jason-users Mailing List for jason

Hi Jomi,

Sorry, I did not see the demo about functions. Now it works, I only need to 
register the function such as:

{ register_function("tools.StdDev") } 

By the way, have you considered including the standard deviation as a built-in 
Jason function?

Best,
Francisco

A Dimarts, 31 d'agost de 2010 08:48:39, Jomi Hubner va escriure:
> Hi Francisco,
> 
> functions are not so "easy" to include in a project as internal actions,
>  you need to inform the architecture about them. There is demo in the
>  folder demo/function that, I guess, can help you.
> 
> Best,
> 
> Jomi
> 
> On 30/08/2010, at 21:23, Francisco Grimaldo Moreno wrote:
> > Dear all,
> >
> > I would like to create a new arithmetic function such as the ones in the
> > package jason.functions. For instance, from the "average" function I have
> > writen a "std_dev" function computing the standard deviation of a list of
> > numbers (see the attached file to be put under the tools directory). Then
> > I try to invoke it such as:
> >
> > .println("Std. Deviation", tools.std_dev(L));
> >
> > However, if I do it this way I get the following warning message:
> >
> > [aslparser] [test.asl:14] warning: The internal action class for
> > 'tools.std_dev(L)' was not loaded! Error:
> > java.lang.ClassNotFoundException: tools.std_dev
> >
> > I have reprogrammed this arithmetic function as a conventional internal
> > action and it works ok but then the programmer needs two steps to get to
> > use the standard deviation.
> >
> > Am I doing something wrong? Can the programmers define new arithmetic
> > functions?
> >
> > Best,
> > Francisco Grimaldo
> 

-- 
-------------------------------------------
Dr. Francisco Grimaldo Moreno (PhD)
(Id. GnuPG: 0xBAEC75A7)

Departament d'Informàtica
Escola Tècnica Superior d'Enginyeria (ETSE)
Universitat de València
Avda. Vicent Andrés Estellés, s/n
46100 Burjassot-València (SPAIN)

Tel: +34 96 354 44 87 (+34 96 354 45 65)
Fax: +34 96 354 47 68
Email: fra...@uv...
Web: http://www.uv.es/grimo
-------------------------------------------

Hi Francisco,

functions are not so "easy" to include in a project as internal actions, you need to inform the architecture about them. There is demo in the folder demo/function that, I guess, can help you.

Best,

Jomi

On 30/08/2010, at 21:23, Francisco Grimaldo Moreno wrote:

> Dear all,
> 
> I would like to create a new arithmetic function such as the ones in the 
> package jason.functions. For instance, from the "average" function I have 
> writen a "std_dev" function computing the standard deviation of a list of 
> numbers (see the attached file to be put under the tools directory). Then I 
> try to invoke it such as: 
> 
> .println("Std. Deviation", tools.std_dev(L));
> 
> However, if I do it this way I get the following warning message: 
> 
> [aslparser] [test.asl:14] warning: The internal action class for 
> 'tools.std_dev(L)' was not loaded! Error: java.lang.ClassNotFoundException: 
> tools.std_dev
> 
> I have reprogrammed this arithmetic function as a conventional internal action 
> and it works ok but then the programmer needs two steps to get to use the 
> standard deviation. 
> 
> Am I doing something wrong? Can the programmers define new arithmetic 
> functions? 
> 
> Best,
> Francisco Grimaldo
> 
> -- 
> -------------------------------------------
> Dr. Francisco Grimaldo Moreno (PhD)
> (Id. GnuPG: 0xBAEC75A7)
> 
> Departament d'Informàtica
> Escola Tècnica Superior d'Enginyeria (ETSE)
> Universitat de València
> Avda. Vicent Andrés Estellés, s/n
> 46100 Burjassot-València (SPAIN)
> 
> Tel: +34 96 354 44 87 (+34 96 354 45 65)
> Fax: +34 96 354 47 68
> Email: fra...@uv...
> Web: http://www.uv.es/grimo
> -------------------------------------------
> <StdDev.java>------------------------------------------------------------------------------
> This SF.net Dev2Dev email is sponsored by:
> 
> Show off your parallel programming skills.
> Enter the Intel(R) Threading Challenge 2010.
> http://p.sf.net/sfu/intel-thread-sfd_______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Dear all,

