hadoop - percentile_approx in hive returning zero -
i have been trying check percentile_approx set of users. intention behind top 25% of customers in data set. so, in order check that, ran following hive query.
select percentile_approx(amount, 0.75) sales
however, value returned query 0.0
. not sure problem is. when run query on sample of few records result expected.
can please shed light on this?
note - trying find percentile in data set containing more 3.3 m records.
generally percentile_approx() works on integer type data. please make sure have applied on column has integers.
Comments
Post a Comment