using talend tdenormalize i'm getting erratic results. data in csv files columns this:
contact id | xx | yy | zz | account id | first name | last name
1|x|y|z| a1 | john | smith
2|x|y|z| a1 | sally | smith
3|x|y|z| b1 | bill | jones
4|x|y|z| b1 | karen | jones
i'm trying tdenormalize create :
1|x|y|z| a1%a1 | john%sally | smith%smith
4|x|y|z| b1%b1 | bill%karen | jones%jones
(this contacts table i'm working on child table accounts table)
as can see i'm getting small subset of records "normalized" have tried tdenormalizesortedrow similar results. it's not clear how talend determines rows aggregate. filled first column account number (overlaying contact id) results in talend screen shot. great.
after lots of tests, seems talend requires other columns of data equal perform tdenormalize correctly. not know other unmatched columns. new function denormalize , select other columns "pull first" value or last value nice feature add.(wink wink).
Comments
Post a Comment