Create Identifier/counter Based On Some Shared Columns And Seperate Based On Other Columns

I want to create a calculated column based on a shared column but the calculated column should 'restart' and be grouped based on a third column. As described in the picture below C

Solution 1:

Use windowing function DENSE_RANK() with an OVER() clause:

DECLARE@tblTABLE(Column1 INT,Column2 INT,Column3 VARCHAR(100));
INSERTINTO@tblVALUES(1,1,'A')
                      ,(1,2,'A') 
                      ,(1,3,'B') 
                      ,(2,1,'A') 
                      ,(2,2,'A') 
                      ,(2,3,'B') 
                      ,(3,1,'A') 
                      ,(3,2,'B') 
                      ,(3,3,'V');

SELECT*
      ,DENSE_RANK() OVER(PARTITIONBY Column1 ORDERBY Column3) AS ComputedColumn
FROM@tbl;

The PARTITION BY will re-start the counter for each new value in column1, while the ORDER BY defines the ranking.

Hint: Do not paste pictures!

For your next question please follow my example to create a stand-alone example reproducing your issue and add the code you've tried yourself.