Galaxy Zoo Talk

Must refresh every 5 classifications or so

  • williamaskew by williamaskew

    Must refresh every 5 classifications or so? is anyone else having GZ freeze up after a few classifications and has to hit refresh button?

    Posted

  • sat666leg by sat666leg

    mine was doing something ? but i was doing a little post in discussions. it kept coming up with a window with kill / wait. quite a few times.

    Posted

  • Budgieye by Budgieye moderator

    Is GZ working now without the screen freezing?

    Posted

  • williamaskew by williamaskew

    Nope its still happening

    Posted

  • Abe_Hoekstra by Abe_Hoekstra

    I don't have this problem, Billy.

    Have you tried deleting your cookies or using another browser?

    Posted

  • williamaskew by williamaskew

    its happening on two different computers. very frustration

    Posted

  • Abe_Hoekstra by Abe_Hoekstra

    Geez, then it could be a GZ problem.

    Posted

  • Abe_Hoekstra by Abe_Hoekstra

    I hope GZ can solve the problems, Billy. I not, you might think about creating another account. Maybe the new one works.

    Posted

  • ElisabethB by ElisabethB moderator

    I've been trying things out with Chrome and Firefox. No problems for me.
    Which browser are you using

    Posted

  • williamaskew by williamaskew

    both chrome and safari aren't working for me. I used a friends computer and the same thing so its in GZ

    Posted

  • Abe_Hoekstra by Abe_Hoekstra

    Maybe you could pm DZM, maybe he can do something.

    Posted

  • ElisabethB by ElisabethB moderator

    DZM is no longer with the Zooniverse.

    @Billy,

    Could it be that you have classified the entire dataset ?

    Posted

  • williamaskew by williamaskew

    how many are there EB? 15000?

    Posted

  • ElisabethB by ElisabethB moderator in response to williamaskew's comment.

    Not sure ! I'll poke Kyle ! 😄

    Posted

  • KWillett by KWillett scientist, admin, translator

    Hi @williamaskew,

    We are coming up on finishing the DECaLS dataset; while they're not entirely finished yet, we've retired almost all the subjects. Since you're a prolific classifier, it might indeed be possible that you've done all of the ones that are currently active (there were about 33,000 total, but all but a few thousand have been retired as of today).

    I'm working on getting some more specific numbers right now. There should still be significant numbers of Illustris images available, so I suggest refreshing if it throws the same errors. Please keep reporting here if it's a continuing issue, though.

    Posted

  • KWillett by KWillett scientist, admin, translator

    Update:

    I feel we should drop some balloons! @williamaskew, you've become the first person to completely finish the DECaLS dataset. That's monumentally impressive - thanks so much for your help in doing more than 33,000 galaxies.

    As I said before, there are several thousand DECaLS images left, but we'll probably be finished with them within one week (maybe less). I will work on putting in a larger sample of Illustris images so that everyone will have new subjects to classify, but that may take a few days. We are working hard on a new set of images from DECaLS DR2, but we will probably have at least a short period when only the Illustris images will be on the site.

    @ElisabethB, @Budgieye, @Capella05 - if anyone has similar problems (especially if they're a very prolific classifier), this is likely the reason. Please let me know if anything different comes up, though.

    Thanks everyone! Time to ramp up our analysis on the new results.

    Posted

  • ElisabethB by ElisabethB moderator

    Woohoo ! Go Billy ! 😄

    Posted

  • Abe_Hoekstra by Abe_Hoekstra

    Wow, Billy, you are awesome! Congrats!

    Posted

  • Budgieye by Budgieye moderator

    Amazing! Congratulations Billy!

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR

    Congrats Billy!


    Illustris only? Maybe an idea to re-use retired objects? Surely a lot of zooites classifying now haven't seen them?

    • Would probably up the chance of new people stick around
    • Now also searching for specifics we weren't at the start of DECaLS images, double dust among others (x-shaped bulges and polar rings neither?)

