Source code for glue.core.data_collection

from contextlib import contextmanager

from glue.core.message import (DataCollectionAddMessage,
from glue.core.registry import Registry
from glue.core.link_manager import LinkManager
from import Data, BaseCartesianData
from glue.core.hub import Hub, HubListener
from glue.core.coordinates import WCSCoordinates
from glue.config import settings, data_translator
from glue.utils import as_list, common_prefix

__all__ = ['DataCollection']

[docs]class DataCollection(HubListener): """The top-level object for interacting with datasets in Glue. DataCollections have the following responsibilities: * Providing a way to retrieve and store data * Broadcasting messages when data are added or removed * Keeping each managed data set's list of :class:`~glue.core.component.DerivedComponent` instances up-to-date * Creating the hub that all other objects should use to communicate with one another (stored in ``self.hub``) """ def __init__(self, data=None): """ :param data: :class:`` object, or list of such objects """ super(DataCollection, self).__init__() self._link_manager = LinkManager(self) self._data = [] self.hub = None self._disable_sync_link_manager = 0 self._subset_groups = [] self.register_to_hub(Hub()) self.extend(as_list(data or [])) self._sg_count = 0 self._link_manager.register_to_hub(self.hub)
[docs] @property def data(self): """ The :class:`` objects in the collection """ return self._data
[docs] def append(self, data): """ Add a new dataset to this collection. Appending emits a DataCollectionAddMessage. It also updates the list of DerivedComponents that each data set can work with. :param data: :class:`` object to add """ if isinstance(data, list): self.extend(data) return if data in self: return if not isinstance(data, BaseCartesianData): raise TypeError("Only BaseCartesianData subclasses can be used at this time") self._data.append(data) if self.hub: data.register_to_hub(self.hub) for s in data.subsets: s.register() msg = DataCollectionAddMessage(self, data) self.hub.broadcast(msg) self._sync_link_manager()
[docs] def extend(self, data): """Add several new datasets to this collection See :meth:`append` for more information :param data: List of data objects to add """ # Wait until all datasets are added to sync the link manager with self._ignore_link_manager_update(): for d in data: self.append(d) self._sync_link_manager()
[docs] def remove(self, data): """ Remove a data set from the collection Emits a DataCollectionDeleteMessage :param data: the object to remove :type data: :class:`` """ if data not in self._data: return self._data.remove(data) Registry().unregister(data, Data) if self.hub: msg = DataCollectionDeleteMessage(self, data) self.hub.broadcast(msg)
[docs] def clear(self): with self._ignore_link_manager_update(): for data in list(self): self.remove(data)
def _sync_link_manager(self): """ update the LinkManager, so all the DerivedComponents for each data set are up-to-date """ if getattr(self, '_disable_sync_link_manager', False): return # Avoid circular calls with self._ignore_link_manager_update(): self._link_manager.update_externally_derivable_components() @contextmanager def _ignore_link_manager_update(self): self._disable_sync_link_manager += 1 yield self._disable_sync_link_manager -= 1 def _merge_link(self, link): pass
[docs] def register_to_hub(self, hub): """ Register managed data objects to a hub. :param hub: The hub to register with :type hub: :class:`~glue.core.hub.Hub` """ if self.hub is hub: return if self.hub is not None: raise RuntimeError("Data Collection already registered " "to a different Hub") if not isinstance(hub, Hub): raise TypeError("Input is not a Hub object: %s" % type(hub)) self.hub = hub # re-assign all data, subset hub instances to this hub for d in self._data: d.register_to_hub(hub) for s in d.subsets: s.register() hub.subscribe(self, ComponentsChangedMessage, lambda msg: self._sync_link_manager(), filter=lambda x: x.sender in self._data)
[docs] def new_subset_group(self, label=None, subset_state=None): """ Create and return a new Subset Group. :param label: The label to assign to the group :type label: str :param subset_state: The state to initialize the group with :type subset_state: :class:`~glue.core.subset.SubsetState` :returns: A new :class:`~glue.core.subset_group.SubsetGroup` """ from glue.core.subset_group import SubsetGroup color = settings.SUBSET_COLORS[self._sg_count % len(settings.SUBSET_COLORS)] self._sg_count += 1 label = label or 'Subset %i' % self._sg_count result = SubsetGroup(color=color, label=label, subset_state=subset_state) self._subset_groups.append(result) result.register(self) return result
