Shift Spectrum to Rest Frame in Marvin-web Display

[bretthandrews]

As FuHeng (Eric) Liang suggested, the spectral plot in Marvin-web would be more useful if it was in rest frame.

[kbwestfall]

Maybe a stupid suggestion, but could we have the rest wavelength be labeled on the top axis so that people could see both the observed and rest wavelength? Would we de-redshift based on the NSA redshift, the stellar velocity, or the gas velocity? To keep it simple, I imagine we want to use the bulk redshift from the NSA.

[bretthandrews]

Dual wavelength axes is a good suggestion. We'll have to look into Dygraphs to see if it has that functionality (it's not that feature-heavy, so I wouldn't be surprised if it cannot).

As for the redshift to use...I think that the gas or stellar velocity would be more helpful in high S/N regimes but could lead to catastrophic errors in low S/N spaxels.

[havok2063]

Yeah I agree, dual axes is a better suggestion than replacement or making it toggleable. I had a quick look and it seems Dygraphs does not allow for multiple x axes. y axes yes. The closest out-of-the-box thing looks like plotting it as an independent series.

[kbwestfall]

If it's easier to toggle, that might be the best solution.

[havok2063]

it wouldn't be trivial to implement, but maybe this means we need to start looking for a new Javascript library that has a bit more functionality, or at the very least, is better documented.

[havok2063]

Just to expand a bit on the work needed to make it toggleable. We would need to write javascript and html to

1. Store the wavelength array on the client side
2. Add a toggle button / checkbox to the interface
3. Store or compute the rest wavelength on the client side
4. Re-render the Dygraphs plot fresh from upon button click.

Most of the work is in steps 2 and 4.

[kbwestfall]

Yeah, doesn't sound trivial. I personally would put this at low priority. It's basically a "polishing" issue, I think.

[havok2063]

I would agree. I would put this in the category of user freedoms or preferences, similar to maps range rescaling.

[kbwestfall]

The rescaling of the maps is arguably more important. I think many more people will play with the maps than will actually look at individual spectra.

[havok2063]

I agree but I think, right now, adding user flexibility on the web interface is a low priority in general, given everything else. I do think moving the DAP custom color scale and custom ranges for defined map-types from the client side into the Marvin data-model, and pushing those custom values to the web, is a priority however.

[albireox]

Just a comment that this was discussed again today in the DevCon and we agreed to look into it. I'm moving it to backlog.

[havok2063]

This has been implemented and merged in 2.3.5 for MPL-9