def setUp(self):
sampler = T2DSampler()
self.logger = Logger().getLogger(__name__)
self.simplePropertyMapper = SimplePropertyMapper()
self.testTable = sampler.getTestTable()
#self.testTables20 = sampler.get20Tables()
self.testTables = sampler.getTablesSubjectIdentification()
class SimplePropertyMappingTestCase(unittest.TestCase):
def setUp(self):
sampler = T2DSampler()
self.testTables = sampler.getTablesPropertyAnnotationDbpediaGoldStandard()
self.propertyMapper = SimplePropertyMapper()
def testMapProperties(self):
logging.disable(logging.DEBUG)
logging.disable(logging.INFO)
for num, table in enumerate(self.testTables):
propertyScores = self.propertyMapper.getScores(table)
propertiesRetrieved = self.propertiesRetrievedByTP(propertyScores, table.propertiesGold)
#(overall, correct) = self.diffProperties(properties, table.propertiesGold)
print "%s, %s" % (propertiesRetrieved, len(table.propertiesGold),)
def propertiesRetrievedByTP(self, propertyScores, propertiesGold):
count = 0
if propertyScores == []:
return count
for _property in propertiesGold:
propertyScore = propertyScores[_property['columnIndex']].get(_property['uri'])
if propertyScore != None:
count += 1
return count
def diffProperties(self, propertiesMapped, propertiesGold):
correct = 0
overall = len(propertiesMapped)
for propertyMapped in propertiesMapped:
#find property with the same columnIndex
for propertyGold in propertiesGold:
if propertyMapped['columnIndex'] == propertyGold['columnIndex']:
if propertyMapped['uri'] == propertyGold['uri']:
correct += 1
return (overall, correct)
def setUp(self):
sampler = T2DSampler()
self.testTables = sampler.getTablesPropertyAnnotationDbpediaGoldStandard()
self.propertyMapper = SimplePropertyMapper()
def getScores(self, table):
simplePropertyMapper = SimplePropertyMapper()
return simplePropertyMapper.getScores(table)
class SimpleCachePropertyMappingBenchTestCase(unittest.TestCase):
def setUp(self):
sampler = T2DSampler()
self.logger = Logger().getLogger(__name__)
self.simplePropertyMapper = SimplePropertyMapper()
self.testTable = sampler.getTestTable()
#self.testTables20 = sampler.get20Tables()
self.testTables = sampler.getTablesSubjectIdentification()
def determineResultsFilename(self, filename):
import os
while os.path.exists(os.path.join("results",filename)):
filename = filename.split(".")
index = str(int(filename.pop()) + 1)
filename.append(index)
filename = ".".join(filename)
filename = os.path.join("results",filename)
return filename
def resultsIterativePrinter(self, row, filename):
import csv
with open(filename, 'a') as csvfile:
spamwriter = csv.writer(csvfile, delimiter=',',
quotechar='"', quoting=csv.QUOTE_MINIMAL)
spamwriter.writerow(row)
def simplePropertyMapping(self, tables, rowsToDisambiguate=20, threshold=10):
resultsFilename = "%srows.%sthreshold.propertyMapping.results.csv.1" %(rowsToDisambiguate, threshold)
resultsFilename = self.determineResultsFilename(resultsFilename)
header = ["tableId","subjectColumn","subjectColumnCorrect","propertyIdentified","correctly","falsePositives","notIdentified","seedListContains","executionTimeFull","executionTimePure","disambiguationTime","classSearchTime","propertySearchTime"]
self.resultsIterativePrinter(header,resultsFilename)
for table in tables:
properties = self.simplePropertyMapper.mapProperties(table,rowsToDisambiguate, threshold)
subjectColumn = self.simplePropertyMapper.subjectColumn
subjectColumnCorrect = self.simplePropertyMapper.subjectColumnCorrect
executionTimeFull = self.simplePropertyMapper.executionTimeFull
executionTimePure = self.simplePropertyMapper.executionTimePure
disambiguationTime = self.simplePropertyMapper.disambiguationTime
classSearchTime = self.simplePropertyMapper.classSearchTime
propertySearchTime = self.simplePropertyMapper.propertySearchTime
seedListContains = self.simplePropertyMapper.seedListContains
falsePositives = 0
falseNegatives = table.getNumberOfProperties()
correct = 0
propertiesString = u""
lastItem = len(properties) - 1
for i, _property in enumerate(properties):
(uri, index) = _property
if table.isProperty(_property):
correct += 1
falseNegatives -= 1
else:
falsePositives += 1
if i == lastItem:
propertiesString += uri
else:
propertiesString += uri + u"|"
result = [table.id, subjectColumn, subjectColumnCorrect, propertiesString, correct, falsePositives,falseNegatives,seedListContains,executionTimeFull,executionTimePure,disambiguationTime,classSearchTime,propertySearchTime]
self.resultsIterativePrinter(result,resultsFilename)
def testMapProperties(self):
#tables = [self.testTable]
tables = self.testTables
self.simplePropertyMapping(tables)