I would like to create a new arithmetic function such as the ones in the 
package jason.functions. For instance, from the "average" function I have 
writen a "std_dev" function computing the standard deviation of a list of 
numbers (see the attached file to be put under the tools directory). Then I 
try to invoke it such as: 

.println("Std. Deviation", tools.std_dev(L));

However, if I do it this way I get the following warning message: 

[aslparser] [test.asl:14] warning: The internal action class for 
'tools.std_dev(L)' was not loaded! Error: java.lang.ClassNotFoundException: 
tools.std_dev

I have reprogrammed this arithmetic function as a conventional internal action 
and it works ok but then the programmer needs two steps to get to use the 
standard deviation. 

Am I doing something wrong? Can the programmers define new arithmetic 
functions? 

Best,
Francisco Grimaldo

-- 
-------------------------------------------
Dr. Francisco Grimaldo Moreno (PhD)
(Id. GnuPG: 0xBAEC75A7)

Departament d'Informàtica
Escola Tècnica Superior d'Enginyeria (ETSE)
Universitat de València
Avda. Vicent Andrés Estellés, s/n
46100 Burjassot-València (SPAIN)

Tel: +34 96 354 44 87 (+34 96 354 45 65)
Fax: +34 96 354 47 68
Email: fra...@uv...
Web: http://www.uv.es/grimo
-------------------------------------------

You can take a look on the demos/gui (specially gui2) for some ideas.

Jomi

On 29/08/2010, at 13:42, SRIshanu wrote:

> Hi all,
> 
> I created a UI form my agent based application using java.
> it requires the user to enter some values which will be used to query a beliefset. When the user enters the data and clicks a button I want that event to be perceived by the agent and read the data sent through the form.
> 
> How do I capture the events happening on the form (button click) to pass the data on the form to my agent to execute a relevant plan.??
> 
> Thanks in advance!
> 
> -- 
> ශ්‍රීශානු | SRIshanu
> http://twitter.com/srishanu
> ------------------------------------------------------------------------------
> Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
> Be part of this innovative community and reach millions of netbook users 
> worldwide. Take advantage of special opportunities to increase revenue and 
> speed time-to-market. Join now, and jumpstart your future.
> http://p.sf.net/sfu/intel-atom-d2d_______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi all,

I created a UI form my agent based application using java.
it requires the user to enter some values which will be used to query a
beliefset. When the user enters the data and clicks a button I want that
event to be perceived by the agent and read the data sent through the form.

How do I capture the events happening on the form (button click) to pass the
data on the form to my agent to execute a relevant plan.??

Thanks in advance!

-- 
ශ්‍රීශානු | SRIshanu
http://twitter.com/srishanu

  Hi Dimitrios,

Sorry it took us so long to reply, both Jomi and I were giving a course 
together with other colleagues, on Multi-Agent Programming, at the 
European Agent Systems Summer School.

Just complementing what Jomi said...

It's difficult to tell if you've got the right model for time update for 
your simulation with what you've told us about it but if you need time 
change to occur both in the environment and through the initiative of 
individual agents, then it is a bit more complicated than usual. Just so 
you understand why you're getting old unnecessary beliefs about time 
updates, this is precisely because they come as information from agents 
rather then perception of the environment. If you keep only one percept 
time_update(_) in your environment model, when Jason does belief update 
it will only keep the latest one and the ones are removed (this is the 
normal working of beliefs that derive specifically from percepts). Now, 
an agent does not forget what it has been told by other agents unless it 
explicitly deletes that belief with the right annotation or the agent 
tell it not to believe so anymore (an you're not overriding the default 
behaviour of the performatives). The .abolish action will delete all 
beliefs with the given pattern, so that might indeed help you in this case.

