=parent)
self._similarity.loc[into, remaining_clusters] = similarity
if self.is_symmetric:
similarity = similarity.rename({'j': 'i'})
else:
similarity = self.compute_similarities(
into, remaining_clusters, dim='i', parent=parent)
self._similarity.loc[remaining_clusters, into] = similarity
except NotImplementedError as e:
if remaining_clusters:
self._similarity.loc[into, remaining_clusters] = -np.inf
self._similarity.loc[remaining_clusters, into] = -np.inf
for cluster in remaining_clusters:
# compute similarity if (and only if) clusters are mergeable
if parent.constraint.mergeable([into, cluster], parent=parent):
similarity = self.compute_similarity(
into, cluster, parent=parent)
self._similarity.loc[into, cluster] = similarity
if not self.is_symmetric:
similarity = self.compute_similarity(
cluster, into, parent=parent)
self._similarity.loc[cluster, into] = similarity