Finding matches in different fields in a table

[ChickenLittle]

Hi, what is the most efficient way to grab each unique value from FieldABC for every person listed(i.e. AB and then CD and then EF then GH, etc without duplicates). And then either do option 1 or option 2 as follows:

Option1: Set up a macro to go through each of those unique values to produce a proc freq between FieldABC and FieldXYZ to see where there are matches.

Option2: Do a proc freq displaying all the unique values from FieldABC vs FieldXYZ to see where there are matches.

The table set up looks like this:

Person                  FieldABC                              FieldXYZ

PersonA               AB                                          CD;EF

PersonB               CD;EF;GH                             EF;GH

PersonC               AB;EF                                    AB;

PersonD               CD                                          EF;GH

Result would look something like this:

                AB          CD          EF           GH         

AB          1              0              0              0

CD          0              0              0              0

EF           0              0              1              0

GH          0              0              0              1

Or results could be something like this:

For value ‘AB’

                     Field XYZ

FieldABC              1

For value ‘EF’

                     Field XYZ

FieldABC              1

[Tom · Accepted solution]

Normalize the data and then just count.

data tall;
  set have;
  do index=1 to countw(fieldabc,';');
    word = scan(fieldabc,index,';');
    output;
  end;
  do index=1 to countw(fieldxyz,';');
    word = scan(fieldxyz,index,';');
    output;
  end;
  keep person word;
run;
proc freq ;
  tables person*word / noprint out=counts;
run;

[ChickenLittle]

Thanks!