Skip to content
Snippets Groups Projects
Commit 95810d7c authored by bbauvin's avatar bbauvin
Browse files

Trying to fix segmentation fault

parent 59e9af6c
Branches
No related tags found
No related merge requests found
Show whitespace changes
Inline Side-by-side
Code/MonoMutliViewClassifiers/Monoview/ExecClassifMonoView.py
+ 0
− 1
View file @ 95810d7c
......@@ -58,7 +58,6 @@ def ExecMonoview(X, Y, name, learningRate, nbFolds, nbCores, databaseType, path,
CL_type = kwargs["CL_type"]
nbClass = kwargs["nbClass"]
X = getValue(X)
print X.data.nbytes + X.indptr.nbytes + X.indices.nbytes
datasetLength = X.shape[0]
clKWARGS = kwargs[kwargs["CL_type"]+"KWARGS"]
......
Code/MonoMutliViewClassifiers/Multiview/GetMultiviewDb.py
+ 31
− 26
View file @ 95810d7c
......@@ -403,23 +403,24 @@ def makeSparseTotalMatrix(sortedRNASeq):
return sparseFull
def getAdjacenceMatrix(sortedRNASeq, k=1):
indices = np.zeros((sortedRNASeq.shape[0]*k*sortedRNASeq.shape[1]), dtype=np.int32)
data = np.ones((sortedRNASeq.shape[0]*k*sortedRNASeq.shape[1]), dtype=bool)
indptr = np.zeros(sortedRNASeq.shape[0]+1, dtype=np.int16)
nbGenes = sortedRNASeq.shape[1]
def getAdjacenceMatrix(RNASeqRanking, sotredRNASeq, k=2):
k=int(k)/2*2
indices = np.zeros((RNASeqRanking.shape[0]*k*RNASeqRanking.shape[1]), dtype=np.int32)
data = np.ones((RNASeqRanking.shape[0]*k*RNASeqRanking.shape[1]), dtype=bool)
indptr = np.zeros(RNASeqRanking.shape[0]+1, dtype=np.int16)
nbGenes = RNASeqRanking.shape[1]
pointer = 0
for patientIndex, patient in enumerate(sortedRNASeq):
for patientIndex in range(RNASeqRanking.shape[0]):
print patientIndex
for i in range(nbGenes):
for j in range(k):
for j in range(k/2):
try:
indices[pointer]=patient[(i-(j+1))]+patient[i]*nbGenes
indices[pointer]=RNASeqRanking[patientIndex, (sotredRNASeq[patientIndex, i]-(j+1))]+i*nbGenes
pointer+=1
except:
pass
try:
indices[pointer]=patient[i+(j+1)]+patient[i]*nbGenes
indices[pointer]=RNASeqRanking[patientIndex, (sotredRNASeq[patientIndex, i]+(j+1))]+i*nbGenes
pointer+=1
except:
pass
......@@ -431,7 +432,7 @@ def getAdjacenceMatrix(sortedRNASeq, k=1):
# data.append(True)
indptr[patientIndex+1] = pointer
mat = sparse.csr_matrix((data, indices, indptr), shape=(sortedRNASeq.shape[0], sortedRNASeq.shape[1]*sortedRNASeq.shape[1]), dtype=bool)
mat = sparse.csr_matrix((data, indices, indptr), shape=(RNASeqRanking.shape[0], RNASeqRanking.shape[1]*RNASeqRanking.shape[1]), dtype=bool)
return mat
......@@ -478,12 +479,16 @@ def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES):
logging.debug("Start:\t Getting Sorted RNASeq Data")
RNASeq = datasetFile["View2"][...]
modifiedRNASeq = np.zeros(datasetFile.get("View2").shape, dtype=int)
sortedRNASeqGeneIndices = np.zeros(datasetFile.get("View2").shape, dtype=int)
RNASeqRanking = np.zeros(datasetFile.get("View2").shape, dtype=int)
for exampleIndex, exampleArray in enumerate(RNASeq):
RNASeqDictionary = dict((index, value) for index, value in enumerate(exampleArray))
sorted_x = sorted(RNASeqDictionary.items(), key=operator.itemgetter(1))
modifiedRNASeq[exampleIndex] = np.array([index for (index, value) in sorted_x], dtype=int)
mrnaseqDset = datasetFile.create_dataset("View4", modifiedRNASeq.shape, data=modifiedRNASeq)
sortedRNASeqDictionary = dict((index, value) for index, value in enumerate(exampleArray))
sortedRNASeqIndicesDict = sorted(sortedRNASeqDictionary.items(), key=operator.itemgetter(1))
sortedRNASeqIndicesArray = np.array([index for (index, value) in sortedRNASeqIndicesDict], dtype=int)
sortedRNASeqGeneIndices[exampleIndex] = sortedRNASeqIndicesArray
for geneIndex in range(RNASeq.shape[1]):
RNASeqRanking[exampleIndex, sortedRNASeqIndicesArray[geneIndex]] = geneIndex
mrnaseqDset = datasetFile.create_dataset("View4", sortedRNASeqGeneIndices.shape, data=sortedRNASeqGeneIndices)
mrnaseqDset.attrs["name"] = "SRNASeq"
mrnaseqDset.attrs["sparse"] = False
logging.debug("Done:\t Getting Sorted RNASeq Data")
......@@ -496,9 +501,9 @@ def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES):
