Module Fval.FSource

Sourcetype t
Sourceval of_float : float -> t

fails on NaNs, but allows infinities.

Sourceval to_float : t -> float

fails on NaNs, but allows infinities.

Sourceval compare : t -> t -> int
Sourceval equal : t -> t -> bool

Those functions distinguish -0. and +0.

Sourceval pretty : Format.formatter -> t -> unit
Sourceval pretty_normal : use_hex:bool -> Format.formatter -> t -> unit
Sourceval plus_zero : t
Sourceval is_finite : t -> bool

Returns true if the float is a finite number.

Sourceval next_float : kind -> float -> float

First double strictly above the argument. Must be called on non-NaN floats. Returns +infty on MAX_FLT. Infinities are left unchanged.

Sourceval prev_float : kind -> float -> float

First double strictly below the argument. Must be called on non-NaN floats. Returns -infty on -MAX_FLT. Infinities are left unchanged.