NUMBER Data type's Negative Scale

Question

This is what I read in the 1z0-051 Study Guide:
The NUMBER data type may optionally be qualified with a precision and a scale. The precision sets the maximum number of digits in the number, and the scale is how many of those digits are to the right of the decimal point. If the scale is negative, this has the effect of replacing the last digits of any number inserted with zeros, which do not count toward the number of digits specified for the precision. If the number of digits exceeds the precision, there will be an error; if it is within the precision but outside the scale, the number will be rounded (up or down) to the nearest value within the scale. I am not sure if I understood this paragraph the correct way or not and hence this query:
I created this table t1
create table t1
(col1 number,
col2 number(4),
col3 number(4,2),
col4 number(4,-2));
INSERT INTO T1 (COL4) VALUES (1234); = 1200 is inserted
INSERT INTO T1 (COL4) VALUES (12.34); = 0 is inserted
INSERT INTO T1 (COL4) VALUES (12345.34); = 12300 is inserted. Should is not throw an error as the whole value should be of length 4 (including the precision).
However, INSERT INTO T1 (COL4) VALUES (123456789.34); throws error.How is this working? Please help!
Edited by: TuX4EvA on Sep 29, 2009 4:12 AM

Answer 1

Please, observe the following example:
CREATE TABLE test1
col1 NUMBER (6, 2),
col2 NUMBER (6, -1),
col3 NUMBER (6, -2),
col4 NUMBER (6, -3)
INSERT INTO test1 (col1, col2, col3, col4) VALUES (1234.89, 1234.89, 1234.89, 1234.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (1000.89, 1000.89, 1000.89, 1000.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (2020.02, 2020.02, 2020.02, 2020.02);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (56, 56, 56, 56);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (56.89, 56.89, 56.89, 56.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (20.89, 20.89, 20.89, 20.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (49.89, 49.89, 49.89, 49.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (50.89, 50.89, 50.89, 50.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (501.89, 501.89, 501.89, 501.89);
INSERT INTO test1 (col1, col2, col3, col4) VALUES (158.23, 158.23, 158.23, 158.23);
COMMIT;
SELECT * FROM test1;When you select from the table, you will receive:
        col1          col2         col3     col4
   1234,89           1230           1200     1000
   1000,89           1000           1000     1000
   2020,02           2020           2000     2000
        56             60            100     0
     56,89             60            100     0
     20,89             20              0     0
     49,89             50              0     0
     50,89             50            100     0
    501,89            500            500     1000
    158,23            160            200     0From this example and according to the oracle documentation SQL Reference one could see that scale equal to -1 means to round the precision to the tens, -2 means to the hundreds, -3 - to the thousands and so on. But I cannot say it as well as it is said on page 44 from [Oracle Database SQL Reference|http://download.oracle.com/docs/cd/B19306_01/server.102/b14200.pdf].
Hope it was useful :)