Cheers,

Rafael

On 29/08/2010 15:27, Jomi Fred Hubner wrote:
> Hi,
>
> for the 3rd question, yes. It could be the case that more than on intention is created the handle the event +time_update(...), since concurrently the environment and the agents can produce them.
>
> I guess the easiest way to solve your problem is to define the plan as atomic so that there is no interleaving between those intentions:
>
> @someLabel[atomic]
> +time_update(Time) : true
> 	<-	-+curtime(Time);			
>                   ... other things here
>
>
> HTH
>
> Jomi
>
> On 19/08/2010, at 18:29, Dimitrios Traskas wrote:
>
>> Hi there,
>>
>> I 've been having a few issues in the last couple of weeks and thought maybe you could help. I will start with the simpler of the problems, abolishing beliefs. I have a discrete event simulation and a SimModel class that holds all the necessary information for my model as well as all the events. My version of Jason is not the latest one but the one before.
>>
>> In my simulation time advances in 2 ways, either with addPercept from the environment or via the instantiation of a new agent who announces a time update with a broadcast. When time advances I have the following belief added:
>>
>> +time_update(Time) : true
>> 	<-	-+curtime(Time);			
>>                   ... other things here
>>
>> The simulation can have thousands of agents and thousands of time_updates.
>>
>> Questions:
>>
>> 1. Do I need to abolish(time_update(_)) every time I get the belief added?
>> 2. If I do need to abolish the belief then why do I notice beliefs left in the agents when I inspect them with Debug?
>> 3. When I have a lot of beliefs of time_update is there a case that achievements goals will trigger more than once for each belief?
>>
>> Finally any recommendations for handling time updates better?
>>
>> Regards
>> Dimitris
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by
>>
>> Make an app they can't live without
>> Enter the BlackBerry Developer Challenge
>> http://p.sf.net/sfu/RIM-dev2dev _______________________________________________
>> Jason-users mailing list
>> Jas...@li...
>> https://lists.sourceforge.net/lists/listinfo/jason-users

Hi again,

Yes, since you have the environment and some agents updating the time 
concurrently, you probably need to synchronise the access to that code 
you included below.

And yes, in the environment, if the agents access other data structures 
apart from the lists of percepts that are handled by Jason, you need to 
synchronise the access to them. We cannot simplify things by making the 
whole action execution in the environment completely synchronised 
because that makes the whole system extremely inefficient.

Best,

Rafael


On 19/08/2010 23:40, Dimitrios Traskas wrote:
> Hello again,
>
> my 2nd question is in regards with some threading problems I
> experienced. Basically my model uses a simulation agent but also an
> instance of a model class. In the model class I handle certain events
> and I noticed that two threads-agents will attempt to access the same
> structures at the same time. I have used the synchronized keyword to
> solve this problem but I am not quite sure if it's the right approach.
>
> Additionally I have a user interface where I control the simulation
> from. This was working ok for some smaller scale problems but now I
> tried it with a larger scale model and I get threading exceptions from
> that user interface. The only thing I update there is time in a textbox
> so that I can see where I am. The code is below:
>
> public void updateTime(Date time){
> final String text = df.format(time);
> txtClock.setText(text);
>      }
>
> Do I need to use synchronized for that too?
>
>
> Regards
> Dimitris
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by
>
> Make an app they can't live without
> Enter the BlackBerry Developer Challenge
> http://p.sf.net/sfu/RIM-dev2dev
>
>
>
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

Hi,

for the 3rd question, yes. It could be the case that more than on intention is created the handle the event +time_update(...), since concurrently the environment and the agents can produce them. 

