float

(float x)

Answers a number widened to a double: there is no single-precision value, so (float 1/3) is 0.3333333333333333 where the oracle's is 0.33333334. A value past the float range (and an infinity) signals Value out of range for float: ..., like the oracle; a non-number signals. As a value a one-argument function.