def get_issues():
n_issues = 0
issues_by_case = defaultdict(list)
meta_file = util_sc.init_metadata(data_dir)
case_file_by_year = util_sc.get_case_from_metafile(meta_file, data_dir)
for year in case_file_by_year:
current_cases = case_file_by_year[year]
#print current_cases
#print len(current_cases)
for case in current_cases:
case_file_path = year + '/' + str(case) + '.xml'
abs_case_file_path = data_dir + '/' + case_file_path
tree = et.ElementTree(file=abs_case_file_path)
case_issues = tree.find("issues").text.split(',')
for i in xrange(0, len(case_issues)):
issue = case_issues[i].strip().lower()
# print "case issues %s" % issue.strip()
if issue not in issue_id:
issue_id[issue] = n_issues
reverse_issue_id[n_issues] = issue
n_issues += 1
issues_by_case[case].append(issue_id[issue])
#print len(issues_by_case)
#print n_issues
return issues_by_case, n_issues
def build_dataset(n_tokens, n_topics):
meta_file = util_sc.init_metadata(data_dir)
case_file_by_year = util_sc.get_case_from_metafile(meta_file, data_dir)