I guess the easiest way to solve your problem is to define the plan as atomic so that there is no interleaving between those intentions:

@someLabel[atomic]
+time_update(Time) : true
	<-	-+curtime(Time);			
                 ... other things here


HTH

Jomi

On 19/08/2010, at 18:29, Dimitrios Traskas wrote:

> Hi there,
> 
> I 've been having a few issues in the last couple of weeks and thought maybe you could help. I will start with the simpler of the problems, abolishing beliefs. I have a discrete event simulation and a SimModel class that holds all the necessary information for my model as well as all the events. My version of Jason is not the latest one but the one before.
> 
> In my simulation time advances in 2 ways, either with addPercept from the environment or via the instantiation of a new agent who announces a time update with a broadcast. When time advances I have the following belief added:
> 
> +time_update(Time) : true
> 	<-	-+curtime(Time);			
>                  ... other things here
> 
> The simulation can have thousands of agents and thousands of time_updates. 
> 
> Questions:
> 
> 1. Do I need to abolish(time_update(_)) every time I get the belief added? 
> 2. If I do need to abolish the belief then why do I notice beliefs left in the agents when I inspect them with Debug?
> 3. When I have a lot of beliefs of time_update is there a case that achievements goals will trigger more than once for each belief?
> 
> Finally any recommendations for handling time updates better?
> 
> Regards
> Dimitris
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by 
> 
> Make an app they can't live without
> Enter the BlackBerry Developer Challenge
> http://p.sf.net/sfu/RIM-dev2dev _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi,

as in Prolog, you can write any logical expression for the 2nd arg of findall, for instance:

	.findall(D, bel(D,R) & D > "08/14/10",L)

Jomi

On 21/08/2010, at 10:57, sachini weerawardhana wrote:

> Hi.
> 
> A question on beliefbase query.
> 
> Suppose this is an agent's beliefbase. Represented in an sql database table.
> 
> 
> 
> date
> Reading
> 
> 
> 08/14/10 03:21 AM	91.95
> 
> 
> 08/14/10 03:21 AM	9.07
> 
> 
> 08/14/10 03:22 AM	14.04
> 
> 
> 08/19/10 10:49 AM	80.55
> 
> 
> 08/19/10 10:53 AM	80.32
> 
> 
> 08/19/10 10:54 AM	88.31
> 
> 
> 08/19/10 10:55 AM	82.73
> 
> 
> 08/19/10 10:55 AM	8.44
> 
> 
> 08/19/10 11:01 AM	47.47
> 
> 
> 08/19/10 11:02 AM	55.17
> 
> 
> 08/19/10 11:02 AM	76.6
> 
> 
> 08/19/10 11:02 AM	70.24
> 
> 
> 08/19/10 11:03 AM	3.68
> 
> 
> 08/19/10 11:29 AM	20.7
> 
> 
> 08/20/10 12:10 AM	22.6
> 
> 
> 08/20/10 12:17 AM	21.96
> 
> 
> 08/20/10 12:34 AM	84.5
> 
> 
> 08/20/10 12:34 AM	5.71
> 
> 
> 08/20/10 12:35 AM	76.27
> 
> 
> 08/20/10 12:36 AM	22.41
> How is it possible to query the beliefbase with .findall() so that it returns only the items corresponding to a given filtering criteria?
> e.g. Find all readings between 08/19/2010 and 08/20/2010
> 
> 
> Thanks
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by 
> 
> Make an app they can't live without
> Enter the BlackBerry Developer Challenge
> http://p.sf.net/sfu/RIM-dev2dev _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi Sachine,

you shouldn't "tell" a list to another agent, the content has to be a predicate, for instance:

	.send(authorizer, tell, all_values(L))

I hope that it solves your problem

Bellow some comments on your code: 


On 25/08/2010, at 04:16, sachini weerawardhana wrote:


> ==========logger.asl=================
> +rainfall(D, R, LOC).
> +riverheight(D, R, LOC).

These plans are not necessary, since they do nothing. You can remove them.

> 
> @p1[atomic]
> +!query_rainfall <-.findall(R, rainfall(_,_,R), LF).

I do not understand this plan. Why to find all beliefs and does nothing with them?

Regards,

Jomi

> 
> @p2[atomic]
> +!query_riverheight <-.findall(R, riverheight(_,_,R), LH).
> 
> 
> ==============================================
> 
> I used [atomic] annotation because the logger agent keeps on adding fresh values to its belief base (+rainfall, +riverheight). I wanted to stop values from being added to the beliefbase while querying took place.
> 
> Please help
> 
> Sachini
> 
> ------------------------------------------------------------------------------
> Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
> Be part of this innovative community and reach millions of netbook users 
> worldwide. Take advantage of special opportunities to increase revenue and 
> speed time-to-market. Join now, and jumpstart your future.
> http://p.sf.net/sfu/intel-atom-d2d_______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

sorry about the 2X post.
I am still stuck on this requirement and seeing no way out :(

can someone please share his/her 2cents :)

Sachini



Hi All,

I am in need of some help (again) with regard to info communication between
agents.

There are two agents involved in this particular scenario.

   - Logger Agent (logs info)
   - Authorizer Agent (instructs the logger agent to send for information he
   needs, periodically)


The logger agent, in his beliefbase maintains two database tables. The
authorizer agent within certain intervals instructs the logger agent to
achieve its data querying goal.
upon querying a list of records is generated. This list has to be sent to
the authorizer so that he can use the values to plot a graph.

I do not know how a list of items can be sent from one agent to the other. I
used the .send(authorizer, tell, L) but it didnt appear to be working.

Below are my agent source code for Authorizer and logger agents.

==========Authorizer.asl==============

!fetchraindata.
!fetchriverdata.

+!fetchraindata <-.send(datalogger, achieve, query_rainfall);
                  .wait(20000);
                  !!fetchraindata.

+!fetchriverdata <-.send(datalogger, achieve, query_riverheight);
                   .wait(20000);
                   !!fetchriverdata.

==========logger.asl=================
+rainfall(D, R, LOC).
+riverheight(D, R, LOC).

@p1[atomic]
+!query_rainfall <-.findall(R, rainfall(_,_,R), LF).

@p2[atomic]
+!query_riverheight <-.findall(R, riverheight(_,_,R), LH).


==============================================

I used [atomic] annotation because the logger agent keeps on adding fresh
values to its belief base (+rainfall, +riverheight). I wanted to stop values
from being added to the beliefbase while querying took place.

Please help

Sachini

Hi.

A question on beliefbase query.

Suppose this is an agent's beliefbase. Represented in an sql database table.



date
Reading


08/14/10 03:21 AM 91.95

08/14/10 03:21 AM 9.07

08/14/10 03:22 AM 14.04

08/19/10 10:49 AM 80.55

08/19/10 10:53 AM 80.32

08/19/10 10:54 AM 88.31

08/19/10 10:55 AM 82.73

08/19/10 10:55 AM 8.44

08/19/10 11:01 AM 47.47

08/19/10 11:02 AM 55.17

08/19/10 11:02 AM 76.6

08/19/10 11:02 AM 70.24

08/19/10 11:03 AM 3.68

08/19/10 11:29 AM 20.7

08/20/10 12:10 AM 22.6

08/20/10 12:17 AM 21.96

08/20/10 12:34 AM 84.5

08/20/10 12:34 AM 5.71

08/20/10 12:35 AM 76.27

08/20/10 12:36 AM 22.41
How is it possible to query the beliefbase with .findall() so that it
returns only the items corresponding to a given filtering criteria?
e.g. Find all readings between 08/19/2010 and 08/20/2010


Thanks

Hi All,

I am in need of some help (again) with regard to info communication between
agents.

There are two agents involved in this particular scenario.

   - Logger Agent (logs info)
   - Authorizer Agent (instructs the logger agent to send for information he
   needs, periodically)


The logger agent, in his beliefbase maintains two database tables. The
authorizer agent within certain intervals instructs the logger agent to
achieve its data querying goal.
upon querying a list of records is generated. This list has to be sent to
the authorizer so that he can use the values to plot a graph.

I do not know how a list of items can be sent from one agent to the other. I
used the .send(authorizer, tell, L) but it didnt appear to be working.

Below are my agent source code for Authorizer and logger agents.

==========Authorizer.asl==============

!fetchraindata.
!fetchriverdata.

+!fetchraindata <-.send(datalogger, achieve, query_rainfall);
                  .wait(20000);
                  !!fetchraindata.

+!fetchriverdata <-.send(datalogger, achieve, query_riverheight);
                   .wait(20000);
                   !!fetchriverdata.

==========logger.asl=================
+rainfall(D, R, LOC).
+riverheight(D, R, LOC).

@p1[atomic]
+!query_rainfall <-.findall(R, rainfall(_,_,R), LF).

@p2[atomic]
+!query_riverheight <-.findall(R, riverheight(_,_,R), LH).


==============================================

I used [atomic] annotation because the logger agent keeps on adding fresh
values to its belief base (+rainfall, +riverheight). I wanted to stop values
from being added to the beliefbase while querying took place.

Please help

Sachini

Hello again,

my 2nd question is in regards with some threading problems I experienced.
Basically my model uses a simulation agent but also an instance of a model
class. In the model class I handle certain events and I noticed that two
threads-agents will attempt to access the same structures at the same time.
I have used the synchronized keyword to solve this problem but I am not
quite sure if it's the right approach.

Additionally I have a user interface where I control the simulation from.
This was working ok for some smaller scale problems but now I tried it with
a larger scale model and I get threading exceptions from that user
interface. The only thing I update there is time in a textbox so that I can
see where I am. The code is below:

public void updateTime(Date time){
     final String text = df.format(time);
     txtClock.setText(text);
    }

Do I need to use synchronized for that too?


Regards
Dimitris

Hi there,

I 've been having a few issues in the last couple of weeks and thought maybe
you could help. I will start with the simpler of the problems, abolishing
beliefs. I have a discrete event simulation and a SimModel class that holds
all the necessary information for my model as well as all the events. My
version of Jason is not the latest one but the one before.

In my simulation time advances in 2 ways, either with addPercept from the
environment or via the instantiation of a new agent who announces a time
update with a broadcast. When time advances I have the following belief
added:

+time_update(Time) : true
<- -+curtime(Time);
                 ... other things here

The simulation can have thousands of agents and thousands of time_updates.

Questions:

1. Do I need to abolish(time_update(_)) every time I get the belief added?
2. If I do need to abolish the belief then why do I notice beliefs left in
the agents when I inspect them with Debug?
3. When I have a lot of beliefs of time_update is there a case that
achievements goals will trigger more than once for each belief?

Finally any recommendations for handling time updates better?

Regards
Dimitris

Hi Joerg,

We couldn't immediately identify what is the problem in your code and 
this is unlikely to be a bug in Jason because other users don't report 
the problem. So I'm afraid you'll have to debug more your code. 
Debugging a distributed system is very hard, but I'm afraid there's 
nothing we can help with, with the information you've given us.

Best,

Rafael

Hi Sachini,

good questions! I'll try to answer them:

On 11/08/2010, at 05:51, sachini weerawardhana wrote:

> Hi all,
> 
> I have a few questions to be solved with regard to belief persistence.
> 
> My system has two agents, a reader agent and a logger agent. Reader agent sends data to the logger periodically, to be saved in a database.
> The database has one table called reading, which has 3 fields; date (data type timestamp), location (data type varchar) and reading(data type decimal) 
> 
> The questions I have are:
> 	• At the moment I am not sure how to record the time the logger receives the message from reader. I put the SQL data type timestamp (although other SQL data types like datetime, time, date are also there) for easy querying purposes. I dont know how to send a correct time stamp value from my jason code to be inserted in the table. How do I do that?

you need to "wrap" the timestamp in a structure with the format "timestamp(Y,M,D,H,M,S)". For instance to add a new line in the table
	+reading(timestamp(2010,8,11,9,55,12),"location x", 23.1); 



> 	• To see if my code works correctly I temporarily removed the timestamp column and tried to record the reading and the location only in the table. This works but a record is added to the table only once (the first reading only). Since the reader repeatedly sends different values after 5s interval I need all the readings to be stored in the table with time at which it was recorded. Can you please show me where I have gone wrong in my code?

the belief base is considered as a set (believe or not). To avoid the above effect, you need to force that each entry in the log is different (I guess the timestamp will do that). 

> 	• When a table has 3 columns is it a must that I provide values corresponding to all columns from my jason code? If i omit a value would it be stored with the default value assigned for that field (or null) like in normal sql?

(I guess) you need to provide all values. 

> Is it not possible to select the fields I want values to be set from my agent code?

No. But you can take a look at the JDBC custom BB (class JDBCPersistentBB in src/jason/bb), extend this class for your needs and use it instead. (for instance, see the getInsert method that creates the SQL insert command for the jason + operator).

HTH,

Jomi

ps.: you can use internal action .time and .date to obtain the current time to build the timestamp.

> 
> I am sending my code herewith.
> 
> Reader code (part)
> ===============
> +!send_reading <- .random(R); 
>                 .send(datalogger, tell, reading("location A", R));
>                 .wait(5000);
>                 !!send_reading.
> 
> Datalogger Code
> =============
> !+reading(LOC, R).
> 
> System.mas2j
> ===========
> 
>  infrastructure: Centralised
> 
>    agents:
>        reader;
>        
>        datalogger beliefBaseClass jason.bb.JDBCPersistentBB("com.mysql.jdbc.Driver","jdbc:mysql://localhost/datalogger","root","sa","[reading(3, reading, columns(date(timestamp),location(varchar(45)),reading(decimal(2))))]");
> 
> Apologies for the long email
> 
> Many thanks in advance.
> Sachini
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by 
> 
> Make an app they can't live without
> Enter the BlackBerry Developer Challenge
> http://p.sf.net/sfu/RIM-dev2dev _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi all,

I have a few questions to be solved with regard to belief persistence.

My system has two agents, a reader agent and a logger agent. Reader agent
sends data to the logger periodically, to be saved in a database.
The database has one table called reading, which has 3 fields; date (data
type timestamp), location (data type varchar) and reading(data type decimal)


The questions I have are:

   1. At the moment I am not sure how to record the time the logger receives
   the message from reader. I put the SQL data type timestamp (although other
   SQL data types like datetime, time, date are also there) for easy querying
   purposes. I dont know how to send a correct time stamp value from my jason
   code to be inserted in the table. How do I do that?
   2. To see if my code works correctly I temporarily removed the timestamp
   column and tried to record the reading and the location only in the table.
   This works but a record is added to the table only once (the first reading
   only). Since the reader repeatedly sends different values after 5s interval
   I need all the readings to be stored in the table with time at which it was
   recorded. Can you please show me where I have gone wrong in my code?
   3. When a table has 3 columns is it a must that I provide values
   corresponding to all columns from my jason code? If i omit a value would it
   be stored with the default value assigned for that field (or null) like in
   normal sql? Is it not possible to select the fields I want values to be set
   from my agent code?


I am sending my code herewith.

Reader code (part)
===============
+!send_reading <- .random(R);
                .send(datalogger, tell, reading("location A", R));
                .wait(5000);
                !!send_reading.

Datalogger Code
=============
!+reading(LOC, R).

System.mas2j
===========

 infrastructure: Centralised

   agents:
       reader;

       datalogger beliefBaseClass
jason.bb.JDBCPersistentBB("com.mysql.jdbc.Driver","jdbc:mysql://localhost/datalogger","root","sa","[reading(3,
reading,
columns(date(timestamp),location(varchar(45)),reading(decimal(2))))]");

Apologies for the long email

Many thanks in advance.
Sachini