def getBoundingRegionInfo(self, region):
self._updateContentsIfNecessary(region.chr)
if region.chr in self._contents:
brInfoHolder = self._contents[region.chr]
#Temporary, to store old preprocessed boundingRegion.shelve files
isDict = isinstance(brInfoHolder, dict)
if isDict:
brStarts = brInfoHolder.keys()
else:
brStarts = brInfoHolder.brStarts
#idx = self._contents[region.chr].keys().bisect_right(region.start)
idx = bisect_right(brStarts, region.start)
if idx > 0:
if isDict:
brInfo = brInfoHolder[brStarts[idx - 1]]
else:
brInfo = brInfoHolder.brInfos[idx - 1]
if region.start < brInfo.end and region.end <= brInfo.end:
return brInfo
if not self._minimalRegion == region:
#
#There are bounding regions in the same chromosome, but not any encompassing the user bin
#Thus the bounding regions are explicitly defined (not just the complete chromosome)
#
from gtrackcore.util.CommonFunctions import prettyPrintTrackName
raise OutsideBoundingRegionError("The analysis region '%s' is outside the bounding regions of track: %s" \
% (region, prettyPrintTrackName(self._trackName)))
return BoundingRegionInfo(region.start, region.end, 0, 0, 0, 0)
def getTrackView(self, region):
allowOverlaps = self._trackFormatReq.allowOverlaps()
borderHandling = self._trackFormatReq.borderHandling()
assert (allowOverlaps is not None)
assert (borderHandling is not None)
origTrackView = self._getRawTrackView(region, borderHandling,
allowOverlaps)
if self.formatConverters is None:
self.formatConverters = getFormatConverters(
origTrackView.trackFormat, self._trackFormatReq)
if self.formatConverters == []:
raise IncompatibleTracksError(prettyPrintTrackName(self.trackName) + ' with format: '\
+ str(origTrackView.trackFormat) +
('(' + origTrackView.trackFormat._val + ')' if origTrackView.trackFormat._val else '') + \
' does not satisfy ' + str(self._trackFormatReq))
if not self.formatConverters[0].canHandle(origTrackView.trackFormat,
self._trackFormatReq):
raise IncompatibleTracksError(getClassName(self.formatConverters[0]) +\
' does not support conversion from ' + str(origTrackView.trackFormat) + \
' to ' + str(self._trackFormatReq))
return self.formatConverters[0].convert(origTrackView)
def getBoundingRegionInfo(self, region):
self._updateContentsIfNecessary(region.chr)
if region.chr in self._contents:
brInfoHolder = self._contents[region.chr]
#Temporary, to store old preprocessed boundingRegion.shelve files
isDict = isinstance(brInfoHolder, dict)
if isDict:
brStarts = brInfoHolder.keys()
else:
brStarts = brInfoHolder.brStarts
#idx = self._contents[region.chr].keys().bisect_right(region.start)
idx = bisect_right(brStarts, region.start)
if idx > 0:
if isDict:
brInfo = brInfoHolder[brStarts[idx-1]]
else:
brInfo = brInfoHolder.brInfos[idx-1]
if region.start < brInfo.end and region.end <= brInfo.end:
return brInfo
if not self._minimalRegion == region:
#
#There are bounding regions in the same chromosome, but not any encompassing the user bin
#Thus the bounding regions are explicitly defined (not just the complete chromosome)
#
from gtrackcore.util.CommonFunctions import prettyPrintTrackName
raise OutsideBoundingRegionError("The analysis region '%s' is outside the bounding regions of track: %s" \
% (region, prettyPrintTrackName(self._trackName)))
return BoundingRegionInfo(region.start, region.end, 0, 0, 0, 0)
def getAllBoundingRegions(self):
if not self.fileExists():
from gtrackcore.util.CommonFunctions import prettyPrintTrackName
raise BoundingRegionsNotAvailableError('Bounding regions not available for track: ' + \
prettyPrintTrackName(self._trackName))
for chr in GenomeInfo.getExtendedChrList(self._genome):
for reg in self.getAllBoundingRegionsForChr(chr):
yield reg
def getAllBoundingRegions(self):
if not self.fileExists():
from gtrackcore.util.CommonFunctions import prettyPrintTrackName
raise BoundingRegionsNotAvailableError('Bounding regions not available for track: ' + \
prettyPrintTrackName(self._trackName))
