def buildNetwork(queryFile, skosFile, featMan):
catQueryDist = findCatQueryDist(queryFile, featMan)
total = set(catQueryDist.keys())
#found = set()
#i = 0
#for line in open(argv[2],'r'):
print len(catQueryDist)
#print skosFile
if 'owl' in skosFile:
found, related, broad = loadOntology(skosFile, catQueryDist)
print len(found), len(broad)
else:
catQueryDist['dow_jones_transportation_average'] = set()
found, related, broad = loadSkosCategories(skosFile, catQueryDist)
notFound = total.difference(found)
categoryNetwork = nx.DiGraph()
#B parent of A
for entry in broad:
categoryNetwork.add_edge(entry[1], entry[0])
print len(notFound), len(related), len(broad)
return categoryNetwork, catQueryDist
task_label_dict,
samePairsSet,\
differentPairsSet, args.outDir)
mergeMetrics(total_metrics_dict, metrics)
else:
#weightMatrix = readWeightMatrix(args.distFile)
if args.algo == 'kmeans':
outSuff = args.outDir+'/kmeans_'
metrics = clusterAllWithKMeans(args.lowerLimit, args.upperLimit,\
featMan, weightMatrix, \
samePairsSet,
differentPairsSet,\
args.outDir)
mergeMetrics(total_metrics_dict, metrics)
elif args.algo == 'cat_kmeans':
catQueryDist = findCatQueryDist(args.featFile,featMan)
outSuff = args.outDir+'/cat_kmeans_'
metrics = clusterCatWithKMeans(args.lowerLimit, args.upperLimit,\
featMan, weightMatrix, samePairsSet,\
differentPairsSet, catQueryDist, \
outSuff)
mergeMetrics(total_metrics_dict, metrics)
elif args.algo == 'cat_kmediods':
catQueryDist = findCatQueryDist(args.featFile,featMan)
outSuff = args.outDir+'/cat_kmeds_'
metrics = clusterCatWithMediods(args.lowerLimit, args.upperLimit,\
featMan, weightMatrix, samePairsSet,\
differentPairsSet,catQueryDist,\
outSuff)
mergeMetrics(total_metrics_dict, metrics)
elif args.algo == 'cat_merge':