Merging one variable in a column of a table with all the values in a different column of another.

[IsoscelesKramer]

I have this dilemma, I tried to do this in excel, but with the sheer amount of data its going to take forever and I have not found any sort of method in excel to perfrom a loop of some sort. 

Basically I have these 5 levels as each level has an addition of code. Bh basically take EBRDBG01 and merge that with all the variables in column PH4, then do the same with EBRDBG02...etc. This same basically needsto be done with level 3, level 4.

Level 3 is easy to compose since Level 2 only has one variable in the table. However Level 3 and Level 4 have several variables in the table and each one needs to be merged with the contents of the PH columns. 

Is this even possible in SAS? I appreciate any advice or pointers, thanks guys!

Also please let me know if you need more clarification, I do apologize for the codes, they tend to get confusing looking as they grow bigger. 

Level 2

EBRD

       +

PH3

BG01

BG02

BG03

BG04

BG05

BG06

BG07

BG08

BG09

BG10

     =

Level 3

EBRDBG01

EBRDBG02

EBRDBG03

EBRDBG04

EBRDBG05

EBRDBG06

EBRDBG07

EBRDBG08

EBRDBG09

EBRDBG10

         +

PH4

BBR01

BBR02

BBR03

BBR04

BBR05

BBR06

BBR07

BBR08

BBR09

BBR10

   =  

Level 4

EBRDBG01BBR01

EBRDBG01BBR02

EBRDBG01BBR03

EBRDBG01BBR04

EBRDBG01BBR05

EBRDBG01BBR06

EBRDBG01BBR07

EBRDBG01BBR08

EBRDBG01BBR09

EBRDBG01BBR10

EBRDBG02BBR01

EBRDBG02BBR02

EBRDBG02BBR03

EBRDBG02BBR04

EBRDBG02BBR05

EBRDBG02BBR06

EBRDBG02BBR07

EBRDBG02BBR08

EBRDBG02BBR09

EBRDBG02BBR10

EBRDBG03BBR01

EBRDBG03BBR02

EBRDBG03BBR03

EBRDBG03BBR04

EBRDBG03BBR05

EBRDBG03BBR06

EBRDBG03BBR07

EBRDBG03BBR08

EBRDBG03BBR09

EBRDBG03BBR10

This continues for EBRDG04, 05, 06, etc. 

        +

PH5

BF001

BF002

BF003

BF004

BF005

BF006

BF007

BF008

BF009

BF010

BF011

BF012

BF013

BF014

BF015

BF016

BF017

This continues for EBRDBG04..05...06...etc.

[Astounding · Accepted solution]

For all combinations, SQL is usually the right tool for the job.  Here's the approach:

proc sql;

create table want as

select strip(ph1) || strip(ph2) || strip(ph3) || strip(ph4) || strip(ph5) as combo

from level1, level2, level3, level4, level5;

quit;

I'm making assumptions about the names of the fields, the names of the incoming data sets, whether you want to strip out blanks, and you may want to set a length for the resulting field.  But this should be a viable approach.

[IsoscelesKramer]

I am a bit new to sas, so would I import the excel file into SAS and run the proc sql command that you have provided? or would I import each column Level 2, PH3, Level 3, PH4, Level 4, PH5 as seperate tables?

Thanks for the help, I really appreciate it!

[Astounding]

They way I pictured it, you would import 5 separate files into SAS:

The file named Level1 would contain the variable PH1, the file named Level2 would contain the variable PH2, etc.

That's the easiest way to allow each variable to have a different number of unique values.

The program could be adapted to handle other data structures, but that's the way it is designed for now.

[IsoscelesKramer]

Thanks so much for the help! I got exactly what I needed.