FLOAT

Single precision floating-point number. A FLOAT column uses 4 bytes and stores approximate values with less precision than DOUBLE.

Syntax

FLOAT[(M,D)] [SIGNED | UNSIGNED | ZEROFILL]

Description

A small (single-precision) floating-point number (see DOUBLE for a regular-size floating point number). Allowable values are:

• -3.402823466E+38 to -1.175494351E-38

• 0

• 1.175494351E-38 to 3.402823466E+38.

These are the theoretical limits, based on the IEEE standard. The actual range might be slightly smaller depending on your hardware or operating system.

M is the total number of digits and D is the number of digits following the decimal point. If M and D are omitted, values are stored to the limits allowed by the hardware. A single-precision floating-point number is accurate to approximately 7 decimal places.

UNSIGNED, if specified, disallows negative values.

Using FLOAT might give you some unexpected problems because all calculations in MariaDB are done with double precision. See Floating Point Accuracy.

For more details on the attributes, see Numeric Data Type Overview.

EXAMPLES

SIGNED and UNSIGNED

The FLOAT data type may be SIGNED (allowing negative values) or UNSIGNED (not allowing negative values).

Example of FLOAT SIGNED (SIGNED is the default):

CREATE TABLE float_signed_example (
  description VARCHAR(20),
  example FLOAT,
  sz6_2 FLOAT(6,2),
  sz20_19 FLOAT(20,19) SIGNED
);

SET @pi = 3.1415926535897932384626433832795;

INSERT INTO float_signed_example VALUES
  ('Pi', @pi, @pi, @pi),
  ('Series', 1234.567890123, 1234.567890123, 1.234567890123),
  ('Negative', -1234.567890123, -1234.567890123, -1.234567890123),
  ('Various', 1234567890, 9999.99, 9.9999999999999999999);

SELECT * FROM float_signed_example;

+-------------+------------+----------+------------------------+
| description | example    | sz6_2    | sz20_19                |
+-------------+------------+----------+------------------------+
| Pi          |    3.14159 |     3.14 |  3.1415927410125732000 |
| Series      |    1234.57 |  1234.57 |  1.2345678806304932000 |
| Negative    |   -1234.57 | -1234.57 | -1.2345678806304932000 |
| Various     | 1234570000 |  9999.99 | 10.0000000000000000000 |
+-------------+------------+----------+------------------------+

Example of FLOAT UNSIGNED:

CREATE TABLE float_unsigned_example (
  description VARCHAR(20),
  example FLOAT UNSIGNED,
  sz6_2 FLOAT(6,2) UNSIGNED,
  sz20_19 FLOAT(20,19) UNSIGNED
);

SET @pi = 3.1415926535897932384626433832795;

INSERT INTO float_unsigned_example VALUES
  ('Pi', @pi, @pi, @pi),
  ('Series', 1234.567890123, 1234.567890123, 1.234567890123),
  ('Various', 1234567890, 9999.99, 9.9999999999999999999);

SELECT * FROM float_unsigned_example;

+-------------+------------+---------+------------------------+
| description | example    | sz6_2   | sz20_19                |
+-------------+------------+---------+------------------------+
| Pi          |    3.14159 |    3.14 |  3.1415927410125732000 |
| Series      |    1234.57 | 1234.57 |  1.2345678806304932000 |
| Various     | 1234570000 | 9999.99 | 10.0000000000000000000 |
+-------------+------------+---------+------------------------+

Out of Range

A value is considered "out of range" when it is too small or too large to be stored in a data type. The size specified when creating the column is the limit for what values can be represented. The limits can also vary based on your hardware and operating system. When SQL_MODE is strict (the default) an out-of-range value generates an error and the operation fails. If strict mode is not in effect, the value is rounded to the nearest valid value and a warning is generated (which might be hidden, depending on your warning settings).

A value whose significant digits must be rounded to fit only generates a warning note about data truncation, since it is only an out-of-range value if the rounding causes the value to overflow. A somewhat strange exception happens when the decimal places are 16 digits or larger: at that point the value can round up to be one digit larger than you would expect to be accepted, but only for the next larger power of 10. For instance, a FLOAT(17,16) should max out at 9.9999999999999999, but that value rounds up to being equivalent to 10 (and 11 overflows).