[docs] def remove_subset_group(self, subset_grp): """ Remove an existing :class:`~glue.core.subset_group.SubsetGroup` """ if subset_grp not in self._subset_groups: return # remove from list first, so that group appears deleted # by the time the first SubsetDelete message is broadcast self._subset_groups.remove(subset_grp) for s in subset_grp.subsets: s.delete() subset_grp.unregister(self.hub)
[docs] def suggest_merge_label(self, *data): """ Determine what merge label to suggest given datasets """ # Find longest common prefix for data suggestion = common_prefix([d.label for d in data]) if len(suggestion) < 3: suggestion = 'Merged data' # Now check if the suggestion already exists, and if so add a suffix labels = self.labels if suggestion in labels: suffix = 2 while "{0} [{1}]".format(suggestion, suffix) in labels: suffix += 1 suggestion = "{0} [{1}]".format(suggestion, suffix) return suggestion
[docs] @property def labels(self): return [d.label for d in self]
[docs] def merge(self, *data, **kwargs): """ Merge two or more datasets into a single dataset. This has the following effects: All components from all datasets are added to the first argument All datasets except the first argument are removed from the collection Any component name conflicts are disambiguated The pixel and world components apart from the first argument are discarded :note: All arguments must have the same shape :param data: One or more :class:`` instances. :returns: self """ if len(data) < 2: raise ValueError("merge requires 2 or more arguments") shp = data[0].shape for d in data: if d.shape != shp: raise ValueError("All arguments must have the same shape") label = kwargs.get('label', data[0].label) master = Data(label=label) self.append(master) master.coords = data[0].coords for i, d in enumerate(data): if isinstance(d.coords, WCSCoordinates): master.coords = d.coords break # Find ambiguous components (ones which have labels in more than one # dataset from collections import Counter clabel_count = Counter([c.label for d in data for c in d.main_components + d.derived_components]) for d in data: for c in d.components: if c in master.components: # already present (via a link) continue # Don't include coordinate components here as they will be # recomputed separately once the first non-coordinate component # is added. if c in d.coordinate_components: continue lbl = c.label if clabel_count[lbl] > 1: lbl = lbl + " [{0}]".format(d.label) c._label = lbl c.parent = master master.add_component(d.get_component(c), c) self.remove(d) return master
[docs] @property def subset_groups(self): """ tuple of current :class:`Subset Groups <glue.core.subset_group.SubsetGroup>` """ return tuple(self._subset_groups)
def __contains__(self, obj): return (obj in self._data or obj in self.subset_groups or any([data.label == obj for data in self._data])) def __getitem__(self, key): if isinstance(key, str): matches = [data for data in self._data if data.label == key] if len(matches) == 0: raise ValueError("No data found with the label '{0}'".format(key)) elif len(matches) > 1: raise ValueError("Several datasets were found with the label '{0}'".format(key)) else: return matches[0] else: return self._data[key] def __setitem__(self, key, data): """ Add a dataset to the data collection. This can be either a :class:`` object, which will then have its label set to the specified key, or another kind of object which will be automatically translated into a :class:`` object. """ if not isinstance(key, str): raise TypeError("item key should be a string, but got {0}".format(type(key))) if not isinstance(data, Data): handler, preferred = data_translator.get_handler_for(data) data = handler.to_data(data) data._preferred_translation = preferred data.label = key for existing_data in self._data[:]: if existing_data.label == key: self.remove(existing_data) self.append(data) def __iter__(self): return iter(self._data) def __len__(self): return len(self._data)
[docs] def index(self, item): return self._data.index(item)
def __str__(self): if len(self) == 1: result = "DataCollection (1 data set)\n\t" else: result = "DataCollection (%i data sets)\n\t" % len(self) result += '\n\t'.join("%3i: %s" % (i, d.label) for i, d in enumerate(self)) return result def __repr__(self): return self.__str__() def __bool__(self): return True def __nonzero__(self): return True