    New eyes see new things

    Posted

  • KWillett by KWillett scientist, admin, translator

    That's an interesting point, @Ghost_Sheep_SWR. I wonder if you could expand on it a bit:

    • We place a really big emphasis on trying to only ask Zooniverse volunteers to do tasks that are worthwhile and leading to new results (eg, don't waste anyone's time). In most images of Galaxy Zoo, we've shown that getting between 30-50 classifications is enough for an accurate classification. We don't want to have people adding votes to the same image if it isn't going to improve our knowledge!
    • Why do you think that using retired objects would have more users stick around? I'm not sure I follow why that'd be the case - I think many users are excited by new images and the idea that they will be the first (or one of the first) to view and classify these galaxies.
    • The last part is quite interesting, I think. I don't know if redoing the entire selection would be worthwhile, but maybe a smaller subsample based on the classifications + tags that users have already identified would be interesting for the truly rare objects.

    Kyle

    Posted

  • williamaskew by williamaskew in response to Ghost_Sheep_SWR's comment.

    Kyle

    I have no problem running through the data again (as long as i dont have to refresh every 5 objects), though new would be better. When i started classifying the DECaLS data i wasnt looking for anything in particular. Since then I have started a new thread "peas of a different colour" and would still like to go through the older DECaLS data for anything we missed. It seems DECaLS picks up the faint of objects much clearer than SDSS so perhaps we have come across some new ones. Not that anything will become of them but its good to use the data while it is available.

    http://talk.galaxyzoo.org/?_ga=1.216109846.189379650.1438018854#/boards/BGZ0000004/discussions/DGZ0001thi?page=10&comment_id=56b8c5a8f724d746520007a6

    Posted

  • Capella05 by Capella05 moderator

    Hi Kyle,

    This is my personnel opinion - as a volunteer who has been classifying for almost 8 years - not in my capacity as a Mod.

    We place a really big emphasis on trying to only ask Zooniverse volunteers to do tasks that are worthwhile and leading to new results (eg, don't waste anyone's time). In most images of Galaxy Zoo, we've shown that getting between 30-50 classifications is enough for an accurate classification. We don't want to have people adding votes to the same image if it isn't going to improve our knowledge!

    Agree - one of the main pillars of citizen science is knowing that we are making a valuable contribution and that our time is not being wasted. I trust that the Zooniverse will not have me classifying images that have no intrinsic scientific value.

    Why do you think that using retired objects would have more users stick around? I'm not sure I follow why that'd be the case - I think many users are excited by new images and the idea that they will be the first (or one of the first) to view and classify these galaxies.

    I disagree with the original poster( @Ghost_Sheep_SWR) - we have had previously retired images on the site (SDSS - just over a year ago- the reason why, has escaped my memory), and instead of encouraging participation it had the opposite effect - we landed up discouraging a lot of new volunteers. They disliked the fact that they were being given old images, that had first been classified 3+ years ago, to classify. As Kyle pointed out, most of the thrill comes with new images that might never of been seen with a human eye before 😃

    As for only having Illustris images? Noooooo! I like classifying galaxies, not sims. Having the odd sim is okay, if it serves a purposes - like weighting our classifications. I think a lot of people might take a 'sabbatical' if that ever happened.

    So....... and this will be controversial - but until we get new data - perhaps a break might be in order?

    This will allow a chance for volunteers to evaluate their collections, and a perfect opportunity for the Scientists to provide feedback for all the interesting objects we have raised here on Talk over the past few years! Also, it would be nice if we (the GZ community), could be given a heads up on future plans / surveys etc. of Galaxy Zoo.

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR

    Of course Kyle,

    For me it's easier to split my answer in parts,


    Illustris

    Although there is maybe just as much scientific value to classifying Illustris images as any 'real-world' set, for me personally it just isn't interesting enough to keep classifying them, for a part because i'm here for 'the real deal'. I believe the normal ratio for Illustris / DECaLS was 1:3, at a certain point i just stopped classifying them. I know this is very very subjective but feel it's reflected in the minimal amount of comments, questions and posts about them. From this point of view an Illustris-only GZ would be very uninteresting, and probably disappointing for people who then come for the first time to check it out. Maybe this is also reflected in the number and/ or quality of Illustris classifications?


