Ah - I didn't realise that. I was trying to avoid the overhead, as I
understood it, of both converting the parameters to Float/Double
objects, and then of checking for overflow.
So '*' and '+' don't do overflow checking - do they need to convert
primitives to/from objects? Specifically, when I call:
(loop [x (double x) y (double y)]
... do maths to x and y
(recur x y)
)
Is it going to have to convert x and y between 'double' and 'Double'
at all, or will it just work with primitive types???
- Korny
On Sat, Feb 7, 2009 at 6:29 PM, Mark H. <mark.hoem...@gmail.com> wrote:
>
> On Feb 6, 7:14 pm, Korny Sietsma <ko...@sietsma.com> wrote:
>> Hi folks - I was trying the example posted below, and I discovered a
>> slight snag - unchecked-* methods don't exist for doubles or floats!
>> (and if you call (unchecked-multiply 1.2 3.4) you get "No matching
>> method found"... which caused some confusion :) )
>>
>> Is this something that might change in the future? Or would it be
>> easy for a clojure newbie to implement them himself? :-}
>
> The question is whether it makes sense to have "unchecked_*" methods
> for floats and doubles. The point of "unchecked_*" methods is to
> declare your willingness to get nonsense answers with fixed-width
> integer operations. Floating-point operations already have semantics
> that make sense for conditions like overflow, and (unlike integer
> operations in Clojure) they don't switch to "bigfloats" on overflow.
>
> If you'd like to have the methods just for completeness' sake, it
> doesn't look too hard to add them. Check out src/jvm/clojure/lang/
> Numbers.java and look for static methods of the form "unchecked_add",
> "unchecked_multiply", etc. For each "unchecked_*" method name and for
> each data type, there should be one such method: e.g.,
>
> static public int unchecked_add(int x, int y){
> return x + y;
> }
>
> static public long unchecked_add(long x, long y){
> return x + y;
> }
>
> mfh
> >
>
--
Kornelis Sietsma korny at my surname dot com
kornys on gmail, twitter, facebook, etc.
"Every jumbled pile of person has a thinking part
that wonders what the part that isn't thinking
isn't thinking of"
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Clojure" group.
To post to this group, send email to clojure@googlegroups.com
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
-~----------~----~----~----~------~----~------~--~---