factorizedLeftBaseMatrix = np.genfromtxt(path+"factorLeft--n-"+str(datasetFile.get("View2").shape[1])+"--k-"+str(100)+".csv", delimiter=',')
except:
factorizedSupBaseMatrix, factorizedLeftBaseMatrix = getBaseMatrices(rnaseqData.shape[1], k)
brnaseqDset = datasetFile.create_dataset("View5", (modifiedRNASeq.shape[0], modifiedRNASeq.shape[1]*k*2), dtype=bool)
for patientIndex, patientSortedArray in enumerate(modifiedRNASeq):
patientMatrix = np.zeros((modifiedRNASeq.shape[1], k * 2), dtype=bool)
brnaseqDset = datasetFile.create_dataset("View5", (sortedRNASeqGeneIndices.shape[0], sortedRNASeqGeneIndices.shape[1]*k*2), dtype=bool)
for patientIndex, patientSortedArray in enumerate(sortedRNASeqGeneIndices):
patientMatrix = np.zeros((sortedRNASeqGeneIndices.shape[1], k * 2), dtype=bool)
for lineIndex, geneIndex in enumerate(patientSortedArray):
patientMatrix[geneIndex]= np.concatenate((factorizedLeftBaseMatrix[lineIndex,:], factorizedSupBaseMatrix[:, lineIndex]))
brnaseqDset[patientIndex] = patientMatrix.flatten()
......@@ -507,7 +512,7 @@ def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES):
logging.debug("Done:\t Getting Binarized RNASeq Data")
# logging.debug("Start:\t Getting Binned RNASeq Data")
# sparseBinnedRNASeq = makeSparseTotalMatrix(modifiedRNASeq)
# sparseBinnedRNASeq = makeSparseTotalMatrix(sortedRNASeqGeneIndices)
# sparseBinnedRNASeqGrp = datasetFile.create_group("View6")
# dataDset = sparseBinnedRNASeqGrp.create_dataset("data", sparseBinnedRNASeq.data.shape, data=sparseBinnedRNASeq.data)
# indicesDset = sparseBinnedRNASeqGrp.create_dataset("indices", sparseBinnedRNASeq.indices.shape, data=sparseBinnedRNASeq.indices)
......@@ -518,7 +523,7 @@ def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES):
# logging.debug("Done:\t Getting Binned RNASeq Data")
logging.debug("Start:\t Getting Adjacence RNASeq Data")
sparseAdjRNASeq = getAdjacenceMatrix(modifiedRNASeq, k=findClosestPowerOfTwo(100)-1)
sparseAdjRNASeq = getAdjacenceMatrix(RNASeqRanking, sortedRNASeqGeneIndices, k=findClosestPowerOfTwo(100)-1)
sparseAdjRNASeqGrp = datasetFile.create_group("View6")
dataDset = sparseAdjRNASeqGrp.create_dataset("data", sparseAdjRNASeq.data.shape, data=sparseAdjRNASeq.data)
indicesDset = sparseAdjRNASeqGrp.create_dataset("indices", sparseAdjRNASeq.indices.shape, data=sparseAdjRNASeq.indices)
......
Code/MonoMutliViewClassifiers/utils/Dataset.py
+ 1
− 8
View file @ 95810d7c
......@@ -12,8 +12,6 @@ def getV(DATASET, viewIndex, usedIndices=None):
DATASET.get("View"+str(viewIndex)).get("indices").value,
DATASET.get("View"+str(viewIndex)).get("indptr").value),
shape=DATASET.get("View"+str(viewIndex)).attrs["shape"])[usedIndices,:]
print sparse_mat.shape
print sparse_mat.indptr
return sparse_mat
......@@ -32,24 +30,19 @@ def getValue(DATASET):
DATASET.get("indices").value,
DATASET.get("indptr").value),
shape=DATASET.attrs["shape"])
print sparse_mat.shape
print sparse_mat.indptr
return sparse_mat
def extractSubset(matrix, usedIndices):
if sparse.issparse(matrix):
newIndptr = np.zeros(len(usedIndices)+1, dtype=np.int16)
oldindptr = matrix.indptr
print oldindptr
for exampleIndexIndex, exampleIndex in enumerate(usedIndices):
newIndptr[exampleIndexIndex+1] = newIndptr[exampleIndexIndex]+(oldindptr[exampleIndex+1]-oldindptr[exampleIndex])
newData = np.ones(newIndptr[-1], dtype=bool)
newIndices = np.zeros(newIndptr[-1], dtype=np.int32)
oldIndices = matrix.indices
print newIndptr
for exampleIndexIndex, exampleIndex in enumerate(usedIndices):
print newIndptr[exampleIndexIndex], newIndptr[exampleIndexIndex+1]
newIndices[newIndptr[exampleIndexIndex]:newIndptr[exampleIndexIndex+1]] = oldIndices[oldindptr[exampleIndex]: oldindptr[exampleIndex+1]]
return sparse.csr_matrix((newData, newIndices, newIndptr), shape=(len(usedIndices), matrix.shape))
return sparse.csr_matrix((newData, newIndices, newIndptr), shape=(len(usedIndices), matrix.shape[1]))
else:
return matrix[usedIndices]
\ No newline at end of file
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please to comment