It looks like you're essentially doing a pivot function. Your best bet is to write a custom UDAF or look at the windowing functions available in recent releases.
Matt On Dec 28, 2013 12:57 PM, "Raj Hadoop" <hadoop...@yahoo.com> wrote: > Dear All Hive Group Members, > > I have the following requirement. > > Input: > > Ticket#|Date of booking|Price > 100|20-Oct-13|54 > 100|21-Oct-13|56 > 100|22-Oct-13|54 > 100|23-Oct-13|55 > 100|27-Oct-13|60 > 100|30-Oct-13|47 > > 101|10-Sep-13|12 > 101|13-Sep-13|14 > 101|20-Oct-13|6 > > > Expected Output: > > Ticket#|Initial|Delta1|Delta2|Delta3|Delta4|Delta5 > 100|20-Oct-13,54| > 21-Oct-13,2|22-Oct-13,0|23-Oct-3,1|27-Oct-13,6|30-Oct-13,-7 > 101|10-Sep-13,12|13-Sep-13,2|20-Oct-13,-6||| > > The number of columns in the expected output is a dynamic list depending > on the number of price changes of a ticket. > > 1) What is the best design to solve the above problem in Hive? > 2) How do we implement it? > > Please advise. > > Regards, > Raj > > > > > >
