Essentially there is no equivalent in DB2.
The logic behind this being that there is no "natural" rowid attached to a certain row: a certain row can be the 4th result row from a certain query, but the same row could as well be the 6th row from the SAME query on the SAME table when the optimizer chose to implement the query differently! (E.g., with or without the use of an index.)
Typically you probably wanted to change a value in the row with rowid=4 because some (independent) program logic found out that this row had a certain property, i.e., "where" condition.
In that case you could do one of three things:
1. Add that condition to your UPDATE statement:
update mytab set c1 = -c1 where CONDITION;
2. Suppose you are running through a result table with a cursor, and at a certain iteration the condition applies for the current row; in that case use
update mytab set c1 = -c1 where CURRENT OF cursorname;
3. Use a subquery to produce the primary key values of mytab that satisfy the condition:
update mytab set c1 = -c1 where mytab.pkey IN ( SELECT pkey FROM mytab WHERE CONDITION );