If you are unable to create a new account, please email support@bspsoftware.com

 

News:

MetaManager - Administrative Tools for IBM Cognos
Pricing starting at $2,100
Download Now    Learn More

Main Menu

How to stitch 4 tables of data back together???

Started by porsche3, 09 Feb 2016 03:06:19 PM

Previous topic - Next topic

porsche3

I have a need were I need to look at a historical table at 4 different points in time and based on a few other filters. The only difference is the points in time. I have the 4 tables and I have the data grouped by Machine Type and Models. I need to stitch them back together. Example below:

Table Y1
Model      Type      Qty
======= ====== =====
2074        001        18
2096        360        205
DSC        Base      22

Table Y2
Model      Type      Qty
======= ====== =====
2074        001        33
2096        360        156
DSC        Base      29
3678       345         16

Table Y3
Model      Type      Qty
======= ====== =====
2074        001        28
2096        360        295
DSC        Base      36
3678       345         25

Table Y4
Model      Type      Qty
======= ====== =====
2074        001        19
2096        360        200
DSC        Base      20
3678       345         18

Stitched Table

Model     Type     Y1       Y2      Y3        Y4
======  ====== ====  ====  =====  ======
2074        001        18         33     28         29
2096        360        205      156   295       200
DSC         Base     22          28    36         20
3678        345        0            16    25        18

Any help would be appreciated!

bdbits

Since this is FM, I am assuming you want to create a query subject - so somewhat depending on the database in use, you could probably use SQL something like:

with u as (
  select model, type, qty as y1, 0 as y2, 0 as y3, 0 as y4
  union
  select model, type, 0 as y1, qty as y2, 0 as y3, 0 as y4
  union
  select model, type, 0 as y1, 0 as y2, qty as y3, 0 as y4
  union
  select model, type, 0 as y1, 0 as y2, 0 as y3, qty as y4
)
select model, type, sum(y1), sum(y2), sum(y3), sum(y4)
group by model, type

Technically speaking, you would likely not need the "as" portion on any but the first select and the rest would be ignored; only order is important on subsequent selects. I included them thinking it might clarify what is intended.