def detectEvents(limit=300,
similarityType=MED_SIM,
minimalTermPerTweet=MIN_TERM_OCCURENCE,
remove_noise_with_poisson_Law=REMOVE_NOISE_WITH_POISSON_LAW,
printEvents=True):
mongoDBHandler = MongoDBHandler()
tweets = mongoDBHandler.getAllTweets(limit=limit)
if similarityType == LED_SIM:
s = LEDSimilarityMatrixBuilder(timeThreshold=TIME_RESOLUTION,
distanceThreshold=DISTANCE_RESOLUTION)
eventDetector = EventDetector(tweets, s)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
elif similarityType == MED_SIM:
s = MEDSimilarityMatrixBuilder(timeResolution=TIME_RESOLUTION,
distanceResolution=DISTANCE_RESOLUTION,
scaleNumber=SCALE_NUMBER,
minSimilarity=MIN_SIMILARITY)
eventDetector = EventDetector(tweets, s)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
else:
eventDetector = OptimisedEventDetectorMEDBased(
tweets,
timeResolution=TIME_RESOLUTION,
distanceResolution=DISTANCE_RESOLUTION,
scaleNumber=SCALE_NUMBER,
minSimilarity=MIN_SIMILARITY)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
print ""
print "-" * 40
print "{0} Event detected : ".format(len(events))
print "-" * 40
if printEvents:
eventDetector.showTopEvents(top=10)
return events
def detectEvents(limit=300,
similarityType=MED_SIM,
minimalTermPerTweet=MIN_TERM_OCCURENCE,
remove_noise_with_poisson_Law=REMOVE_NOISE_WITH_POISSON_LAW,
printEvents=True,
printInFile="events.txt",
useOnlyHashtags=False,
mongoDBName='Twitter',
mongoCollectionName="tweets"):
staringTime = time.time()
mongoDBHandler = MongoDBHandler(database_name=mongoDBName,
collection_name=mongoCollectionName)
tweets = mongoDBHandler.getAllTweets(limit=limit)
if similarityType == LED_SIM:
s = LEDSimilarityMatrixBuilder(timeThreshold=TIME_RESOLUTION,
distanceThreshold=DISTANCE_RESOLUTION,
useOnlyHashtags=useOnlyHashtags)
eventDetector = EventDetector(tweets, s)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
elif similarityType == MED_SIM:
s = MEDSimilarityMatrixBuilder(timeResolution=TIME_RESOLUTION,
distanceResolution=DISTANCE_RESOLUTION,
scaleNumber=SCALE_NUMBER,
minSimilarity=MIN_SIMILARITY,
useOnlyHashtags=useOnlyHashtags)
eventDetector = EventDetector(tweets, s)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
else:
eventDetector = OptimisedEventDetectorMEDBased(
tweets,
timeResolution=TIME_RESOLUTION,
distanceResolution=DISTANCE_RESOLUTION,
scaleNumber=SCALE_NUMBER,
minSimilarity=MIN_SIMILARITY,
useOnlyHashtags=useOnlyHashtags)
events = eventDetector.getEvents(
minimalTermPerTweet=minimalTermPerTweet,
remove_noise_with_poisson_Law=remove_noise_with_poisson_Law)
print ""
print "-" * 40
print "{0} Event detected : ".format(len(events))
print "-" * 40
if printEvents:
eventDetector.showTopEvents(top=len(events))
if printInFile:
txtFile = open(printInFile, 'w')
elapsed_time = (time.time() - staringTime)
txtFile.write("Total elapsed time : {0}s\n".format(elapsed_time))
txtFile.write("-" * 40 + "\n")
SEPARATOR = "\t|"
HEADER = "|" + SEPARATOR.join([
"Median time", "estimated duration (s)", "mean latitude",
"mean longitude", "radius (m)", "user number", "tweets number",
"top hashtags"
]) + SEPARATOR + "\n"
txtFile.write(HEADER)
txtFile.write("-" * 40 + "\n")
for event in events:
line = eventDetector.getStringOfEvent(event)
txtFile.write(line + "\n")
txtFile.write("-" * 40 + "\n")
txtFile.close()
return events
