CODEWITHSUNDEEP
pythontwitterpython-3.x
Edgar Derby
Edgar Derby

Reputation: 2825

Count number of results for a particular word on Twitter (API v1.1)

The question has already been asked a couple of times. For example here or here. Yet, no accepted answer was found. Also, both the questions refer to Twitter API v1.0, which is no longer in use. Thus, I thought it could be beneficial to share a simple piece of code I wrote in order to have the number of tweets containing a given keyword (or phrase).

If you have any feedbacks, do not hesitate to reply.

Upvotes: 5

Views: 5537

Answers (1)

Edgar Derby
Edgar Derby

Reputation: 2825

Here it is:

#Import the required modules
from twython import Twython
import json
import csv

#Set parameters
keyword = 'kittens'; #The desired keyword(s)
tweetsXiteration = 100; #Where 100 is the max
dateFrom = '2014-02-01'; #Inclusive (YYYY-MM-DD)
dateTo = '2014-02-02'; #Exclusive (YYYY-MM-DD)
done = False; #Must be false

#Setting the OAuth
Consumer_Key = 'XXX';
Consumer_Secret = 'XXX';
Access_Token = 'XXX';
Access_Token_Secret = 'XXX';

#Connection established with Twitter API v1.1
twitter = Twython(Consumer_Key, Consumer_Secret, Access_Token, Access_Token_Secret);

#Twitter is queried
response = twitter.search(q = keyword, count = tweetsXiteration, since = dateFrom, until = dateTo, result_type = 'mixed');

#Results (partial)
countTweets = len(response['statuses']);

#If all the tweets have been fetched, then we are done
if not ('next_results' in response['search_metadata']): 
    done = True;

#If not all the tweets have been fetched, then...
while (done == False):

    #Parsing information for maxID
    parse1 = response['search_metadata']['next_results'].split("&");
    parse2 = parse1[0].split("?max_id=");
    parse3 = parse2[1];
    maxID = parse3;

    #Twitter is queried (again, this time with the addition of 'max_id')
    response = twitter.search(q = keyword, count = tweetsXiteration, since = dateFrom, until = dateTo, max_id = maxID, include_entities = 1, result_type = 'mixed');

    #Updating the total amount of tweets fetched
    countTweets = countTweets + len(response['statuses']);       

    #If all the tweets have been fetched, then we are done
    if not ('next_results' in response['search_metadata']): 
        done = True;

print(countTweets);

Keep in mind that:

  1. You need to authenticate trough OAuth;
  2. You can only fetch results no older than a week;
  3. If you want to search for more than one word, you need to use "" if you are only interested in the results containing the two words in that particular order (e.g., '"Stack Overflow"').

Additional information can be found here or on Twitter official documentation.

Upvotes: 6

Related Questions