I have 3 tables that I want to merge, each with a different column of interest. I also have an id variable that I want to do separate merges "within" id. The idea is that I want to merge X, Y, and Z by date (within ID), and have missing values if that date does not exist for a particular variable.
Table X:
ID Date X
1 2012-01-01 101
1 2012-01-02 102
1 2012-01-03 103
1 2012-01-04 104
1 2012-01-05 105
2 2012-01-01 150
Table Y:
ID Date Y
1 2012-01-01 301
1 2012-01-02 302
1 2012-01-03 303
1 2012-01-11 311
2 2012-01-01 350
Table Z:
ID Date Z
1 2012-01-01 401
1 2012-01-03 403
1 2012-01-04 404
1 2012-01-11 411
1 2012-01-21 421
2 2012-01-01 450
Desired Result Table:
ID Date X Y Z
1 2012-01-01 101 301 401
1 2012-01-02 102 302 .
1 2012-01-03 103 303 403
1 2012-01-04 104 . 404
1 2012-01-05 105 . .
1 2012-01-11 . 311 411
1 2012-01-21 . . 421
2 2012-01-01 150 350 450
Any ideas how to write this SQL statement? I've tried messing around with "full joins" and where statements for cross products, but I keep getting duplicate values for some of my ID-date combinations, or sometimes no ID.
Any help would be appreciated.