Raises a number to a power.
POWER(<number>, <exponent>)
<number> ^ <exponent>
POWER(2, 10) gives 1024. The exponent can be fractional or negative: POWER(9, 0.5) is the same as a square root, and POWER(2, -1) is a half. For the root itself there's a shorter function, SQRT.
| power_of_two | square_root | half |
|---|---|---|
| 1024 | 3.0000000000000000 | 0.5 |
| after_3_years |
|---|
| 1157.63 |
1000 at 5% a year: every year the amount is multiplied by 1.05, so over three years — by 1.05 to the third power.
| megabytes |
|---|
| 4.76837158203125 |
A megabyte has 1024 · 1024 bytes — that is, 1024 to the second power.
Integer arguments give double precision, and with it — rounding errors: POWER(1.1::float8, 2) gives 1.2100000000000002. With a decimal literal the calculation is done in numeric, and POWER(1.1, 2) gives exactly 1.21.
^ raises to a power too: 2 ^ 10. But 2 ** 10, as in some programming languages, throws operator does not exist.
A negative number raised to a fractional power is an error: POWER(-8, 1.0 / 3) says a negative number raised to a non-integer power yields a complex result.