Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ft_master/cherrypick_optimization #68

Open
wants to merge 2 commits into
base: hb_master
Choose a base branch
from
+28 −14
Open

ft_master/cherrypick_optimization #68

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
63c3f13
CHERRYPICK to master: Optimize raw overlap algorithm
borissim Apr 13, 2018
023ad59
CHERRYPICK to master: Optimize shuffling of numpy arrays
borissim Apr 13, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 16 additions & 9 deletions lib/hb/gold/statistic/RawOverlapStat.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -67,7 +67,7 @@ def _findAllStartAndEndEvents(t1s, t1e, t2s, t2e):
t2CodedEnds= t2e * 8 +2

allSortedCodedEvents = numpy.concatenate( (t1CodedStarts,t1CodedEnds,t2CodedStarts,t2CodedEnds) )
allSortedCodedEvents.sort()
allSortedCodedEvents.sort(kind='mergesort')

allEventCodes = (allSortedCodedEvents % 8) -4

Expand All @@ -83,14 +83,21 @@ def _findAllStartAndEndEvents(t1s, t1e, t2s, t2e):

@classmethod
def _computeRawOverlap(cls, t1s, t1e, t2s, t2e, binSize):
allSortedDecodedEvents, allEventLengths, cumulativeCoverStatus = cls._findAllStartAndEndEvents(t1s, t1e, t2s, t2e)

tn,fp,fn,tp = [long((allEventLengths[ cumulativeCoverStatus[:-1] == status ]).sum()) for status in range(4)]

if len(allSortedDecodedEvents)>0:
tn += allSortedDecodedEvents[0] + (binSize - allSortedDecodedEvents[-1])
else:
tn+=binSize
# allSortedDecodedEvents, allEventLengths, cumulativeCoverStatus = cls._findAllStartAndEndEvents(t1s, t1e, t2s, t2e)
#
# tn,fp,fn,tp = [long((allEventLengths[ cumulativeCoverStatus[:-1] == status ]).sum()) for status in range(4)]
#
# if len(allSortedDecodedEvents)>0:
# tn += allSortedDecodedEvents[0] + (binSize - allSortedDecodedEvents[-1])
# else:
# tn+=binSize

starts = numpy.sort(numpy.concatenate((t1s, t2s)), kind="mergesort")
ends = numpy.sort(numpy.concatenate((t1e, t2e)), kind="mergesort")
tp = long(numpy.sum(numpy.maximum(ends[:-1] - starts[1:], 0)))
fp = long(numpy.sum(t1e - t1s)) - tp
fn = long(numpy.sum(t2e - t2s)) - tp
tn = binSize - tp - fp - fn

return tn,fp,fn,tp

Expand Down
8 changes: 6 additions & 2 deletions lib/hb/gold/track/RandomizedSegsTrack.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,9 @@ def _permuteSegs(self, starts, ends, vals, strands, ids, edges, weights, extras)
segLens = ends-starts
#permuting order (of length-elements) of both pools
if vals is None and strands is None:
numpy.random.shuffle(segLens)
# numpy.random.shuffle(segLens)
permutIndexes = numpy.random.permutation(len(segLens))
segLens = segLens[permutIndexes]
else:
permutIndexes = numpy.random.permutation( len(segLens) )
segLens = segLens[permutIndexes]
Expand All @@ -61,7 +63,9 @@ def _permuteIntersegs(self, starts, ends, binLen):
#add start and end-case of bin. Double-check with statistician..
intersegLens = numpy.append(intersegLens, [starts[0], binLen-ends[-1]])

numpy.random.shuffle(intersegLens)
# numpy.random.shuffle(intersegLens)
permutIndexes = numpy.random.permutation(len(intersegLens))
intersegLens = intersegLens[permutIndexes]
return intersegLens

def _sampleIntervals(self, totalSpace, numElements):
Expand Down
3 changes: 2 additions & 1 deletion lib/hb/gold/track/ShuffledMarksTrack.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@ def _checkTrackFormat(self, origTV):

def _createRandomizedNumpyArrays(self, binLen, starts, ends, vals, strands, ids, edges, weights, extras, region):
newVals = numpy.copy(vals)
numpy.random.shuffle(newVals)
permutIndexes = numpy.random.permutation(len(newVals))
newVals = newVals[permutIndexes]

return starts, ends, newVals, strands, ids, edges, weights, extras
3 changes: 2 additions & 1 deletion lib/hb/quick/extra/StandardizeTrackFiles.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -730,7 +730,8 @@ def parseFile(cls,inFn, outFn, numElements=None, **kwargs):

import numpy.random as nr
lineNumbers = range(len(lines))
nr.shuffle(lineNumbers)
permutIndexes = nr.permutation(len(lineNumbers))
lineNumbers = lineNumbers[permutIndexes]
lineNumbers = lineNumbers[0:numElements]

for lineNum in lineNumbers:
Expand Down
3 changes: 2 additions & 1 deletion lib/hb/quick/track/SegsSampledByDistanceToReferenceTrack.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,8 @@ def _createRandomizedNumpyArraysFromDistanceToReference(self, binLen, starts, en

assert len(sampledPositions) == numElements
sampledElementLengths = elementLengths
numpy.random.shuffle(sampledElementLengths)
permutIndexes = numpy.random.permutation(len(sampledElementLengths))
sampledElementLengths = sampledElementLengths[permutIndexes]
sampledPositions = numpy.array(sampledPositions)
sampledPositions.sort()
sampledStarts = (sampledPositions - (sampledElementLengths/2)).astype('int')
Expand Down