Generate Random Data, Using bind variables to improve performance

create table salary_rates
(
dept char(1),
rate number(2,1)
);

alter table salary_rates cache;

create table emp
(
id number,
dept char,
salary number
);

create unique index emp_id_idx on emp(id);

create sequence emp_id;


CREATE OR REPLACE TRIGGER emp_trg
BEFORE INSERT ON emp
FOR EACH ROW
WHEN (new.id IS NULL)
BEGIN
SELECT emp_id.NEXTVAL INTO :new.id FROM dual;
END;
/

--------Salary rates Loading data------------------------

insert into salary_rates values( 'A',0.1);
insert into salary_rates values( 'B',0.2);
insert into salary_rates values( 'C',0.3);
insert into salary_rates values( 'D',0.4);
commit;
--------Salary rates Loading data ------------------------

------------------------- Generate random depts -------------------------
CREATE OR REPLACE FUNCTION GET_RANDOM_DPT RETURN CHAR
IS
X NUMBER;
RTV CHAR;
BEGIN
X:=ROUND(dbms_random.value(1, 5));

IF (X=1) THEN
RTV := 'A';
ELSIF (X=2) THEN
RTV := 'B';
ELSIF (X=3) THEN
RTV := 'C';
ELSE
RTV := 'D';
END IF;

RETURN RTV;
END;
/
------------------------- Generate random depts -------------------------

create or replace procedure gen_data ( emp_population number)
is
begin
FOR i IN 1 .. emp_population loop
insert into emp(dept,salary) values(GET_RANDOM_DPT,ROUND(dbms_random.value(1,1000000)));
commit;
END LOOP;
end;
/
-----------------------------------------------------------------------------------------------------------

/* Scenario : Update employee salary , ONLY those employees current salary is less than 3000 according the employee's department increase rates.
------------------------------------------------------------------------------------------------------
update 1
Elapsed: 00:00:01.70


CREATE OR REPLACE PROCEDURE UPDATE1
AS
rate_ number;
begin
---open c_emp;
for i IN ( select id,dept,salary from emp) loop
if (i.salary < 3000 ) then
select rate into rate_ from salary_rates where dept=i.dept;
update emp set salary=(salary * rate_/100)+salary where id=i.id;
commit;
else
null;
end if;
end loop;
END UPDATE1;



CREATE OR REPLACE PROCEDURE UPDATE2
AS
rate_ number;
type rc is ref cursor;
q rc;
begin
for i IN ( select id,dept,salary from emp) loop
if (i.salary < 3007 ) then
open q for ' select rate from salary_rates where dept= :y ' using i.dept;
fetch q into rate_;
execute immediate ( ' update emp set salary=(salary * :r / 100) + salary where id= :x') using rate_,i.id;
commit;
else
null;
end if;
end loop;
END UPDATE2;
Elapsed: 00:00:01.32