TRUNCATE float_signed_example;

-- Disable strict mode
SET sql_mode=(SELECT REPLACE(@@sql_mode, 'STRICT_TRANS_TABLES', ''));

INSERT INTO float_signed_example VALUES
  ('OK', 3.402823466e38, NULL, NULL),
  ('Overflow', 4.0e38, NULL, NULL),
  ('OK', -3.402823466e38, NULL, NULL),
  ('Underflow', -4.0e38, NULL, NULL),
  ('OK', NULL, 9999.99, NULL),
  ('Rounded overflow', NULL, 9999.995, NULL),
  ('Overflow', NULL, 10000, NULL),
  ('OK', NULL, -9999.99, NULL),
  ('Rounded underflow', NULL, -9999.995, NULL),
  ('Underflow', NULL, -10000, NULL),
  ('OK', NULL, NULL, 9.9999999999999999),
  ('Rounded OK', NULL, NULL, 9.99999999999999995),
  ('Overflow', NULL, NULL, 11),
  ('OK', NULL, NULL, -9.9999999999999999),
  ('Rounded OK', NULL, NULL, -9.99999999999999995),
  ('Underflow', NULL, NULL, -11);

Warning (sql 1264): Out of range value for column 'example' at row 2
Warning (sql 1264): Out of range value for column 'example' at row 4
Warning (sql 1264): Out of range value for column 'sz6_2' at row 6
Warning (sql 1264): Out of range value for column 'sz6_2' at row 7
Warning (sql 1264): Out of range value for column 'sz6_2' at row 9
Warning (sql 1264): Out of range value for column 'sz6_2' at row 10
Warning (sql 1264): Out of range value for column 'sz20_19' at row 13
Warning (sql 1264): Out of range value for column 'sz20_19' at row 16

SELECT * FROM float_signed_example;

+-------------------+-------------+----------+-------------------------+
| description       | example     | sz6_2    | sz20_19                 |
+-------------------+-------------+----------+-------------------------+
| OK                |  3.40282e38 |     NULL |                    NULL |
| Overflow          |  3.40282e38 |     NULL |                    NULL |
| OK                | -3.40282e38 |     NULL |                    NULL |
| Underflow         | -3.40282e38 |     NULL |                    NULL |
| OK                |        NULL |  9999.99 |                    NULL |
| Rounded overflow  |        NULL |  9999.99 |                    NULL |
| Overflow          |        NULL |  9999.99 |                    NULL |
| OK                |        NULL | -9999.99 |                    NULL |
| Rounded underflow |        NULL | -9999.99 |                    NULL |
| Underflow         |        NULL | -9999.99 |                    NULL |
| OK                |        NULL |     NULL |  10.0000000000000000000 |
| Rounded OK        |        NULL |     NULL |  10.0000000000000000000 |
| Overflow          |        NULL |     NULL |  10.0000000000000000000 |
| OK                |        NULL |     NULL | -10.0000000000000000000 |
| Rounded OK        |        NULL |     NULL | -10.0000000000000000000 |
| Underflow         |        NULL |     NULL | -10.0000000000000000000 |
+-------------------+-------------+----------+-------------------------+

TRUNCATE float_unsigned_example;

-- Disable strict mode
SET sql_mode=(SELECT REPLACE(@@sql_mode, 'STRICT_TRANS_TABLES', ''));

INSERT INTO float_unsigned_example VALUES
  ('OK', 3.402823466e38, NULL, NULL),
  ('Overflow', 4.0e38, NULL, NULL),
  ('Underflow', -1, NULL, NULL);

Warning (sql 1264): Out of range value for column 'example' at row 2
Warning (sql 1264): Out of range value for column 'example' at row 3

SELECT * FROM float_unsigned_example;

+-------------+------------+-------+---------+
| description | example    | sz6_2 | sz20_19 |
+-------------+------------+-------+---------+
| OK          | 3.40282e38 |  NULL |    NULL |
| Overflow    | 3.40282e38 |  NULL |    NULL |
| Underflow   |          0 |  NULL |    NULL |
+-------------+------------+-------+---------+

FLOAT ZEROFILL

A special type of FLOAT UNSIGNED is FLOAT ZEROFILL, which pads out the values with leading zeros in SELECT results. The number of leading zeros are just enough to pad the field out to the length of the type's field size (counting the decimal point), but the zeros are not included in an expression result or in a UNION SELECT column.

Using FLOAT ZEROFILL works the same way as FLOAT UNSIGNED for most operations except a simple SELECT. For example, with the following test table setup:

CREATE TABLE float_zerofill_example (
  description VARCHAR(20),
  example FLOAT ZEROFILL,
  sz6_2 FLOAT(6,2) ZEROFILL,
  sz20_19 FLOAT(20,19) ZEROFILL
);

SET @pi = 3.1415926535897932384626433832795;

INSERT INTO float_zerofill_example VALUES
  ('Pi', @pi, @pi, @pi),
  ('Series', 1234.567890123, 1234.567890123, 1.234567890123),
  ('Various', 1234567890, 9999.99, 9.9999999999999999999);

SELECT * FROM float_zerofill_example;

+-------------+--------------+---------+------------------------+
| description | example      | sz6_2   | sz20_19                |
+-------------+--------------+---------+------------------------+
| Pi          | 000003.14159 |  003.14 |  3.1415927410125732000 |
| Series      | 000001234.57 | 1234.57 |  1.2345678806304932000 |
| Various     | 001234570000 | 9999.99 | 10.0000000000000000000 |
+-------------+--------------+---------+------------------------+

_{This page is licensed: GPLv2, originally from fill_help_tables.sql}