How to check if your results are "any good"

Message boards : Rosetta@home Science : How to check if your results are "any good"

To post messages, you must log in.

AuthorMessage
Profile Tim&Maria

Send message
Joined: 11 Nov 06
Posts: 1
Credit: 194
RAC: 0
Message 31468 - Posted: 20 Nov 2006, 18:04:04 UTC

I have just recently joined, and am now dedicating 100% of my CPU to the rosetta project. I am asking all my friends to join, but most don't want to, and i am baffled as to why. That's another issue.. anyway...

How can i check if the results my computer has generated are any good? I' havn't researched too much, but i'm guessing that each time you recieve credit, that is another "random" stab at the dark of getting a closest match as possible to the actual thing that you are mapping. I take it that at the end of the project, the Rosetta guys take the best result and use that.

Question: how do we know if our results are any good? If they are close to being the optimum thingy for what we are searching. I know i'm being very vaig, but i still havn't fully come to a conclusion about how it all works etc.

Thankyou,
Tim
ID: 31468 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile anders n

Send message
Joined: 19 Sep 05
Posts: 403
Credit: 537,991
RAC: 0
Message 31470 - Posted: 20 Nov 2006, 18:31:39 UTC - in response to Message 31468.  

I have just recently joined, and am now dedicating 100% of my CPU to the rosetta project. I am asking all my friends to join, but most don't want to, and i am baffled as to why. That's another issue.. anyway...

How can i check if the results my computer has generated are any good? I' havn't researched too much, but i'm guessing that each time you recieve credit, that is another "random" stab at the dark of getting a closest match as possible to the actual thing that you are mapping. I take it that at the end of the project, the Rosetta guys take the best result and use that.

Question: how do we know if our results are any good? If they are close to being the optimum thingy for what we are searching. I know i'm being very vaig, but i still havn't fully come to a conclusion about how it all works etc.

Thankyou,
Tim



Hi Tim and welkome to Rosetta.

On Your account page there is this line ,Reports and plots of results from

active work units , press view and you can se your work done.

Hope this helps.

Happy cruching

Anders n

ID: 31470 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile River~~
Avatar

Send message
Joined: 15 Dec 05
Posts: 761
Credit: 285,578
RAC: 0
Message 31476 - Posted: 20 Nov 2006, 20:09:41 UTC
Last modified: 20 Nov 2006, 20:17:35 UTC

What anders says is correct.

But I'd encourage you *not* to take these plots too seriously. Sure it is nice to see how your box has done; it is even nicer if you get the honour one day to be named "predictor of the day" (PotD) on the front page.

But actually, all these random efforts are equally valuable. The whole point is that nobody knows which ones will produce the best answers, and that the bigger the ensemble the more likely it is that the best answer will be close to the right answer. So by all means look at your own results from time to time, and enjoy thinking that your box(es) did the work represented by the red dots on the plot; but please don't get into thinking that your results are less (or more) valuable because they are further (nearer) to the best one.

I find the most interesting thing is to follow a particular plot over several days, and see how the random pattern builds up - both my dots and those for all the other data.

R~~
ID: 31476 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
hugothehermit

Send message
Joined: 26 Sep 05
Posts: 238
Credit: 314,893
RAC: 0
Message 31488 - Posted: 21 Nov 2006, 3:23:28 UTC
Last modified: 21 Nov 2006, 3:39:12 UTC

Question: how do we know if our results are any good? If they are close to being the optimum thingy for what we are searching. I know i'm being very vaig, but i still havn't fully come to a conclusion about how it all works etc.


G'day Tim,
I've added this just in case you are asking how to tell if your results a valid, (didn't error out)

There's a couple of ways to get to that information:

1. Click your own name in a thread like this, and you will go to an Account data for Tim Butler screen, then click on Computers, then on Computer ID, then Results.

or

2. Bookmark This page (this is your results page)

or

3. Go to the Rosetta@Home home page (scroll down to the) Returning participants (section and click on) Your account - view stats, modify preferences (scroll down to the section called) Work done (click the) Results View.

I have just recently joined, and am now dedicating 100% of my CPU to the rosetta project. I am asking all my friends to join, but most don't want to, and i am baffled as to why. That's another issue.. anyway...


Yes I think most of us feel that way, here is a project that is really important but most people when you talk to them just shug their shoulders.

Oh well I hope that helps
-Hugo

Edit: I should of added that no-one can see the amout of information that you can about your computers, just click on my name (and my computers etc... ) to see the difference between the two
ID: 31488 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Rosetta@home Science : How to check if your results are "any good"



©2024 University of Washington
https://www.bakerlab.org