def add_collection_novelty(): #based on title
all_paper_titles = {} #average novelty is
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(
json.load(open('citeulike_collection/users/' + user + '.json')))
for paper in stat['papers']:
title = paper['title']
if not title in all_paper_titles:
all_paper_titles[title] = 1
else:
all_paper_titles[title] += 1
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(
json.load(open('citeulike_collection/users/' + user + '.json')))
paper_novelties = []
for paper in stat['papers']:
title = paper['title']
paper_novelties += [1.0 / float(all_paper_titles[paper['title']])]
nov = 0
if len(paper_novelties) > 0:
nov = numpy.mean(paper_novelties)
by_user[user]['collection_novelty'] = nov
print nov, paper_novelties
def add_collection_novelty():#based on title
all_paper_titles = {}#average novelty is
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(json.load(open('citeulike_collection/users/'+user+'.json')))
for paper in stat['papers']:
title = paper['title'],paper
def add_collection_novelty(): #based on title
all_paper_titles = {} #average novelty is
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(
json.load(open('citeulike_collection/users/' + user + '.json')))
for paper in stat['papers']:
title = paper['title'], paper
def generate_collected_csv():
for user in any_user:
# print "w user",user
stat = download_and_save_users.citulike_user_object_to_stats(json.load(open('citeulike_collection/users/'+user+'.json')))
# print stat
out_row = [user[:-1]]
for set in ['rank','log','web','ice']:
if any_user[user][set]:
out_row.append(set)
out_row.append(stat['collected'])
csv_out.writerow(out_row)
def generate_collected_csv():
for user in any_user:
# print "w user",user
stat = download_and_save_users.citulike_user_object_to_stats(
json.load(open('citeulike_collection/users/' + user + '.json')))
# print stat
out_row = [user[:-1]]
for set in ['rank', 'log', 'web', 'ice']:
if any_user[user][set]:
out_row.append(set)
out_row.append(stat['collected'])
csv_out.writerow(out_row)
def add_collection_novelty():#based on title
all_paper_titles = {}#average novelty is
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(json.load(open('citeulike_collection/users/'+user+'.json')))
for paper in stat['papers']:
title = paper['title']
if not title in all_paper_titles:
all_paper_titles[title] = 1
else:
all_paper_titles[title] += 1
for user in any_user:
print user
stat = download_and_save_users.citulike_user_object_to_stats(json.load(open('citeulike_collection/users/'+user+'.json')))
paper_novelties = []
for paper in stat['papers']:
title = paper['title']
paper_novelties += [1.0/ float(all_paper_titles[paper['title']])]
nov = 0;
if len(paper_novelties) > 0:
nov = numpy.mean(paper_novelties)
by_user[user]['collection_novelty'] = nov
print nov,paper_novelties
def compute_collection_metrics():
for user in any_user:
print "Computing collected for...", user
stat = download_and_save_users.citulike_user_object_to_stats(
json.load(open('citeulike_collection/users/' + user + '.json')))
out_row = [user[:-1]] #[user]#[user[:-1]]
for set in ['rank', 'log', 'web', 'ice']:
if any_user[user][set]:
out_row.append(set)
out_row.append(stat['collected'])
out_row.append(user[len(user) - 1])
#out_row.append()
keywords = {}
for paper in stat['papers']:
if "keywords" in paper:
print "paper keywords", paper['keywords']
for keyword in paper['keywords']:
keywords[keyword] = 1
by_user[user]['keyword_variety'] = len(keywords.keys())
print "out row", out_row
for i in range(len(collection_metrics)):
by_user[user][collection_metrics[i]] = out_row[i]
collected_depths = []
for paper in stat['papers']:
print paper['title'] #gbgbgbgbgb
if paper['title'] in paper_title_to_depth:
print paper_title_to_depth[paper['title']]
collected_depths += [paper_title_to_depth[paper['title']]]
else:
title = "nonenone"
if ":" in paper['title']:
title = paper['title'].split(":")[0]
if title in paper_title_to_depth:
print paper_title_to_depth[title]
collected_depths += [paper_title_to_depth[title]]
else:
print "???"
if len(collected_depths) > 0:
by_user[user]['collected_depth'] = numpy.mean(collected_depths)
else:
by_user[user]['collected_depth'] = 0
def compute_collection_metrics():
for user in any_user:
print "Computing collected for...",user
stat = download_and_save_users.citulike_user_object_to_stats(json.load(open('citeulike_collection/users/'+user+'.json')))
out_row = [user[:-1]]#[user]#[user[:-1]]
for set in ['rank','log','web','ice']:
if any_user[user][set]:
out_row.append(set)
out_row.append(stat['collected'])
out_row.append(user[len(user)-1])
#out_row.append()
keywords = {}
for paper in stat['papers']:
if "keywords" in paper:
print "paper keywords",paper['keywords']
for keyword in paper['keywords']:
keywords[keyword] = 1
by_user[user]['keyword_variety'] = len(keywords.keys())
print "out row",out_row
for i in range(len(collection_metrics)):
by_user[user][collection_metrics[i]] = out_row[i]
collected_depths = []
for paper in stat['papers']:
print paper['title']#gbgbgbgbgb
if paper['title'] in paper_title_to_depth:
print paper_title_to_depth[paper['title']]
collected_depths += [paper_title_to_depth[paper['title']]]
else:
title = "nonenone"
if ":" in paper['title']:
title = paper['title'].split(":")[0]
if title in paper_title_to_depth:
print paper_title_to_depth[title]
collected_depths += [paper_title_to_depth[title]]
else:
print "???"
if len(collected_depths) > 0:
by_user[user]['collected_depth'] = numpy.mean( collected_depths )
else:
by_user[user]['collected_depth'] = 0
