@@ -1,17 +1,20 @@ from content_analyst_api import extractKeyWords -from twitter_api import getTwits +from twitter_api import getTwits, Twit +from twitter_api2 import query_twitter from flickr_api import getFlickers def dig_web(): - twits = getTwits() + #twits = getTwits() + twits = query_twitter('brhackday') + #Twit(set(["brhackday"])) photos = getFlickers() words = [] - for twit in twits: - twit.content_tags = extractKeyWords(twit) - words.append(twit) - #for photo in photos: - # photo.content_tags = photo.tags - # words.append(photo) + #for twit in twits: + # twit.content_tags = extractKeyWords(twit) + # words.append(twit) + for photo in photos: + photo.content_tags = photo.tags + words.append(photo) return words if __name__ == "__main__": analyser.py @@ -35,7 +35,6 @@ class Cluster: def create_clusters(contextables, clusters, tags): - print "Start creating new clusters" new_clusters = [] for cluster in clusters: for tag in tags: @@ -43,20 +42,23 @@ def create_clusters(contextables, clusters, tags): tag_cluster = CLUSTERS[frozenset([tag])] new_cluster = Cluster(cluster.tags.union(tag_cluster.tags)) new_cluster.populate(cluster.contextables.union(tag_cluster.contextables)) - new_clusters.append(new_cluster) + if new_cluster.contextables: + new_clusters.append(new_cluster) return list(sorted(new_clusters, key=lambda i: len(i))) def merge_clusters(contextables, number=5): - print "Merging" - tags = order_tags(contextables)[:160] + tags = order_tags(contextables)[:260] new_clusters = [] - print "First clusters" for tag in tags: cluster = Cluster(frozenset([tag])) cluster.populate(contextables) new_clusters.append(cluster) for i in range(number-1): - new_clusters = create_clusters(contextables, new_clusters, tags)[:160/(i+2)] + nclusters = create_clusters(contextables, new_clusters, tags)[:260/(i+2)] + if nclusters: + new_clusters = nclusters + else: + return new_clusters return new_clusters def order_clusters(clusters): clusters.py @@ -37,7 +37,7 @@ def getFlickers(): tags = 0 photos = [] flickr = authFlickr() - recent = flickr.photos_getrecent(user_id=userid, per_page='500',format='json',extras='tags') + recent = flickr.photos_search(tags="brhackday08", per_page='500',format='json',extras='tags') recent_json = get_json(recent) #print "Down: "+str(len(recent_json['photos']['photo'])) for photo in recent_json['photos']['photo']: @@ -56,4 +56,4 @@ def getFlickers(): if __name__=='__main__': while True: - getFlickers() + print len(getFlickers()) flickr_api.py @@ -1,4 +1,4 @@ from analyser import dig_web from clusters import merge_clusters -mc = merge_clusters(dig_web(), 5) +mc = merge_clusters(dig_web(), 2) main.py @@ -5,7 +5,6 @@ url = 'http://search.twitter.com/search.json?q=%s' def query_twitter(term): s = urlopen(url % (term)).read() dic = eval(s.replace('null', 'None')) - #print dic return dic if __name__ == '__main__': twitter_api2.py 0ae8441adc64fcd585ca943b7612b1bf881188eb Bruno Gola bgola@pedal.(none) http://github.com/Fabs/eventomeeter/commit/fc6261f154cc804bb2228b9a2c3c22a2d2d67da9 fc6261f154cc804bb2228b9a2c3c22a2d2d67da9 2008-11-09T06:20:43-08:00 2008-11-09T06:20:43-08:00 ... 55ea1cdc2662116708a280bf053e87934d631a01 Bruno Gola bgola@pedal.(none)