Converts this value to Float.
Rounding may lose precision. Values too large for Float become signed infinity; very small values may round to signed zero. NaN and infinities are retained.
the floating-point representation.