for chr in GenomeInfo.getExtendedChrList(self._genome):
for reg in self.getAllBoundingRegionsForChr(chr):
yield reg
def _getRawTrackView(self, region, borderHandling, allowOverlaps):
assert len(region) == 1
from collections import OrderedDict
from gtrackcore.track.memmap.CommonMemmapFunctions import findEmptyVal
from gtrackcore.track.core.TrackView import TrackView
import numpy as np
geSource = ExternalTrackManager.getGESourceFromGalaxyOrVirtualTN(self.trackName, region.genome)
prefixList = geSource.getPrefixList()
valDataType = geSource.getValDataType()
valDim = geSource.getValDim()
weightDataType = geSource.getEdgeWeightDataType()
weightDim = geSource.getEdgeWeightDim()
startList, endList, valList, strandList, idList, edgesList, weightsList = [None]*7
extraLists=OrderedDict()
tf = TrackFormat.createInstanceFromPrefixList(prefixList, valDataType, valDim, \
weightDataType, weightDim)
if allowOverlaps and (tf.isDense() or geSource.hasNoOverlappingElements()):
raise IncompatibleTracksError(prettyPrintTrackName(self.trackName) + ' with format: '\
+ str(tf) + ' does not satisfy ' + str(self._trackFormatReq))
denseAndInterval = tf.isDense() and tf.isInterval()
numEls = 2 if denseAndInterval else 1
if valDataType == 'S':
valDataType = 'S2'
if weightDataType == 'S':
weightDataType = 'S2'
for prefix in prefixList:
if prefix == 'start':
startList = np.array([-1], dtype='int32')
elif prefix == 'end':
if denseAndInterval:
endList = np.array([0, 1], dtype='int32')
else:
endList = np.array([0], dtype='int32')
elif prefix == 'val':
valList = np.array([findEmptyVal(valDataType)] * valDim * numEls, \
dtype=valDataType).reshape((numEls, valDim) if valDim > 1 else numEls)
elif prefix == 'strand':
strandList = np.array([1] * numEls, dtype='int8')
elif prefix == 'id':
idList = np.array([''] * numEls, dtype='S1')
elif prefix == 'edges':
edgesList = np.array([['']] * numEls, dtype='S1')
elif prefix == 'weights':
weightsList = np.array([[[findEmptyVal(weightDataType)]]] * weightDim * numEls, \
dtype=weightDataType).reshape((numEls, 1, weightDim) if weightDim > 1 else (numEls, 1))
else:
extraLists[prefix] = np.array([''] * numEls, dtype='S1')
return TrackView(region, startList, endList, valList, strandList, idList, edgesList, weightsList, borderHandling, allowOverlaps, extraLists)
def _getBoundingRegionShelve(self, trackName):
if trackName in [None, []] or ExternalTrackManager.isVirtualTrack(trackName):
brShelve = None
else:
brShelve = BoundingRegionShelve(self.genome, trackName, allowOverlaps=False)
if not brShelve.fileExists():
raise BoundingRegionsNotAvailableError('Bounding regions not available for track: ' + \
prettyPrintTrackName(trackName))
return brShelve
def getTrackView(self, region):
allowOverlaps = self._trackFormatReq.allowOverlaps()
borderHandling = self._trackFormatReq.borderHandling()
assert(allowOverlaps is not None)
assert(borderHandling is not None)
origTrackView = self._getRawTrackView(region, borderHandling, allowOverlaps)
if self.formatConverters is None:
self.formatConverters = getFormatConverters(origTrackView.trackFormat, self._trackFormatReq)
if self.formatConverters == []:
raise IncompatibleTracksError(prettyPrintTrackName(self.trackName) + ' with format: '\
+ str(origTrackView.trackFormat) +
('(' + origTrackView.trackFormat._val + ')' if origTrackView.trackFormat._val else '') + \
' does not satisfy ' + str(self._trackFormatReq))
if not self.formatConverters[0].canHandle(origTrackView.trackFormat, self._trackFormatReq):
raise IncompatibleTracksError(getClassName(self.formatConverters[0]) +\
' does not support conversion from ' + str(origTrackView.trackFormat) + \
' to ' + str(self._trackFormatReq))
return self.formatConverters[0].convert(origTrackView)
def _getRawTrackView(self, region, borderHandling, allowOverlaps):
assert len(region) == 1
from collections import OrderedDict
from gtrackcore.track.memmap.CommonMemmapFunctions import findEmptyVal
from gtrackcore.track.core.TrackView import TrackView
