module Number: BatNumber
exception Overflow
Arithmetic overflow.
This kind of exception is raised by "safe" numeric modules whenever the number which should be returned is too large to be represented.
Non-"safe" numeric modules will return a result which depends on
the internal representation. For instance, with module Int
,
max_num + 1
returns min_num
. By opposition, with module
Safe_int
, max_num + 1
raises Overflow
.
exception NaN
Not a Number
This kind of exception is raised by "safe" modules whenever the number which should be returned is not a number.
For instance, with module Safe_float
, 0.0 / 0.0
raises NaN
.
By opposition, with module Float
, 0.0 / 0.0
does not interrupt
computation and returns a special value nan
.
type 'a
numeric = {
|
zero : |
|
one : |
|
neg : |
|
succ : |
|
pred : |
|
abs : |
|
add : |
|
sub : |
|
mul : |
|
div : |
|
modulo : |
|
pow : |
|
compare : |
|
of_int : |
|
to_int : |
|
of_string : |
|
to_string : |
|
of_float : |
|
to_float : |
}
The smallest set of operations supported by every set of numbers.
This is presented as record to permit lightweight typeclass-style computation.
module type Infix =sig
..end
The infix operators available with any type of numbers
module type Compare =sig
..end
And if you are ready to drop generic comparison operators, then you can open this one as well
module type RefOps =sig
..end
Reference operators ala C.
module type Numeric =sig
..end
The full set of operations of a type of numbers
module type Bounded =sig
..end
module type Discrete =sig
..end