General Azoo project statistics.
-
by MvGulik
Reading this:
[DZM, June 23 2015] On a better note, you incredible guys are closing in on 2 million classifications; WOW. (Less than 7,000 to go.)
I think Azoo could use some general stats page, where stuff like that, and more, could be viewed.
For the moment that's about it from me on the subject.
Ps: I use quote-links that are generally best opened into a new browser-tab.
(160301:0800):[pro:cls:cmp:Usr] AZ:2523969:159568:11515 AP:1916075:38155:10682 BD:422219:126186:3452 CS:2626905:360611:23028 CG:3115880:93977:52365
.
(160401:1300):[pro:cls:cmp:Usr] AZ:2579425:167588:11854 AP:1916815:38192:10688 BD:539697:158108:4763 CS:2630537:360611:23092 CG:3117104:94028:52365
.
160501:0900 [pro:cls:cmp:Usr] AZ:2635121:177437:12213 AP:1916982:38201:10689 BD:548107:158108:4816 CS:2630537:360611:23092 CG:3118297:94034:52365
.
160522:0800 [pro:cls:cmp:Usr] AZ:2664780:182628:12333
.
Posted
-
by DZM admin
I'm very well aware that we need public-facing stats. I talk about it all the time! Trying to make it happen for new projects.
In the meantime, I can tell you that this project currently has 2,324,440 classifications. 😃
Posted
-
by MvGulik
Today (May 1 2016) AZ is at:
2,6M Classifications,
with 177k marked compleet,
and 12k registered users.Posted
-
by zutopian in response to MvGulik's comment.
Today AZ is at:
2,6M Classifications,
with 177k marked compleet,
and 12k registered users.There were initially 3 Mio. images to be classified, as far as I know! Are there just 177,000 images, whose classifications are complete?
How many persons are asked to classify each image?
So we would know, how many classifications are still "necessary"!Posted
-
by MvGulik
(by @zutopian) There were initially 3 Mio. images to be classified, as far as I know!
I can't remember having seen any information on the total number of AZ-sets. If you have some related links please post them, so I can take a look.
(zutopian) How many persons are asked to classify each image? So we would know, how many classifications are still "necessary"!
Currently I have no available information on that. ... Hoping someone else might have some.
I do have have some data I can scan, but I'm not sure when I get around to that (as some code-rewriting is needed for that).
Posted
-
by zutopian in response to MvGulik's comment.
I can't remember having seen any information on the total number of AZ-sets. If you have some related links please post them, so I can take a look.
"3 Mio. images" had been mentioned in below Zooniverse blog post.
Well, I am not sure, if "uses 3 Mio. images" means, that 3 Mio. images are asked to be classified!
Actually, there are "sequences of 4 images" displayed for classification!
Were there initially 750,000 "sequences of 4 images" (177,000 complete)?Anyone Fancy an Asteroid?, June 27, 2013
(...) a system that uses more than 3 million images, taken data from the Catalina Sky Survey (...).
https://blog.zooniverse.org/2013/06/27/anyone-fancy-an-asteroid/
PS: Where did you find the below data, about which you informed?
Today AZ is at: 2,6M Classifications, with 177k marked compleet, and 12k registered users.
Posted
-
by MvGulik in response to zutopian's comment.
https://blog.zooniverse.org/2013/06/27/anyone-fancy-an-asteroid/
Aha, thanks.
Yea, I also find that "3 million images" a bit ambiguous.
Presuming that 3 million is actual AZ-sets, its suggests that around 3M/289 = 10380.~ CSS master frame sets where used. Which, although a bit high given the current CSS master frame count (03325) on the Azoo Updates page, it seems not an improbably number.
(289 (1717 grid), which is the number of AZ-sets pulled from a single CSS master frame/set.*)10380 divided by 4 (images per set) would becoming 2595.~, which is just to low. But ... Divided by a general average of 3, as not all sets have actually 4 images per set, we get 3460.~. Which, as far as I know, seems a highly plausibly outcome too.
I'm not sure how this holds up in relation to the current classification count. I will need to do a bit more thinking on that one, and see if I can pull some helpful stats from the AZ data I have stored (like the general number of classifications on finished AZ-sets).
PS: Where did you find the below data, about which you informed?
The only thing I feel comfortably with is saying that I'm pulling that data from the Ouroboros api.
Sorry.Posted
-
by MvGulik in response to zutopian's comment.
@zutopian: How many persons are asked to classify each image?
The answer on that is 30, for good AZ set's that is.
(Based on a 47k sample from the start of the available AZ-set's)Needles to say that none good AZ-set's, cq: marked as bad, get kicked / retired a lot sooner.
Posted
-
by MvGulik
Seems the current last available AZ-set - AAZ000ldjv (== 997,339) - has been the last one for some time.
(Feb 1 2015 - @planetaryscience .. May 17 2016 (now))Base on the completed AZ-Sets count of 177,437, that still leaves a significant number of AZ-sets to be completed. Some of those are already partially completed of course, but I think its save to say AZ is not running out of work any time soon.
Almost forgot ...Considering there is this nice "Total AZ-set's" value.
AZ_Sets per masterframe: 289 = 17 * 17.
That gives 3451 masterframes. 3451 = 997339 / 289.CSS-ID's not included in Update-Page:
- v(2016-05-08): 126 = 3451 - 3325.
- v(2016-06-06): 23 = 3451 - 3428.
Posted
-
by MvGulik
Some AZ related graphical ... pixel art.
...
(Pick your zoom-tool with care.)Posted