def __init__(self, geSource):
self._geSource = self._decorateGESource(geSource)
self._boundingRegionsAndGEsCorrespond = None
self._areValsCategorical = TrackFormat.createInstanceFromGeSource(geSource).getValTypeName() == 'Category'
self._areEdgeWeightsCategorical = TrackFormat.createInstanceFromGeSource(geSource).getWeightTypeName() == 'Category'
self._valCategories = set()
self._edgeWeightCategories = set()
self._numElements = OrderedDefaultDict(int)
self._maxStrLens = OrderedDefaultDict(partial(self._initMaxStrLens, self._getMaxStrLensKeys()))
self._maxNumEdges = OrderedDefaultDict(int)
self._hasCalculatedStats = False
def evaluatePvalueAndNullDistributionList(observedAndMcSamplesTuple, tail, rawStatisticMainClassName):
resultsDict = OrderedDict()
#TODO: What is received is not a list of tuples, it is a tuple of the real result which is a
# TrackStructure whose result is a list of raw values and list of such track structures.
# Need to find a way to handle it.
observedResult = observedAndMcSamplesTuple[0]
mcSamplesTsList = observedAndMcSamplesTuple[1]
#TODO: What about categorial ts results?
isPairedTsResult = all([val.isPairedTs() for val in observedResult.values()])
observedResultDict = OrderedDict()
mcSamplesResultDict = OrderedDefaultDict(list)
if isPairedTsResult:
for pairedTs in observedResult.values():
trackTitle = pairedTs['reference'].metadata['title']
assert trackTitle not in observedResultDict, "%s already in observed results dict" % trackTitle
observedResultDict[trackTitle] = pairedTs.result
for mcSampleTs in mcSamplesTsList:
for pairedTs in mcSampleTs.values():
trackTitle = pairedTs['reference'].metadata['title']
mcSamplesResultDict[trackTitle].append(pairedTs.result)
else: #isFlat?
raise Exception('not implemented yet!')
for trackTitle, observation in observedResultDict.iteritems():
resultsDict[trackTitle] = evaluatePvalueAndNullDistribution((observation, mcSamplesResultDict[trackTitle]), tail, rawStatisticMainClassName)
return resultsDict
def __init__(self, statistic=None):
self._statClassList = [statistic] if statistic else []
self._analysisParts = []
self._analysisOptionsDict = OrderedDefaultDict(list)