我对两个表执行了内部联接。但是，我无法对其中一个克隆进行求和：
执行的查询：

sample1 = load '/user/tweets/samples.csv' using PigStorage AS (line:chararray);
words = FOREACH sample1 GENERATE FLATTEN(TOKENIZE(REPLACE(LOWER(TRIM(line)),'[\\p{Punct},\\p{Cntrl}]',''))) AS word
newinnerjoin = join words by word, wordlexion by lexword;

下面是表的输出：newinnerjoin

(important,important,2)
(irritated,irritated,-3)
(promoting,promoting,1)
(promoting,promoting,1)
(appreciate,appreciate,2)
(confidence,confidence,2)

我想对内部连接结果的第3列执行聚合。所以，我想把总和计算为2+-3+1+1+2+2=5有没有办法不把内部连接结果存储在csv文件中？请告知。
谢谢