As of this morning (feb 19) we started using Mersenne twister instead of drand48() to generate the random numbers for our particle swarm optimization. I'm hoping that this will improve the results a little bit, by adding some more variety to the random numbers we're generating.

---

Good call.

I also added median values to the output and graphs. It'll probably take them awhile to catch up with the median values however.

---

What we have in this post is pretty much preliminary work, so we'd really like to hear your ideas and suggestions as to what you'd like to know is going on server-side, and what we're doing as a whole.

Right now we would like to get these graphs (and some other ones) automatically generated in their own webpage, with information like what users have given us the best fitness and things like that.

Pretty much we could have whatever information you guys think would be interesting, and we'll also have things we're interested about and looking into on their for our own benefit.

Basically, we'd like a webpage we could check to easily know the status of our different searches, and I thought it would be more interesting with some user input, so we could make that information more interesting to you guys, and also add things you'd be interested in as crunchers.

I would suggest, no urge that the project utilize this new webpage as a means to provide a vastly greater educational opportunity for the volunteer participants. Thus, I believe it should provide information on both the fundamental physics and the computational aspects of the project. By computational aspects, I most certainly do not mean any discussion of credits or who has the bigger, faster hardware--such discussions are all too prevalent as it is on the forum. Some suggested reading materials on both subject matters would be excellent additions.

Allow your volunteers to learn if they so choose. The lack of such learning resources is, in my opinion, a great weakness in the BOINC concept.

Best regards,
EigenState

The current situation in ps_s21_9 looks interesting. Is this a case of the particles moving from one local maximum to another, or are we seeing some outliers?

The current situation in ps_s21_9 looks interesting. Is this a case of the particles moving from one local maximum to another, or are we seeing some outliers?

Not quite sure :) That's why I'm letting it keep running.

---

Hehe, that's cool. What about ps_s23_9? You can still just about see when the 'best' values shot up. I'm still not sure what you're aiming for - do you want the worst values to converge to the best ones, while finding the highest average overall? (if so, why aren't you showing the worst values in your graph?)

By the way, if the goal is for the particles to converge (i.e. all of them moving toward the global maximum?), could you add the mean absolute deviation or the interquartile range? (I guess if you add them at the bottom of the graphs you'd need a second range for the vertical axis, which could go on the right)

I would suggest that the bottom line of numbers be made X10000, to eliminate the mess of 0's.

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Hehe, that's cool. What about ps_s23_9? You can still just about see when the 'best' values shot up. I'm still not sure what you're aiming for - do you want the worst values to converge to the best ones, while finding the highest average overall? (if so, why aren't you showing the worst values in your graph?)

I just was playing around and took out the worst values because you could see a bit better how the best was progressing that way. It would be nice for the worst values to converge to the best, because that's an indication that the search is over, but it's not really necessary. We're just looking to find the highest fitness

By the way, if the goal is for the particles to converge (i.e. all of them moving toward the global maximum?), could you add the mean absolute deviation or the interquartile range? (I guess if you add them at the bottom of the graphs you'd need a second range for the vertical axis, which could go on the right)

Yeah I have our searches generate standard deviation and a few other values to the logs but haven't made a graph for that yet. It should be pretty easy to add one :)

---

With regard to a screensaver, I have to admit that at times it is nice to be able to look at the data, but, with the runtimes so short now it does not seem to make much sense. My "slowest" system when I was running it last week I think was down to 10 minutes per task ... that is not much time to do much of anything of interest...

and the high end at 12 seconds, well, blink and you missed it all ...

Looks like the graphs haven't been updated in a few days - I thought they were scripted to run automatically.. Perhaps you could add a timestamp to the bottom of the post so we know when it was last updated.

Looks like the graphs haven't been updated in a few days - I thought they were scripted to run automatically.. Perhaps you could add a timestamp to the bottom of the post so we know when it was last updated.

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

---

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

I'll click the Red X on him.

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

I'll click the Red X on him.

Dave will blame it on a freshman.

---

me@rescam.org

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

I'll click the Red X on him.

Dave will blame it on a freshman.

I think you have something there. A freshman prodding Dave with a cattle prod. Now that's an idea for a screensaver :P

---

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

I'll click the Red X on him.

Dave will blame it on a freshman.

I think you have something there. A freshman prodding Dave with a cattle prod. Now that's an idea for a screensaver :P

I might have to enroll.

---

me@rescam.org

Dave was supposed to have gotten them run automatically :P I guess he didn't... I'm going to bug him some more.

Keep that cattle prod handy!

I'll click the Red X on him.

Dave will blame it on a freshman.

I think you have something there. A freshman prodding Dave with a cattle prod. Now that's an idea for a screensaver :P

I might have to enroll.

The next hit internet game. Prod Dave & the sequel Prod Dave Again!!

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Updated these to the newer searches which are using some limited redundancy to check new local/global best particles ;) Hopefully these should improve better than the last set.

---