Par_incr.Cutoff
type 'a t =
| Never
Never cut-off the computation
*)| Always
Always cut-off the computation
*)| Phys_equal
Cut off the computation based on physical equality (==)
| Eq of 'a -> 'a -> bool
Cutoff the computation based on the function passed
*)| F of oldval:'a -> newval:'a -> bool
Same functionality as Eq
but the equality function uses named arguments
Defines different computation cutoff strategies
val attach : 'a t -> 'a incremental -> 'a incremental
Cutoff.attach cutoff incr
attaches the given cutoff strategy to the incremental incr
. The default cutoff in most cases is `Phys_equal`. So, this is useful in cases where you want somewhat different cutoff condition. Take floating point related computations for example, you may choose to ignore difference in values within some delta. For such cases, you can use this.