Chat

discussion

Version 2 Feature Requests

Started by mschwamb
Default_user
about 6 years ago

drmshake:

Could there be an improvement with communication? Maybe connect the sites membership with places like facebook - I feel there are a lot of discussions that go "undiscussed" because of poor navigation on discussion boards

A good place to have a discussion is on the message boards - anyone can create a thread

~Meg

Default_user
about 6 years ago

A "subscription" option on posts would be nice, so one doesn't have to search for answers. IE Email when anyone posts to a topic or optionally when anyone answers a post.

Default_user
about 6 years ago

Improve the speed of the interface! Often it takes just half a second to see that you netted another quiet star with too much noise to see any transits, but then it takes another 3 clicks - each of which can take multiple seconds to respond - just to say "next!"

I don't mind spending extra time looking at potentially interesting light curves. I do mind having to wait precious seconds waiting just for the interface to switch from "variability" to "are there any transits"? There is no reason why this should take so long when you're not even downloading new data!

Default_user
about 6 years ago

Being able to see all the quarters of data for the star would be useful as well.

And being able to annotate transits on a light curve that you found in some other way (i.e. not through the Classify button).

Default_user
about 6 years ago

More simulated data would be good too. The current tutorial doesn't do a great job at showing some of the harder transits to catch - e.g. a wide but fairly shallow transit in a star with a lot of noise etc.

I think this would also relieve the occasional tedium of hitting a dozen or more stars with no noticeable transits.

Default_user
about 6 years ago
lindrothu in response to PatrikD

PatrikD:

Being able to see all the quarters of data for the star would be useful as well.

And being able to annotate transits on a light curve that you found in some other way (i.e. not through the Classify button).

You can see the other quarters if you answer yes to discuss this star. However what I wanted to ask was why one cannot then mark those other quarters if one chooses to?

I have seen a number of possible transits, and seen more in other quarters at which point one would like to mark all the data for the star.

Default_user
about 6 years ago

One more suggestion -

The noise on the light curve makes it hard for the naked eye to spot some of the more low-amplitude transits. You have the little moving average at the bottom, but that one is really only useful as an index into the data. I would suggest given the eye some more information to find patterns: below the raw data timepoints, display a moving average with a very short window. Below that one, you can draw a second moving average, this time with a longer time window, etc. Each of these would be offset vertically by some amount, but they could all be drawn on the same time series. You may need to increase the amplitude a bit on the longer-window averages though.

By filtering across longer time periods, you remove a lot of the spurious noise the eye otherwise latches onto. Transits of different width and depth will stand out better at different levels of filtering, so these additional curves will help the eye to pick out the patterns more easily.

Default_user
about 6 years ago

And while on analysis, it would be nice to have a vertical line from the cursor to the time axis so that one could check the timing between suspected, or obvious, transits easily. Perhaps even fix such lines with a click and display the time between adjacent lines.

Default_user
about 6 years ago

It would be nice if Talk page had a third drop down box added under "RECENT DISCUSSIONS" cell to allow display of 25/50/75/100 recent posts at a time, in addition to the options we've already got for sorting by time or source.

Default_user
about 6 years ago

A handy tool for the next version would be a view mode that would zoom in to a very narrow view on the X axis... an aperture of perhaps half a day, and then animate the sequence. I've found from my experience with other types of waveforms, most notably audio, that patterns can get lost in the noise when seeing the big picture, but looking at narrow snapshots doesn't give enough data to see a discernible pattern. However, when you can see the data animated at a high magnification, frequency components you never would have noticed stand right out. It's not very handy for picking out transients in the short term, but would greatly help in getting variable stars pigeonholed into the proper category... I just read up in the discussion, and this is pretty much the same thing that PatrickD suggested 3 posts ago. Also, if data exists for longer periods than one quarter, then this method might be key to detecting the transients of an exoplanet with a long orbital period that might have otherwise been mistaken for a spurious measurement artifact.

Please Log In to make comments.