Topic: Gridcoin (GRC) - first coin utilizing BOINC - Official Thread - page 87. (Read 597505 times)

i wasnt watching grc for a while, i had few inconsistently answers about that, if i need for every pc one wallet,
so im asking i dont know maybe fifth time, can i use e.g 4x pc with 1 wallet?
can be magnitude more than 400?

* Update on the CryptoBlox Block Explorer *

CryptoBlox branched a version of Iquidis' Block Explorer for Gridcoin Research and it is very impressive.  It has all of Abes features, plus drill ability for certain things like balance by user, a richlist, graphs, color codes by TX type, search ability, a full fledged powerful API so you can do things like sum a balance with a json api call by address, a users guide, and it seems to be very well built, and its really fast.   RTM I know you have put a lot of work into the alternate version and everyone appreciates that but we really did not know what the state of development was at the time as you know.

Crypto synced research and he is currently looking into creating a tabbed web site with GRC exchange rate, a news feed, and the explorer all in one.  Its pretty impressive.  I did a walkthrough with one of his new coins and it looks very professional.

I think the ETA for the new explorer is approx the time we decommission classic.

Rob

B1bl3 NEWS ‏@B1bl3 via twitter.com 20 hours ago

We added #Gridcoin to http://syncbit.io/  Bitcoin Network .Create your page,your group,invite friends for videochat,tips,games #bitcoin

#Gridcoin becomes 8th in the World on Average #Boinc credits!

Twitter---> http://twitter.com/GridcoinNetwork/status/579355607263621120

Reddit---->http://www.reddit.com/r/gridcoin/comments/2zu7g1/gridcoin_becomes_8th_in_the_world_on_average/

new design iteration of the webpage, what do you think?

http://grctestsite.azurewebsites.net/

Gridcoin is in TOP5 of worldwide BOINC teams (100000+) measured by daily computations!

http://boincstats.com/en/stats/-1/team/list/12/0/0

52up    4up    2up    5    Gridcoin    46,097,421,720    1,656,540,213    11,106,393,664    40,780,095,130    1,380,643,580    -

Update on Gridcoin Finance...  Although I would normally post this on the finance thread, I dont think anyone has been there in 6 months - and the site has been down for 30 days so Ill just post an update here to reinvigorate interest in it.
 
I brought the site back online - finance.gridcoin.us is back up.
 
One of my todo items is to add the close trade button for one to allow people tracking their own fake portfolios to close trades and two, to allow me to close the trades during expiration week (for house managed accounts).
 
During auditing of the ROI a couple of days ago, I found some of our trades are already expired, (these are 4 legged trades), another words, the short put or short call already expired in the past meaning that trade is actually already closed.
 
So, I wanted to keep the ROI accurate, so I made the mark to market system automatically mark an expired trade at the expired value (ie assessed as of the date the box expires) for accuracy.
 
During that process, I found two trades (out of about 50) that had a component in the trade with an unquoteable equity.  One was Master Card, had a 900 call that could not be retrieved.  The other one was AAPL had a box that was quoted on 7-14-2014 (probably when we started the site) that due to the reverse split, could not determine an accurate value of the box.  I just deleted these two trades.  Note that these were actually not Losing trades, so this definitely does not skew the ROI to be in our favor - they are just cases that we have to have better programming to prevent being executed and marked - and code to deal with them.  So I still believe the ROI is accurate.  Next I did a quick check on each option, and 100% of them have marked quotes received by the data service. Another words, over time, I have quotes in the database within 3 days of every option price, with changes per trade.  So there is no way that the system created trades and buffaloed the data over time or skimmed out losers.  They are all there and I have history of the mark to market for each option as time passed (although I only marked them once per week since I have to manually click a button to mark the portfolio to market in the admin panel).  But anyway, the state of the options now: they are all marked with actual quotes and that means the rows should be accurate.
 
