Let's get programmy!

Akonyl · Post by **Akonyl** » May 20th, 2012, 10:43 pm

except that you're obviously leaving out code from that example, since there's only three printouts there and you're listing 4 lines of output :V

ranger · Post by **ranger** » May 20th, 2012, 10:49 pm

^

yeah..assuming the toString method is just returning one line w/o any new line characters, the sample code you're giving is incomplete..

Callid · Post by **Callid** » May 21st, 2012, 5:49 am

Ah, yeah, there used to be more lines because I also tested it with these lines twice for different objects, to see if they'd influence each other (that they did not). So, it was like this, with both second lines sometimes commented out and sometimes not. I just copied to much output

Code: Select all

        GregorianCalendar greg = new GregorianCalendar(1992,7,12,17,30);
        //System.out.println(greg.getTime().toString());
        greg.setTimeZone(TimeZone.getTimeZone("GMT+1"));
        System.out.println(greg.getTime().toString());
        greg.setTimeZone(TimeZone.getTimeZone("GMT-1"));
        System.out.println(greg.getTime().toString());
        
        GregorianCalendar greg2 = new GregorianCalendar(1992,7,12,17,30);
        System.out.println(greg2.getTime().toString());
        greg2.setTimeZone(TimeZone.getTimeZone("GMT+1"));
        System.out.println(greg2.getTime().toString());
        greg2.setTimeZone(TimeZone.getTimeZone("GMT-1"));
        System.out.println(greg2.getTime().toString());

Kogorou · Post by **Kogorou** » May 21st, 2012, 6:04 am

I tried that code too and indeed there seems to something wrong.
I have to debug is a little bit more

Akonyl · Post by **Akonyl** » May 21st, 2012, 11:54 am

ah, that's actually pretty funny then. Although, the problem you posted the second time is the complete opposite of the first problem you posted, and actually makes sense.

but rather than just telling you why (because that doesn't make anyone a better programmer :V), I'll just ask the question: given the two lines of code,

Code: Select all

GregorianCalendar greg = new GregorianCalendar(1992,7,12,17,30);
System.out.println(greg.getTime().toString());

what do you actually expect to happen? You're forcing the program to make a certain large assumption when you call getTime() here :V

Callid · Post by **Callid** » May 21st, 2012, 12:26 pm

Akonyl wrote: ah, that's actually pretty funny then. Although, the problem you posted the second time is the complete opposite of the first problem you posted, and actually makes sense.

but rather than just telling you why (because that doesn't make anyone a better programmer :V), I'll just ask the question: given the two lines of code,
Code: Select all
GregorianCalendar greg = new GregorianCalendar(1992,7,12,17,30);
System.out.println(greg.getTime().toString());
what do you actually expect to happen? You're forcing the program to make a certain large assumption when you call getTime() here :V

Yeah, I assumed that was the reason. Still, it's quite funny that getTime() in fact finalizes the time. I would have expected that that'd need a special method, e.g. ".setFix()" or something

You wouldn't usually expect getTime to do that :x

Akonyl · Post by **Akonyl** » May 21st, 2012, 1:15 pm

It's not finalizing the time, you can still easily add time to the date. :V

c-square · Post by **c-square** » May 21st, 2012, 1:21 pm

Looks like GregorianCalendar has a 'time' member that gets set or cleared based on certain actions. Whether or not this member is set determines if the hour is properly reported after a timezone change. Very strange...

Code: Select all

	        Calendar greg = new GregorianCalendar(1992,7,12,17,20,10);
	        System.out.println("Scenario 1");
	        System.out.println(greg);
	        System.out.println("-- Setting Time Zone --");
	        greg.setTimeZone(TimeZone.getTimeZone("GMT+1"));
	        System.out.println(greg);
	        System.out.println("1) " + greg.getTime());
	        System.out.println("\n-----\n");
	        
	        GregorianCalendar greg2 = new GregorianCalendar(1992,7,12,17,20,10);
	        System.out.println("Scenario 2");
	        System.out.println(greg2);
	        System.out.println("-- Printing Time --");
	        System.out.println("1) " + greg2.getTime());
	        System.out.println(greg2);
	        System.out.println("-- Setting Time Zone --");
	        greg2.setTimeZone(TimeZone.getTimeZone("GMT+1"));
	        System.out.println(greg2);
	        System.out.println("2) " + greg2.getTime());
	        System.out.println("\n-----\n");
	        
