Re: [rules-users] programmatic manipulation of 4.0 salience?
Scott,
not sure if it fits your problem, but you may also consider using
dynamic saliences. Yesterday I was able to convert 3.x rules to use it
worked very well.
In my case, I have facts of different sizes (some are collections of
other facts, some are individual facts) and in some rules I need to
find first MxN pairs (collection facts on both sides), then Mx1 or 1xN
pairs, and only then 1x1 pairs. In other rules I need to find first
1x1 pairs, then Mx1 or 1xN pairs and only then MxN.
Instead of writting a custom ConflictResolver, what I did was to use
one dynamic salience:
salience (1000 / (c.size * t.size))
for finding smallest pairs first, and
salience (-1000 / (c.size * t.size)
for finding the largest pairs first.
The 1000 is to get around the fact that saliences are integers and not
float/double (afaik)
Hope it helps,
-- yuri
On 8/17/07, Manjax23 <manjunatha.n(a)coreobjects.com> wrote:
Salience salience = new SalienceInteger( Integer.parseInt( salienceText ) );
rule.setSalience( salience );
Cheers,
manjax23
Scott Reed-4 wrote:
>
> My 3.1 app has a mechanism that allowed the user to change the salience of
> some rules before running
> them, before loading up WorkingMemory. This was quite simple,
> rule.getSalience() returned a rule's
> int salience value and rule.setSalience(int) set it. Now in 4.0 I see
> salience is no longer just an
> int, but a Salience object with a simple constructor and complicated
> getValue(Tuple,WorkingMemory)
> method to access the int value.
>
> It appears to be still easy to set the salience of a rule:
> rule.setSalience( new
> SalienceInteger(int)) but getting the int salience from a Salience object
> seems to have been removed
> from joe-blow coder's reach.
>
> I would be very grateful if someone would provide me with an example that
> gets the int salience
> associated with a rule. I have no idea where or why I need to get the
> Tuple and WorkingMemory. Can I
> just pass nulls in for those two arguments?
>
> Thanks,
> Scott
> _______________________________________________
> rules-users mailing list
> rules-users(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users
>
>
--
View this message in context:
http://www.nabble.com/programmatic-manipulation-of-4.0-salience--tf428373...
Sent from the drools - user mailing list archive at Nabble.com.
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users