import numpy as np
geSource = ExternalTrackManager.getGESourceFromGalaxyOrVirtualTN(
self.trackName, region.genome)
prefixList = geSource.getPrefixList()
valDataType = geSource.getValDataType()
valDim = geSource.getValDim()
weightDataType = geSource.getEdgeWeightDataType()
weightDim = geSource.getEdgeWeightDim()
startList, endList, valList, strandList, idList, edgesList, weightsList = [
None
] * 7
extraLists = OrderedDict()
tf = TrackFormat.createInstanceFromPrefixList(prefixList, valDataType, valDim, \
weightDataType, weightDim)
if allowOverlaps and (tf.isDense()
or geSource.hasNoOverlappingElements()):
raise IncompatibleTracksError(prettyPrintTrackName(self.trackName) + ' with format: '\
+ str(tf) + ' does not satisfy ' + str(self._trackFormatReq))
denseAndInterval = tf.isDense() and tf.isInterval()
numEls = 2 if denseAndInterval else 1
if valDataType == 'S':
valDataType = 'S2'
if weightDataType == 'S':
weightDataType = 'S2'
for prefix in prefixList:
if prefix == 'start':
startList = np.array([-1], dtype='int32')
elif prefix == 'end':
if denseAndInterval:
endList = np.array([0, 1], dtype='int32')
else:
endList = np.array([0], dtype='int32')
elif prefix == 'val':
valList = np.array([findEmptyVal(valDataType)] * valDim * numEls, \
dtype=valDataType).reshape((numEls, valDim) if valDim > 1 else numEls)
elif prefix == 'strand':
strandList = np.array([1] * numEls, dtype='int8')
elif prefix == 'id':
idList = np.array([''] * numEls, dtype='S1')
elif prefix == 'edges':
edgesList = np.array([['']] * numEls, dtype='S1')
elif prefix == 'weights':
weightsList = np.array([[[findEmptyVal(weightDataType)]]] * weightDim * numEls, \
dtype=weightDataType).reshape((numEls, 1, weightDim) if weightDim > 1 else (numEls, 1))
else:
extraLists[prefix] = np.array([''] * numEls, dtype='S1')
return TrackView(region, startList, endList, valList, strandList,
idList, edgesList, weightsList, borderHandling,
allowOverlaps, extraLists)
if self.PASS_ON_EXCEPTIONS:
raise
else:
self._printExceptionMsg(e, trackName, Error=False)
except Exception, e:
collector.removeEntry()
if self.PASS_ON_EXCEPTIONS:
raise
else:
self._printExceptionMsg(e, trackName, Error=True)
self._calcAndStoreSubTrackCount(trackName)
if self._raiseIfAnyWarnings and len(self._warningTrackNames) > 0:
raise Warning('Warnings occurred in the following tracks: ' + \
', '.join(prettyPrintTrackName(tn) for tn in self._warningTrackNames))
return atLeastOneFinalized
def _allTrackNames(self):
raise AbstractClassError
def _allGESourceManagers(self, trackName, allowOverlaps):
collector = PreProcMetaDataCollector(self._genome, trackName)
if allowOverlaps == False and collector.overlapRuleHasBeenFinalized(True):
for i in range(1):
self._status = 'Trying to prepare preprocessing for track "%s"' % ':'.join(trackName) + \
(' (allowOverlaps: %s)' % allowOverlaps)
yield self._getGESourceManagerFromTrack(trackName)
else:
for geSource in self._allGESources(trackName):
if allowOverlaps == True:
if self.PASS_ON_EXCEPTIONS:
raise
else:
self._printExceptionMsg(e, trackName, Error=False)
except Exception, e:
collector.removeEntry()
if self.PASS_ON_EXCEPTIONS:
raise
else:
self._printExceptionMsg(e, trackName, Error=True)
self._calcAndStoreSubTrackCount(trackName)
if self._raiseIfAnyWarnings and len(self._warningTrackNames) > 0:
raise Warning('Warnings occurred in the following tracks: ' + \
', '.join(prettyPrintTrackName(tn) for tn in self._warningTrackNames))
return atLeastOneFinalized
def _allTrackNames(self):
raise AbstractClassError
def _allGESourceManagers(self, trackName, allowOverlaps):
collector = PreProcMetaDataCollector(self._genome, trackName)
if allowOverlaps == False and collector.overlapRuleHasBeenFinalized(
True):
for i in range(1):
self._status = 'Trying to prepare preprocessing for track "%s"' % ':'.join(trackName) + \
(' (allowOverlaps: %s)' % allowOverlaps)
yield self._getGESourceManagerFromTrack(trackName)
else:
for geSource in self._allGESources(trackName):