 
If you have been taking snapshots you will probably remember we had a 100% total ROI in 6 months, but some of the listed trades showed a very big loss (like -50%) and now you will see they are more like -5%.  I did NOT change ANY trade execution data, the issue was, those trades were actually expired, and the system was still marking them with Todays prices which is actually not possible, since the options are expired, the position price has to be assessed as of the day they expired, and they are now.  (And to add complexity to the issue, when I say today, I mean 2 of the 4 legs of the Live options in a box were marked with todays prices while the 2 legs that were expired were Deep in the money with full intrinsic value, showing a loss greater than would have been possible).  But to put people at ease, when we go back to the expiry date of those rows, the system will automatically detect the earliest end date out of the 4 legs, and assess the row as of the expiration date giving a fully valid realistic price which is what we see right now).
 
So to summarize the first iteration of finding, executing, marking and reporting the trades appears to be working fine and now we need to start over, re-analyze current conditions, enter new trades and watch a new cycle and see if we can repeat the performance.
 
But its very promising in that this cycle was a complete success with 100% ROI .
If anyone wants to audit the individual legs feel free to drill into all the data and shoot holes in it.
 
One useful report I would like to add is what percent of trades did we find, what percent were profitable.  The other positive feedback that I see based on the results is the winners tended to be larger than the losers by at least 1 magnitude.  Meaning the losers we picked became very small losers with small percentages and the winners became large winning percentage.  This lends credence to the original algorithm in the sense that one of the criteria to pick a trade requires the profit/risk ratio to be high before it is considered to be traded - meaning that if you can earn $100 on $100 with a $20 risk it is considered where earning $10 on $100 with a $75 risk is not considered.  The results mirror the selection process since you can see, the losers bore out to be approximately capped at the levels of the original assessed risk (not because we have risk caps per row, because the Actual risk in the position was low due to the nature of the black scholes model for that trade).
 
Rob

I have a first version of the gridcoin machine learning project as well as a whitepaper explaining excactly what it does.
 
The BOINC project is at: http://ec2-54-93-162...s.com/boinctest
 
What the client does is basically:
- download training data, data to be predicted and a program
- the program find patterns in the training data with a random forests and makes predictions according to those patterns for the data to be predcited
- the predictions are uploaded again to the server
- on the server the different predictions from the different clients are aggregated (for why the prediction are different, see the paper)
 
The link to the whitepaper is here: https://drive.google...XhiNUtBREk/view
 
What is needed now is a discussion, under what exact circumstances does gridcoin support commercial projects/DACs?
 
What do you think of this project?
 
The thread of the project is here: https://cryptocointa...19-gridcoin-ml/, maybe it's better to have some of the discussion there.

You need RAC > 100 in a project that Gridcoin recognizes your research.

And in Leiden you should wait, up to 2 days will it take from the point when you reached 100 RAC.

sorry quite different, can anyone tell me what I should do? please see below

"Errors" : "Project missing in [Netsoft] credit verification node. Please contact Netsoft to add this project. (Wait at least 24 hours for new projects to propagate)."
} ??
.....................................................
Command" : "cpids"

},
{
"Project" : "constellation",
"CPID" : "64c0e7135926261547f8d260a6b2aad9",
"RAC" : 10.00000000,
"Team" : "gridcoin",
"Verified RAC" : 9.00000000,
"Verified Team" : "gridcoin",
"Is my CPID Valid?" : false,
"CPID Link" : "http://boinc.netsoft-online.com/get_user.php?cpid=64xxxx
"Errors" : "Verified RAC too low."
},
{
"Project" : "leiden classical",
"CPID" : "64c0e7135926261547f8d260a6b2aad9",
"RAC" : 185.00000000,
"Team" : "gridcoin",
"Verified RAC" : 0.00000000,
"Verified Team" : "",
"Is my CPID Valid?" : false,
"CPID Link" : "http://boinc.netsoft-online.com/get_user.php?cpid=64,xxxxxx
"Errors" : "Project missing in [Netsoft] credit verification node. Please contact Netsoft to add this project. (Wait at least 24 hours for new projects to propagate)."
}
]