	        GregorianCalendar greg3 = new GregorianCalendar(1992,7,12,17,20,10);
	        System.out.println("Scenario 3");
	        System.out.println(greg3);
	        System.out.println("-- Printing Time --");
	        System.out.println("1)" + greg3.getTime());
	        System.out.println(greg3);
	        System.out.println("-- Adding Year --");
	        greg3.add(Calendar.YEAR, 1);
	        System.out.println(greg3);
	        System.out.println("-- Setting Time Zone --");
	        greg3.setTimeZone(TimeZone.getTimeZone("GMT+1"));
	        System.out.println(greg3);
	        System.out.println("2)" + greg3.getTime());
	        System.out.println("\n-----\n");

Code: Select all

Scenario 1
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
-- Setting Time Zone --
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="GMT+01:00",offset=3600000,dstSavings=0,useDaylight=false,transitions=0,lastRule=null],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
1) Wed Aug 12 12:20:10 EDT 1992

-----

Scenario 2
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
-- Printing Time --
1) Wed Aug 12 17:20:10 EDT 1992
java.util.GregorianCalendar[time=713654410000,areFieldsSet=true,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
-- Setting Time Zone --
java.util.GregorianCalendar[time=713654410000,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="GMT+01:00",offset=3600000,dstSavings=0,useDaylight=false,transitions=0,lastRule=null],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
2) Wed Aug 12 17:20:10 EDT 1992

-----

Scenario 3
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
-- Printing Time --
1)Wed Aug 12 17:20:10 EDT 1992
java.util.GregorianCalendar[time=713654410000,areFieldsSet=true,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=?,YEAR=1992,MONTH=7,WEEK_OF_YEAR=?,WEEK_OF_MONTH=?,DAY_OF_MONTH=12,DAY_OF_YEAR=?,DAY_OF_WEEK=?,DAY_OF_WEEK_IN_MONTH=?,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=?,ZONE_OFFSET=?,DST_OFFSET=?]
-- Adding Year --
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=true,lenient=true,zone=sun.util.calendar.ZoneInfo[id="America/New_York",offset=-18000000,dstSavings=3600000,useDaylight=true,transitions=235,lastRule=java.util.SimpleTimeZone[id=America/New_York,offset=-18000000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=3,startMonth=2,startDay=8,startDayOfWeek=1,startTime=7200000,startTimeMode=0,endMode=3,endMonth=10,endDay=1,endDayOfWeek=1,endTime=7200000,endTimeMode=0]],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=1,YEAR=1993,MONTH=7,WEEK_OF_YEAR=33,WEEK_OF_MONTH=3,DAY_OF_MONTH=12,DAY_OF_YEAR=225,DAY_OF_WEEK=4,DAY_OF_WEEK_IN_MONTH=2,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=0,ZONE_OFFSET=-18000000,DST_OFFSET=3600000]
-- Setting Time Zone --
java.util.GregorianCalendar[time=?,areFieldsSet=false,areAllFieldsSet=false,lenient=true,zone=sun.util.calendar.ZoneInfo[id="GMT+01:00",offset=3600000,dstSavings=0,useDaylight=false,transitions=0,lastRule=null],firstDayOfWeek=1,minimalDaysInFirstWeek=1,ERA=1,YEAR=1993,MONTH=7,WEEK_OF_YEAR=33,WEEK_OF_MONTH=3,DAY_OF_MONTH=12,DAY_OF_YEAR=225,DAY_OF_WEEK=4,DAY_OF_WEEK_IN_MONTH=2,AM_PM=1,HOUR=5,HOUR_OF_DAY=17,MINUTE=20,SECOND=10,MILLISECOND=0,ZONE_OFFSET=-18000000,DST_OFFSET=3600000]
2)Thu Aug 12 12:20:10 EDT 1993

-----

Akonyl · Post by **Akonyl** » May 21st, 2012, 1:40 pm

it would be stranger if it didn't have a time member

The issue is that when you construct a GregorianCalendar in that way, you're specifying a year/month/day/etc, but that's it. This doesn't actually mean anything to the system though because it doesn't know what timezone this is supposed to be in reference to, so until you call .setTimeZone for the first time, it has no idea what the actual system time is supposed to be (hence time=?, time being the time since the Epoch).

so, if you don't call setTimeZone before getTime, when you call getTime and it tries to make the Date object (which represents a specific point in time regardless of timezone (again, time since Epoch) ) to return, the current time held by the Calendar isn't a set point in time. To fix this (so that it actually can return something), it sets the time held by the calendar to an actual point in time based on your local timezone (which it already knows) and the month/year/day/etc that you passed into the constructor.

