Floating point numbers are represented internally in IEEE 754 double-precision format. This internal representation can be accessed for advanced operations and input/output purposes.

float>bits ( x -- n )

double>bits ( x -- n )

bits>float ( n -- x )

bits>double ( n -- x )

Constructing floating point NaNs:

<fp-nan> ( payload -- nan )

Floating point numbers are discrete:

prev-float ( m -- n )

next-float ( m -- n )

Introspection on floating point numbers:

fp-special? ( x -- ? )

fp-nan? ( x -- ? )

fp-qnan? ( x -- ? )

fp-snan? ( x -- ? )

fp-infinity? ( x -- ? )

fp-nan-payload ( x -- bits )

Comparing two floating point numbers for bitwise equality:

fp-bitwise= ( x y -- ? )

See also
NAN: