how to merge data sets with different coding system

[Reeza]

That's not making sense. Please post your proc contents. 

[Bal23]

93

var1

Num

8

# Variable Type Len Format Informat Label 6

var2

Char

4

[Reeza]

No formats?

What does each one of those variables look like - to make sure?

[Bal23]

var1  0234     0389      0423

var2   234       389        423

[Reeza · Accepted solution]

data one;
input var1;
cards;
123
456
789
234
;
run;

data two;
input var2 $;
cards;
0123
0456
0789
234
;
run;

data two_formatted;
set two;
var1=input(var2, best12.);
run;

proc sort data=one; by var1; run;
proc sort data=two_formatted; by var1; run;

data merged;
merge one two_formatted;
by var1;
run;

[Peter_C]

In sql a join could convert the char to num
On input( table1.leadz, best15.) = inpur( table2.nolead, best15.)

But that join fails when the input function returns missing for unsuitable strings

To achieve similar results on a data step merge the numeric form of these values need to be converted in steps before the merge.
I/O impact can be minimised by these prior steps creating table1 and table2 as views