Calls after the first call to setTimeZone don't actually change anything, because changing the frame of reference doesn't change the actual true time.

edit: although I guess I should clarify from my previous post that I guess you could say that getTime() "finalizes" the time such that it causes the time to be set, but when I hear "finalize" in reference to java I think of "uneditable forevermore" like the java "final" keyword

edit2: actually, after looking at the output of c-square's example, it looks like it behaves a little differently than I thought. When the Calendar is constructed, it assumes your local timezone, which you can then override however often you want with setTimeZone(), as long as time is still undefined. The time is then only actually defined when you call getTime() (or probably, call any function that tries to make use of the actual absolute time). So yeah, it is a little weirder than it probably should be :V

Callid · Post by **Callid** » May 21st, 2012, 5:44 pm

I have solved the problem by using the constructor with the TimeZone, followed by a .set(...) for the time. It's a bit strange there is no constructor for both, TimeZone and time

Code: Select all

                TimeZone tz;
                if (msg.length > 6) {
                    tz = TimeZone.getTimeZone("GMT" + msg[6]);
                } else {
                    tz = TimeZone.getTimeZone("GMT");
                }
                int[] date = formatDate(msg[4], sender);
                int[] time = formatTime(msg[5], sender);
                if (time == null || date == null) {
                    return;
                }
                GregorianCalendar greg = new GregorianCalendar(tz);
                greg.set(date[0], date[1], date[2], time[0], time[1], time[2]);

Also, DipBot just managed to, for the first time, create and finish a phase! *is totally proud*

Though I suppose it does need a bit of work :x

Spoiler:

Note: All of the timer lines, except for the seconds, are off. The deadline was created 71 seconds before it ended.

Giogio · Post by **Giogio** » June 25th, 2012, 11:27 am

I was wondering: Is there someone into AI's / neural networks here?

I've had an idea to have actually "life-like" intelligences evolve by (super-short version) putting them into an environment, allowing them to fight & eat each other, cooperate, and so on.
However, I never started it so far since i guess it would need a whole lot of computing time, and I'm not very knowledgeable about how to initialize such neural networks.
Also, maybe it was already done?

So, anyone interested in that stuff?

Akonyl · Post by **Akonyl** » June 25th, 2012, 11:50 am

tbh, you would probably be better off using genetic algorithms than neural networks if you wanted to do something like that. I'm not gonna pretend to be an expert on neural nets, but in general they're better suited to things like pattern recognition for the most part, I imagine using them to actually make complex decisions would require a much more in-depth knowledge of things, unless you split the problem apart, which then sorta defeats the purpose of a net. Not to mention, a neural net of any complexity would be absolute hell to debug if a problem ever rose up.

especially if you're going the competition route, that's a classic example of something people use genetic algorithms for. Not to mention, when something goes wrong you'll actually be able to fix it

Giogio · Post by **Giogio** » June 25th, 2012, 3:58 pm

Sure. I wanted to combine the two things, to get suiting neural net topologies.
That's the main problem for me right now, how to put the topology into a genetic code in a way that allows reasonable mutations, but differs from the learning process I wanted to have too.

Basically, the genetic algorithm gives the starting point of every brain (and maybe, somehow, the way it learns), and then the individual is thrown into a world where it has to learn and adapt to various situations.
That way, I hope, you could get nets that aren't limited to one special problem, but evolve to be able to learn fast. Much more like actual life forms. Maybe I'll get ants ^^

Does that make any sense to you?

Akonyl · Post by **Akonyl** » June 25th, 2012, 4:17 pm

well, if you're using a genetic algorithm to set up the topologies, I guess my question is what your question really is in the first place

It seemed to me at first that you were wondering how to know how to connect the neurons, initial axon weights, etc, but if a genetic algorithm's going to take care of that then I'm not sure what your question actually is.

you also have to be careful how you train the net though / how large you make it, because you can end up over-training the net for specific problems which ends up making it worse overall. Well, theoretically, I've done so little with neural nets that it was never an issue

as for ants, that might also be a bit of a problem since the majority of swarm intelligence (like ants) only work because the entire population works on the same simple set of rules, and if you're running a bunch of different brains around in whatever the world is, the chance of them working together is pretty small, I would think. Besides, it's a lot easier to program an ant from scratch than to hope that a learning program would develop similar behavior (especially since ants use pheremone trails, which your things wouldn't unless they were specifically programmed to)

Kogorou · Post by **Kogorou** » June 25th, 2012, 4:39 pm

Well first of all we should define what an AI is.
Unless someone found a PROPER definition for that everyone can say that he or she has build an AI that is programmed to solve on special problem. Those AIs can even evolve and improve the way they solve the problems.

From what I get you want to build a swarm computation program.
you only need to programm one and start a couple of that program. All you need to make sure is that there is some king of connection between them.
For learning purpouse I would recommend to use a database. Just store the information and build an algorithm that can improve they way it's been stored, sort out any redundancy.

Just allow me one question: why?

Let's get programmy!

Is Assembly the best programming language?

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!

Re: Let's get programmy!