rank in hive

 select date, project, variant, page_title,
        rank() over (PARTITION by date, project, variant ORDER BY tmp.pageviews desc) as rank
    (select concat(year, lpad(month, 2, '0'), lpad(day, 2, '0')) as date,
            project,
            variant,
            page_title,
            sum(view_count) as pageviews
       from wmf.pageview_hourly
      where year = 2015
        and month = 7
        and day = 1
        and agent_type = 'user'
        and page_title != '-'
        and project = 'etwiki'
      group by year, month, day, project, variant, page_title
    ) tmp
  where rank <= 1000
;