|
« Return to Thread: Re: JESS: getting a handle on fact in LHS
Re: JESS: getting a handle on fact in LHS
by Ernest Friedman-Hill
::
Rate this Message:
Using a query to find the existing fact is one way to do it; if you're
not seeing the actual fact in the slot afterwards, it's probably just
a coding error in using the query interface. For a query like
(defquery find-some-fact
(declare (variables ?a ?b))
?f <- (some-fact (a ?a) (b ?b)))
Then you'd get the fact, if it exists, like
(bind ?result (run-query* find-some-fact A B))
(if (?result next) then
(bind ?fact (?result get ?f))
else
(bind ?fact (assert (some-fact (a A) (b B)))))
You could put all that into a deffunction so it wouldn't detract from
the flow of your other code.
An alternative way to do all this would be to have two versions of the
original rule: one with a "not" pattern for the sub-fact and one with
a positive pattern. The "not" one should create the subfact, while the
other can directly use the matched fact.
On Jun 29, 2009, at 7:12 AM, axelrad wrote:
>
> Hello,
>
> I have a template that has a fact as a slot.
>
> ex)
>
> (deftemplate mainfact
> (slot subfact (type OBEJCT))
> (slot index))
>
> (deftemplate subfact
> (slot name)
> (slot content))
>
> I have a rule that instantiates (assert) the mainfact and subfact if
> my
> condition is met. My problem is that if the subfact exists already
> it gets
> asserted to false. What I need is to get a handle on the fact if it
> exists
> and assert the mainfact with the existing subfact and a new current
> index.
> What I do now is along the lines of:
>
> assert mainfact (subfact xxx) (index n)
>
> works great if the subfact doesn't exist yet: mainfact xxx n
> but if the subfact exists my result is: mainfact (false) (n)
>
> what i need is : mainfact (existing fact) (n)
>
> I need to compare the subfacts for when they are the same and keep
> the one
> with the higher index value. I've tried doing a query and putting
> that value
> in but i get a jess query object instead of a fact object, and later
> when i
> need to compare the values they are not considered the same object,
> even if
> the slot values are correct, so I don't get the correct results when
> trying
> to keep the mainfact with higher index value.
>
> I can test if the fact exists already, but how can I get a handle on
> it to
> place that value in the slot?
>
> Thanks in advance for any help!
>
> Monica
> --
> View this message in context: http://www.nabble.com/getting-a-handle-on-fact-in-LHS-tp24252267p24252267.html
> Sent from the Jess mailing list archive at Nabble.com.
>
>
>
> --------------------------------------------------------------------
> To unsubscribe, send the words 'unsubscribe jess-users
> you@...'
> in the BODY of a message to majordomo@..., NOT to the list
> (use your own address!) List problems? Notify owner-jess-users@...
> .
> --------------------------------------------------------------------
---------------------------------------------------------
Ernest Friedman-Hill
Informatics & Decision Sciences Phone: (925) 294-2154
Sandia National Labs
PO Box 969, MS 9012 ejfried@...
Livermore, CA 94550 http://www.jessrules.com
--------------------------------------------------------------------
To unsubscribe, send the words 'unsubscribe jess-users you@...'
in the BODY of a message to majordomo@..., NOT to the list
(use your own address!) List problems? Notify owner-jess-users@....
--------------------------------------------------------------------
not seeing the actual fact in the slot afterwards, it's probably just
a coding error in using the query interface. For a query like
(defquery find-some-fact
(declare (variables ?a ?b))
?f <- (some-fact (a ?a) (b ?b)))
Then you'd get the fact, if it exists, like
(bind ?result (run-query* find-some-fact A B))
(if (?result next) then
(bind ?fact (?result get ?f))
else
(bind ?fact (assert (some-fact (a A) (b B)))))
You could put all that into a deffunction so it wouldn't detract from
the flow of your other code.
An alternative way to do all this would be to have two versions of the
original rule: one with a "not" pattern for the sub-fact and one with
a positive pattern. The "not" one should create the subfact, while the
other can directly use the matched fact.
On Jun 29, 2009, at 7:12 AM, axelrad wrote:
>
> Hello,
>
> I have a template that has a fact as a slot.
>
> ex)
>
> (deftemplate mainfact
> (slot subfact (type OBEJCT))
> (slot index))
>
> (deftemplate subfact
> (slot name)
> (slot content))
>
> I have a rule that instantiates (assert) the mainfact and subfact if
> my
> condition is met. My problem is that if the subfact exists already
> it gets
> asserted to false. What I need is to get a handle on the fact if it
> exists
> and assert the mainfact with the existing subfact and a new current
> index.
> What I do now is along the lines of:
>
> assert mainfact (subfact xxx) (index n)
>
> works great if the subfact doesn't exist yet: mainfact xxx n
> but if the subfact exists my result is: mainfact (false) (n)
>
> what i need is : mainfact (existing fact) (n)
>
> I need to compare the subfacts for when they are the same and keep
> the one
> with the higher index value. I've tried doing a query and putting
> that value
> in but i get a jess query object instead of a fact object, and later
> when i
> need to compare the values they are not considered the same object,
> even if
> the slot values are correct, so I don't get the correct results when
> trying
> to keep the mainfact with higher index value.
>
> I can test if the fact exists already, but how can I get a handle on
> it to
> place that value in the slot?
>
> Thanks in advance for any help!
>
> Monica
> --
> View this message in context: http://www.nabble.com/getting-a-handle-on-fact-in-LHS-tp24252267p24252267.html
> Sent from the Jess mailing list archive at Nabble.com.
>
>
>
> --------------------------------------------------------------------
> To unsubscribe, send the words 'unsubscribe jess-users
> you@...'
> in the BODY of a message to majordomo@..., NOT to the list
> (use your own address!) List problems? Notify owner-jess-users@...
> .
> --------------------------------------------------------------------
---------------------------------------------------------
Ernest Friedman-Hill
Informatics & Decision Sciences Phone: (925) 294-2154
Sandia National Labs
PO Box 969, MS 9012 ejfried@...
Livermore, CA 94550 http://www.jessrules.com
--------------------------------------------------------------------
To unsubscribe, send the words 'unsubscribe jess-users you@...'
in the BODY of a message to majordomo@..., NOT to the list
(use your own address!) List problems? Notify owner-jess-users@....
--------------------------------------------------------------------
« Return to Thread: Re: JESS: getting a handle on fact in LHS
| Free embeddable forum powered by Nabble | Forum Help |
