Actually its collect or collectset + concat_ws udf. Edward
On Sun, May 13, 2012 at 9:04 PM, Edward Capriolo <edlinuxg...@gmail.com> wrote: > I just through a derivative of collectset called collect but we need > some type of string join that takes an array and a delimiter. > > https://github.com/edwardcapriolo/hive-collect > > On Fri, May 11, 2012 at 4:05 PM, Edward Capriolo <edlinuxg...@gmail.com> > wrote: >> The main issue with group_concat is that aggregates have to keep each >> column in memory and that is a big problem. If The user knows the list >> will be small you could write a UDAF like collectset, collect which >> puts each value into a list and then you can lateral view that list. >> >> Edward >> >> On Fri, May 11, 2012 at 4:02 PM, Saurabh S <saurab...@live.com> wrote: >>> >>> As far as I understand, there is no equivalent of MySQL group_concat() in >>> Hive. This stackoverflow question is from Sept 2010: >>> http://stackoverflow.com/questions/3703740/combine-multiple-rows-into-one-space-separated-string >>> >>> Does anyone know any other method to create a delimited list from from >>> table rows? There are a few methods listed in Anthony Molinaro's SQL >>> Cookbook, but all of them use over(partition by) method which is also not >>> available in Hive. >>>
