Public Data Access Overview / Discussion Forum

Dear all,

I am trying to look for the interacting galaxies (maybe pairs, triplets, etc.) but I am not get use to Data Access here. I tried to use the tool Visualize Galaxies and Halos to see the galaxies as well as the example code on web-based Lab but these are the furthest I can go, I could see one galaxy at max.

Let's take an example, in the paper of Patton+ 2020, figure 2, they illustrate some pairs of interacting galaxies with different separation. In short, I want to ask that how can I visualize several galaxies at once? Sorry if the question bother you.

Thank you in advance,
Thang Huynh

Under "render method" you should select "sphMap", instead of "sphMap_subhalo". The first includes all particles/cells in the entire halo, while the second only includes the particles/cells in a single subhalo.

(Almost all mergers will be made of two subhalos, so if you only visualize individual subhalos, you will not see mergers).

Thank you very much, it is greatly appreciated.

One question, are there any ways that I can search for the mergers/interacting galaxies more convenient or I must, let say, randomly choosing the subhaloID?

Another question, for the data, if I use the Subhalo, it means that data belong to that galaxy right? So how can I take the data of a pair? Is it using the data in Parent Halo?

Thank you in advance,
Thang Huynh

I see a few possibilities to search for mergers:

(1) There is a published paper analyzing mergers from the simulation you are interested in. You can get the sample (i.e. list of subhalo IDs).

(2) You can use visual inspection, i.e. make lots of images of halos and identify those with ongoing mergers.

(3) You can define a quantitative selection, at a particular time, e.g. the stellar mass ratio of the central versus most massive satellite galaxy. Or, subhalos within a given stellar mass ratio (and distance) range. You can easily implement these using the group catalogs.

(4) You can use the merger tree, to identify actual merger events. Then you can look at the snapshots immediately preceding these events, when the individual galaxies still exist.

Once you identify a pair I would carefully look at subhalo vs halo data, to see what the differences are.

Thank you for such informative instructions.

So according to Task 2 in the API CookBook, there is a command to find subhalos with a range of mass:
search_query = "?mass__gt=" + str(mass_min) + "&mass__lt=" + str(mass_max)
Are there ways to retrieve the range for stellar mass ratio and distance of galaxies like that? Using which keywords?

No, there are an infinite number of such derivable quantities, and this simple search cannot accept e.g. mathematical expressions.

If you want to do more complex searches, you should either (i) download the group catalog to your local computer and use the example scripts, or (ii) use the online Lab service.

The code you would write, to load and search the catalogs, will be identical in both cases.

According to your above statement:
"Under 'render method' you should select 'sphMap', instead of 'sphMap_subhalo'. The first includes all particles/cells in the entire halo, while the second only includes the particles/cells in a single subhalo."

If I take the data of SubhaloID 490053 using API https://www.tng-project.org/api/TNG100-1/snapshots/99/subhalos/490053/ like this, take "MagneticField" as an example, what I retrieve is the data similar to the one I visualize using "sphMap" or "sphMap_subhalo"? In case it is just the "sphMap_subhalo" meaning I have to get the API of the HaloID, right?

For central subhalos, they will be very similar (but not identical).

You should test on a larger halo (with many satellites) to see the difference, e.g. halo 100 instead of halo 1596.

So for 490053 and 490054, they are an interacting pair and belong to HaloID 1596. But if I use the API for 490053, the data will just for only 490053 and there are nothing related to 490054, right? So it'd better to use the HaloID? In such case, how do I know how many SubhaloIDs a HaloID contains?

Please check out the group catalog documentation, you will find the relevant field(s).

Sorry but I am still wondering about how to take data of interaction? In the Visualization tool, it is just straightforward by changing the Render_Method from Subhalo to Halo but I don't know if there is a similar way for using the API data. If I just use the data from SubhaloID 490053, there would be no data of SubID 490054. I try an approach that take the Halo data and then cut a part by taking 'pos_x/y/z' of 490053 and plus/minus with 'halfmassradius' of the SubID to take the data of the 2 interacting IDs from the Halo. I want to ask that if my approach is right and are there any others?

Yes you should take the halo data, of the (parent) halo containing those two subhalos, instead of the subhalo data alone.

The halo data will contain both subhalos. If you plot the stars, gas, or DM, you should clearly see the ongoing merger.

After that, you mention "cut a part", that depends on what you're trying to do. There probably isn't any need to make any further cuts.

Thank you for your information.

By the way, I mentioned "cut a part" because for some Halo, there are many galaxies but I am just interested in the interactions so I made a cut around the region of the pair.

Dear Dr. Nelson,

As far as I know, using GroupNsubs in FoF Halos gives us info of how many subhalos that halo contains, and GroupFirstSub returns the central subhalo of that halo.

However, is there a way to find which subhaloID that halo contains? I mean every ID of the subhalos instead of just a total of number. If not, is there any convenient way? I am thinking of running back and forth starting from the subhaloID from GroupFirstSub and using SubhaloGrNr in Subfind Subhalos to check for the ParentHalo but that seems to be large time-complexity.

Thank you in advance,
Thang Huynh

Please see this thread.

Thank you very much, it is really helpful.

Hello,
I have followed the steps and information that dear Nelson comments, of which I visually detected 200 pairs, but they were substantially reduced when I reviewed again and again, so the visualizer was of great help, however it did not help me. It has become clear how to generate or how to obtain, or what steps to follow to be able to obtain pairs as shown in the image according to Jun-Sung Moon et al (2021), if your comments or scoldings in this regard are very useful, since this has made me led to my manuscript being rejected because it did not have images like these. If you could please guide me. thanks thanks

Perhaps the process is very obvious, however I prefer to ask.

In that example image you have the pairs of Subfind IDs available. You should check if, in general, one is a central and one is a satellite. I would guess not. Therefore, those pairs are pairs of centrals, separated by distances of 10-30 kpc in general.

If you want to find such pairs, you would write down a relevant criterion and apply it to the subhalo masses/positions.

If you want to plot such pairs, you would load the particles that belong to both, and plot them together, with different colors.

Once I have the pairs, how could I plot them together?
I am using python, but concatenating does not work, I do not know if there is any adequate process to generate the union as the author did (I wrote to the author of the article but he did not answer), well I appreciate your comments
greetings

I'd suggest to google e.g. "matplotlib scatter plot two sets of points". If you have questions about plotting, it would make more sense to ask where matplotlib is discussed, and/or follow a tutorial.

ohhh now
Yes I will try it, thanks for the suggestion