def computeDiversityCoefficient(numberNodesInTree,sampleNameList,dataArray):
#@sample is a list of (name of node,rank of node,sampleHitList)
sample,_,numberNodes = takeNodesInTree(dataArray[7],sampleNameList)
resultNodes = []
for node in sample:
assignments = 0
if not (len(node) == 3):
print " \n/!\ ERROR: List of samples wrong:",len(node),"."
raise ValueError
name,rank,sampleList = node[0],node[1],node[2]
for hit in sampleList:
if not (len(hit) == 2):
print "\n/!\ ERROR: Sample Hit List wrong:",len(hit),"."
raise ValueError
assignments += hit[1]
resultNodes.append(((name,rank),assignments))
if not numberNodes:
if not (len(sample)):
print "\n/!\ ERROR: [BUG] [percentage/computeSamplesInAllMetadatum] Empty set of samples."
raise ValueError
return 0,resultNodes
if not numberNodesInTree:
print "\n/!\ ERROR: Taxonomic Tree is empty: whole tree:",numberNodesInTree,"."
raise ValueError
return (numberNodes/numberNodesInTree),resultNodes
def compute(tree,sampleNameList1,sampleNameList2):
nodeList1,numberAT1,numberN1 = takeNodesInTree(tree,sampleNameList1)
nodeList2,numberAT2,numberN2 = takeNodesInTree(tree,sampleNameList2)
common, in1, in2 = [],[],[]
#@numberC is the number of nodes in common
numberC = 0
for node in nodeList1:
boolean,sampleHitList = memAndSampleHitList(node,nodeList2)
#If @sampleHitList is not empty
#That is if node also belongs to the second list of samples
if boolean:
common += [(node[0],node[1],node[2],sampleHitList)]
numberC += 1
else:
in1 += [node]
for node in nodeList1:
#@boolean answers: Is it a node that does not belong to the first list of samples?
boolean = True
for cNode in common:
if (node[0] == cNode[0] and node[1] == cNode[1]):
boolean = False
if boolean:
in2 += [node]
return common,in1,in2,numberAT1,numberAT2,numberN1,numberN2,numberC
def enumerateCommonPatterns(tree,sampleNameList1,sampleNameList2):
commonPatternsList = []
#@nodesList1 is the list of nodes (name,rank,sampleHitList) for sampleNameList1 (see misc.py)
nodesList1,_,_ = takeNodesInTree(tree,sampleNameList1)
#We consider every node of the tree as a potential root for a pattern
for node in nodesList1:
pattern = []
numberAssignments = 0
numberNodes = 0
name,rank = node[0],node[1]
#Gets the subtree (of the whole taxonomic tree) rooted at node
root = tree.search(name,rank)
#@candidateNodes is the list of TaxoTree nodes that can potentially be added to the pattern
candidateNodes = [root]
while candidateNodes:
child = candidateNodes.pop()
sampleHitList = child.sampleHitList
#Checking if child has been assigned in both samples
isInSampleList1 = []
isInSampleList2 = []
for x in sampleHitList:
if inSample(x,sampleNameList1):
isInSampleList1.append(x)
if inSample(x,sampleNameList2):
isInSampleList2.append(x)
#If both lists are not empty, then node has been assigned in both samples
if isInSampleList1 and isInSampleList2:
#Merge the elements of both lists, deleting duplicates
#e.g. if OPNA-J90 belongs to both sampleNameLists it would corrupt the result
#as the assignments in this sample to child would be duplicated
#(assuming there is no duplicate in each list)
isInSample = mergeList(isInSampleList1,isInSampleList2)
pattern.append((child.name,child.rank))
numberNodes += 1
for x in isInSample:
numberAssignments += x[1]
candidateNodes += child.children
#if the pattern is non-empty
if pattern:
commonPatternsList.append((pattern,numberAssignments,numberNodes))
return commonPatternsList
def enumerateSpecificPatterns(tree,sampleNameListPattern,sampleNameListOther):
specificPatternsList = []
#List from where samples in both lists are deleted and only elements from sampleNameListPattern remain
sampleNameListPatternTrimmed = trimList(sampleNameListPattern,sampleNameListOther)
#@nodesList is the list of nodes (name,rank,sampleHitList) for sampleNameListPatternTrimmed
nodesList,_,_ = takeNodesInTree(tree,sampleNameListPatternTrimmed)
#Pretty much the same procedure than for @enumerateCommonPatterns
for node in nodesList:
pattern = []
numberAssignments = 0
numberNodes = 0
name,rank = node[0],node[1]
root = tree.search(name,rank)
candidateNodes = [root]
while candidateNodes:
child = candidateNodes.pop()
sampleHitList = child.sampleHitList
isInSampleListPattern = []
for x in sampleHitList:
if inSample(x,sampleNameListPatternTrimmed):
isInSampleListPattern.append(x)
if inSample(x,sampleNameListOther):
#This node is assigned in the samples of sampleNameListOther
#so it is discarded from the pattern
isInSamplePattern = []
break
if isInSampleListPattern:
pattern.append((child.name,child.rank))
numberNodes += 1
for x in isInSampleListPattern:
numberAssignments += x[1]
candidateNodes += child.children
#if the pattern is non-empty
if pattern:
specificPatternsList.append((pattern,numberAssignments,numberNodes))
return specificPatternsList
