Hi, Bernd, What you described sounds like a bug to me. If you already have something set up to exercise these operations already and can share with us, we would very much like to have those test cases. Otherwise, I will likely have time to look into it later today.
Thanks. John On 5/12/2010 9:09 AM, Bernd Jaenichen wrote: > Hi John, > > when i select "DISTINCT(c_1) as n, c_2" using the table::select > interface on a table with more than one row i get the expected result: > > "n" => number of distinct rows according to the value of column c_2 > "count0" => number of matching rows according to the value of column c_2 > "c_2" => respective value of column c_2 > > > > If i select the same on a table with only one row i get: > > "n" => value of column c_1 > "count0" => 1 > "c_2" => value of column c_2 > > > > If i select the same on a table with two rows having different values in > c_2 i get this: > > "n" => value of column c_1 (i.e. 47) > "count0" => 1 > "c_2" => value of column c_2 (i.e. 2) > > "n" => value of column c_1 (i.e. 48) > "count0" => 1 > "c_2" => value of column c_2 (i.e. 1) > > > > I'm not sure if this is a feature or a bug could you > please explain it to me. > > > > > Kind regards, > > Bernd Jaenichen > > _______________________________________________ FastBit-users mailing list [email protected] https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
