def post(self):
query_string = self.request.body_arguments.get('query')
query = TwitterSearchOrder()
query.set_keywords(query_string)
query.set_language('en')
query.set_include_entities(False)
results = TwitterClient.search_tweets(query)
tweets = [tweet['text'] for tweet in results['content']['statuses']]
tweet_results = indicoio.batch_text_tags(tweets)
n_tweets = float(len(tweet_results))
scores = defaultdict(float)
for tweet in tweet_results:
for category, score in tweet.items():
scores[category] += score / n_tweets
category = max(scores, key=lambda x: scores[x])
data = {
'scores': scores,
'category': category
}
self.write(json.dumps(data))
def main():
entries = feedparser.parse(feed)['entries']
titles = [entry.get('title') for entry in entries]
title_tags = batch_text_tags(titles)
for entry, tags in zip(entries, title_tags):
entry['tags'] = tags
entries = [parsed(entry) for entry in entries]
# render template with additional jinja2 data
return render_template('main.html', entries=entries)
def post(self):
query_string = self.request.body_arguments.get('query')
query = TwitterSearchOrder()
query.set_keywords(query_string)
query.set_language('en')
query.set_include_entities(False)
results = TwitterClient.search_tweets(query)
tweets = [tweet['text'] for tweet in results['content']['statuses']]
tweet_results = indicoio.batch_text_tags(tweets)
n_tweets = float(len(tweet_results))
scores = defaultdict(float)
for tweet in tweet_results:
for category, score in tweet.items():
scores[category] += score / n_tweets
category = max(scores, key=lambda x: scores[x])
data = {'scores': scores, 'category': category}
self.write(json.dumps(data))
def test_batch_texttags(self):
test_data = ["On Monday, president Barack Obama will be..."]
response = batch_text_tags(test_data, api_key=self.api_key)
self.assertTrue(isinstance(response, list))
def test_batch_texttags(self):
test_data = ["On Monday, president Barack Obama will be..."]
response = batch_text_tags(test_data, api_key=self.api_key)
self.assertTrue(isinstance(response, list))
from sklearn.datasets import fetch_20newsgroups import indicoio print "Downloading data set" posts = fetch_20newsgroups().data joblib.dump(posts, "data/posts.jl") print "Tagging %d posts" % len(posts) post_tags = indicoio.batch_text_tags(posts) joblib.dump(post_tags, "data/post_tags.jl")