    New zooites

    I think it would be good to try to get more long-term zooites, many new names only stay 1 day or around that. More people means more classifications in a sporter amount of time. Although the task seems rather simple; classify what you see; as a relative new zooite i can tell you the learning curve actually is really steep. All kinds of artifacts, objects, terms and abbreviations, links to information sources, spectra etc..

    In this way the retired objects form a giant database of questions and answers, both on the imaged objects and in general. So when using this retired images, its very likely that when an image provokes a question the answer might already be in the comments of that object.


    New knowledge / object types

    I've learned a lot the last past months, so for me GalaxyZoo really is a two-way process which reinforces itself. It's not only the classifications that count, but also the knowledge floating back to me.

    I really enjoy learning about Pea's or how to identify an asteroid found in an image. This in turn improves the way i classify and look for different objects. Hence I really disagree with 'wasting someones time' with completed images.

    Lastly in time people learn more about finding information on objects, so i will look / classify an image very different now than i did 5 months ago. An image that has 30 classifications, might on the 31st time be checked on it's spectrum for example.

    And also as mentioned before, something as a double dust collection is relative new, retired images haven't been checked for this (among other things)


    Pffff, so a lot of text but hope this helps / clarifies a bit. In any case, if an image has 30 classifications would it matter if you'd add 20?

    Here to learn, discover and contribute, not in any particular order 😉

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR in response to Capella05's comment.

    I did not know about the retired images and the effect this had. My comment was inspired by a Illustrus-only GZ, in combination with DECaLS retired images, not from 3 years ago!

    Posted

  • Capella05 by Capella05 moderator in response to Ghost_Sheep_SWR's comment.

    If any survey was 'revisited' it would have the same effect - consider this 'hypothetical' situation:

    • User 'A' starts classifying, and comes across an amazing image that has him / her jumping in their respective seats.
    • They go to Talk to tag it / discuss it...
    • Only to discover there are multiple comments about the object, some of them going back years...
    • #dissapointed #whyamIclassifyingOldImages #whatsThePoint would all be valid hashtags, and it wouldn't be the first time I have seen similar tags....

    This happened - in fact several current posters did have quite a moan about it.

    I would like to see GZ try too recapture the excitement we had in the first iterations of GZ - and with a bit of downtime and spring cleaning - I feel that this might be an opportunity to do so.

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR in response to Capella05's comment.

    This might also be a realistic situation yes, maybe slightly less with images from 2015.

    Illustris only GZ or break is the same for me 😃

    But a break then sounds actually good, until a good new set at least. DECaLS DR2 or maybe another starting after a break sounds exciting, and maybe generate a little buzz

    Posted

  • KWillett by KWillett scientist, admin, translator

    Down to ~2500 DECaLS galaxies. A couple other users have also completed all the active subjects; should start ramping up a bit faster now.

    Thank you everyone for the comments. I'm collating these and will bring them up at our science team meeting this Monday - do please let me know (either here or by PM) if you have any more thoughts on this.

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR

    Just wondering what the main difference is of a new DECaLS DR2 set in comparison with DR1, and also will it contain new objects / subjects?

    And any other possible surveys / sets you are taking in consideration ( or hoping for)?

    Thanks

    Posted

  • KWillett by KWillett scientist, admin, translator

    99.78% done. Watching it tick down ...

    DECaLS DR2 is simply the latest data release from the survey, so it would include a substantial amount of new targets that the telescope has imaged in the past year. The quality of the images would be similar, although we might decide to include different galaxies from the catalog based on our scientific use cases. That hasn't yet been decided.

    We are indeed considering other surveys, but I can't say too much about them yet. We'll let everyone know as soon as we've reached an agreement.

    Posted

  • KWillett by KWillett scientist, admin, translator

    Hi everyone,

    Thanks sincerely for the discussion here. I've posted an update on our current progress at http://talk.galaxyzoo.org/#/boards/BGZ0000003/discussions/DGZ0001ucb

    Please let me know if you have any questions. @williamaskew and others - I've reset your status on the DECaLS DR1 set, so you should be able to see those galaxies in your interface again.

    Posted