number 和binary_float binary_double 性能比较

Number类型和float类型的比较

http://blog.csdn.net/wyzxg/article/details/6184877

SQL> create table zfk_test03
2 ( number_type number,float_type binary_float,double_type binary_double);

Table created

SQL> insert into zfk_test03 select rownum,rownum,rownum from dba_objects;

139815 rows inserted

SQL> commit;

Commit complete

SQL> set timing on;

SQL> select sum(sin(number_type)) from zfk_test03;

SUM(SIN(NUMBER_TYPE))
---------------------
1.40583495761752

Executed in 3.031 seconds

再次运算下
SQL> select sum(sin(number_type)) from zfk_test03;

SUM(SIN(NUMBER_TYPE))
---------------------
1.40583495761752

Executed in 2.797 seconds

SQL> select sum(sin(float_type)) from zfk_test03;

SUM(SIN(FLOAT_TYPE))
--------------------
1.40583495761753

Executed in 0.578 seconds

SQL> select sum(sin(double_type)) from zfk_test03;

SUM(SIN(DOUBLE_TYPE))
---------------------
1.40583495761753

Executed in 0.453 seconds

看样子FLOAT确实比NUMBER快NB

SQL> select sum(number_type) from zfk_test03;

SUM(NUMBER_TYPE)
----------------
9774187020

Executed in 0.438 seconds

SQL> select sum(number_type) from zfk_test03;

SUM(NUMBER_TYPE)
----------------
9774187020

Executed in 0.375 seconds

SQL> select sum(float_type) from zfk_test03;

SUM(FLOAT_TYPE)
---------------
9774106620

Executed in 0.406 seconds

SQL> select sum(float_type) from zfk_test03;

SUM(FLOAT_TYPE)
---------------
9774106620

Executed in 0.375 seconds

SQL> select sum(double_type) from zfk_test03;

SUM(DOUBLE_TYPE)
----------------
9774187020

Executed in 0.407 seconds

SQL> select sum(double_type) from zfk_test03;

SUM(DOUBLE_TYPE)
----------------
9774187020

Executed in 0.375 seconds

性能方面是一样的.

一般我们用数字类型基本上都是 SUM()和count() 使用SIN()函数的场景很少