def get_sentiment():
"""Function triggered when Flask API is requested
Returns:
str: JSON containing analysis results
"""
# USER REQUEST PARAMETERS
hashtag = request.args.get('hashtag', '')
if hashtag == "":
return "Please specify a non null hashtag"
nb_days = request.args.get('nb_days', 7, type=int)
nb_days = int(min(max(nb_days, 1), 7))
nb_tweets = max(request.args.get('nb_tweets', nb_days * 10),
nb_days,
type=int)
get_topic_words = bool(int(request.args.get('get_topic_words', "1")))
n_topics = request.args.get('n_topics', 1, type=int)
n_words_per_topic = request.args.get('n_words_per_topic', 10, type=int)
lda_passes = request.args.get('lda_passes', 4, type=int)
return_tweets = bool(int(request.args.get('return_tweets', "0")))
language = request.args.get('language', "en")
# TWITTER REQUEST PARAMETERS
days_offsets = range(-nb_days + 1, 1)
query_key_value = " -is:retweet -is:quote lang:" + language
tweet_fields = "created_at,public_metrics,author_id"
max_nb_tweets_per_day = nb_tweets // len(days_offsets)
query_string = "#" + hashtag.strip() + query_key_value
# COMPUTE RESULTS
tweets = get_tweets(query_string, days_offsets, tweet_fields,
max_nb_tweets_per_day, nb_tweets, search_tweets_args)
sentiments_df, cleaned_tweets_texts, filtered_tweets_df = compute_sentiment(
tweets, model, tokenizer)
if get_topic_words:
top_topics = get_topics_from_tweets(
NLTK_DATA_PATH,
cleaned_tweets_texts,
n_topics=n_topics,
n_words_per_topic=n_words_per_topic,
n_passes=lda_passes,
force_download=False)
if return_tweets:
sentiments_tweets_df = pd.concat(
(sentiments_df, filtered_tweets_df.reset_index(drop=True)), axis=1)
results = {"sentiments_json": sentiments_tweets_df.to_json()}
else:
results = {"sentiments_json": sentiments_df.to_json()}
if get_topic_words:
results["top_topics_json"] = top_topics.to_json()
return json.dumps(results)
def main():
""" prints timeline and stats"""
if len(argv) != 4:
print 'usage: tweets.py screen_name tweet_count stat_count'
return
scr_name = argv[1]
twt_count, stat_count = int(argv[2]), int(argv[3])
tweets = get_tweets(scr_name, twt_count)
#print extract_tweets(tweets)
print len(tweets), 'tweets displayed'
tags, mentions = get_stats(tweets, scr_name, stat_count)
print_stats(tags, stat_count)
print_stats(mentions, stat_count)
def test_get_tweets_total_nb_tweets(collect_results_mock):
# collect_results_mock mocks collect_results from twitter api
expected_tweets = [[{"text": "hello"}, {}], [{"text": "hi"}, {}]]
collect_results_mock.side_effect = expected_tweets
expected_tweets = [et[0] for et in expected_tweets]
total_nb_tweets = 10
query_string = "#december -is:retweet -is:quote lang:en"
days_offsets = range(-1, 1)
tweet_fields = " "
max_nb_tweets_per_day = 2
search_tweets_args = {}
returned_tweets = get_tweets(query_string, days_offsets, tweet_fields,
max_nb_tweets_per_day, total_nb_tweets,
search_tweets_args)
assert len(returned_tweets) <= total_nb_tweets
assert returned_tweets == expected_tweets
from utils import get_tweets, result_file, write_results
from constants import *
def tweets_place(tweets):
tuples = \
tweets.filter(lambda t: t[COLUMNS.index('country_code')] == 'US' and t[COLUMNS.index('place_type')] == 'city')\
.map(lambda t: (t[COLUMNS.index('place_name')], 1))\
.aggregateByKey(0, (lambda x, y: x + y), (lambda rdd1, rdd2: rdd1 + rdd2))\
.sortByKey()\
.sortBy(lambda t: t[1], False)\
.map(lambda t: '%s\t%s' % (t[0], t[1]))\
.collect()
return tuples
if __name__ == "__main__":
task = '5'
tweets = get_tweets(task, False)
result_file = open(result_file(task), "w")
results = tweets_place(tweets)
write_results(result_file, results, cols=['place_name', 'num_tweets'])
def __init__(self, trainf_name, testf_name, smoothing_factor): self.vocab, self.vocabLength = getFV(trainf_name) self.train_tweets = get_tweets(trainf_name) self.tweets_labeled_yes, self.tweets_labeled_no = split_into_classes(trainf_name) self.test_tweets = get_tweets(testf_name) self.smoothing_factor = smoothing_factor