bieber_screennames.pig

-- load all of the parsed tweets as (id, timestamp, screenname, tweet)

raw = LOAD 'parsed' USING PigStorage('\t') AS (id:chararray,timestamp:chararray,screenname:chararray,tweet:chararray);

-- limit the twitter data to where screenname matches bieber, case-insensitive
fltr = FILTER raw BY screenname matches '.*?[Bb][Ii][Ee][Bb][Ee][Rr].*?';

-- group by screenname to count the tweets
grpd = GROUP fltr BY screenname;

-- count the tweets per screenname.  note that these are total tweets, not
-- bieber tweets :D

cntd = FOREACH grpd GENERATE $0,COUNT($1) as cnt;

-- order the list of screennames by count.
srtd = ORDER cntd BY cnt;

-- write the output file
store srtd INTO 'usernameswithbieber2';