How to compute the count of duplicates found in both Data1 and Data2 and in Data 3, and in one or the other?

Hello,
For testing and verification purposes, I am trying to compute the count of good duplicates found in both Data1 and Data2  and in Data3, and in one or the other. These are the steps I am trying to accomplish to get to my final results:

#1.  Create a list (Data4) of unique Data1 and Data2 where variable rec_flag=A, B, C, D.
#2.  Build a duplicate (Data5) file from the list in #1 by generating a record linking each two records where Var3 and Var4 are linked to the same Var6 and Var7.
#3.  Match Data5 file with Data3 where variable rec_flag=B, C, D using key variables Var3, Var4, Var6, and Var7.
#4.  Display in the output .lis the number of records in:
     #4.1 Both Data5 and Data3
     #4.2 Data5 only
     #4.3 Data3 only
 
The issue that I am having is with #2 where I need to build a duplicate file Data5.

Here is my SAS code below with missing step #2:

libname mydir "/home/myprofile/links"; 
%macro cnt_lnks;

 /*  #1 */
proc sort data=mydir.Data1
          out=Data1out(where=(rec_flag='A', 'B', 'C', 'D') keep=Var3 Var4 Var6 Var7 rec_flag);
          by Var3 Var4 Var6 Var7;
run;

proc sort data=mydir.Data2
          out=Data2out(where=(rec_flag='A', 'B', 'C', 'D') keep=Var3 Var4 Var6 Var7 rec_flag);
          by Var3 Var4 Var6 Var7;
run;

/*********************************************
 Datastep: merge Data1 & Data2
           by Var3 Var4 Var6 Var7
 ********************************************/
data all_links1;
  merge Data2out (in=ina)
       Data1out (in=inb);
  by Var3 Var4 Var6 Var7;
  if ina and inb;
run;

/***********************************
 Remove duplicate records
 In: all_links1
 out: n_links
 ************************************/
proc sort data=all_links1 nodupkey out=n_links;
     by Var3 Var4 Var6 Var7;
run;

/* #2  -  how can I build this one?  */



/******************************************
#3
 Create Data3out: keep recs where rec_flag=B,C,D
 in: Data3
 out: Data3out
 *******************************************/
data Data3out(where=(rec_flag in('B','C','D')) keep=Var3 Var4 Var6 Var7 rec_flag);
 set mydir.Data3;
run;

/*******************************************
 sort Data3out by Var3 Var4 Var6 Var7
 *******************************************/
proc sort data= Data3out; by Var3 Var4 Var6 Var7; run;

/****************************************
 Datastep: merge n_links & Data3out
 in: n_links & Data3out
 out: all_links2
 *****************************************/

data all_links2;
 merge n_links (in=ina) Data3out (in=inb);
 by Var3 Var4 Var6 Var7;
 if ina and inb;
run;

/*****************************************
 #4 
Counts:
 1. Total recs: Combined Data2out & Data1out, where rec_flag=A & nodups
 2. Total recs: Data3out, where rec_flag=B,C,D
 3. Total recs: combined [Data2out & Data1out] & Data3out
 ***********************************************/

PROC SQL NOPRINT;
    SELECT COUNT(*)
    INTO:combo1
    FROM n_links;

    SELECT COUNT(*)
    INTO:Data3out2
    FROM Data3out;

    SELECT COUNT(*)
    INTO:combo2
    FROM all_links2;
quit;

 /**********************************************
  Output counts
  out: Counts.lis
  **********************************************/

  data _null_;
   file "Counts.lis";
   put 50*'*';
   put " ";
   put "Additional counts for Data3out and Data1out/Data2out duplicates";
   put " ";
   put "Data1out and Data2out where REC_FLAG=A: &combo1";
   put "Data3out & Data1out and Data2out: &combo2";
   put "Data3out where REC_FLAG= B,C,D: &Data3out1";
  run;
%mend cnt_lnks;
%cnt_lnks;

Select allOpen in new window