ED Astrometrics: Maps and Visualizations

Orvidius

Orvidius
Taen said:
Looking at it now.. something odd is happening with the helium / hydrogen average, at least in some of the boxels... How is the average being calculated? An example of a problem boxel is Wepooe EA-A e, where the average is lower than both the maximum and the minimum.
Click to expand...
Orvidius said:
That's strange. Probably a stupid typo in the code, or something, or referencing the wrong variable. I'll look.
Click to expand...

OK, I think what's happening here is that there are planets that lack these elements, and so in the database I have "null". The average adds these in as zeroes, but the "min" value looks for something "defined" (that is, a true zero is OK, but not null). That's clearly confusing. I probably should skip planets with no values for those elements.
 
Last edited:

Taen

Taen
Ooh!
Orvidius said:
For boxel letter, you mean the "XX-X" or the mass code?
Click to expand...

The mass code. (Sorry, last night was rough and I'm tired. My older son tried sleeping in a separate room from his younger brother, and it did not go well. There was much knocking on doors and crying and waking.)

I'm not sure what the standard nomenclature is, but in my headcanon its:

XX-X 'Position Code' [A-H] 'mass code', then if there's a number followed by a hyphen it's that's the 'offset', followed lastly by the system number.
 

Taen

Taen
Rudi Raumkraut said:
Nice new approach to analyse the data we have :)

Which one is the most explored boxel?
Or maybe a top10 ;)
Click to expand...

Depends probably on how you define explored. Boxel sizes range from 1 system to thousands. And I'm not sure if EDSM yet outputs whether or not a system was fully scanned, and if it doesn't it's going to be difficult to work out which systems have had all of their bodies scanned. On top of that, some data such as helium percentages has only been in the journals for the last couple of years, so some older completely scanned systems will still be lacking.

One benefit of the FSS is that we're now getting more systems that are completely scanned as of December 2018, simply because the time it takes has been cut down by something like 90%. So I think the dataset has been fleshed out a bit since then.

I'll be curious to know which sectors on EDSM have the most fully scanned systems (ideally that aren't made up entirely by stars)

Next on my list is to try and find boxels that are only partially in the suppression zone around The Core.
 

Taen

Taen
Orvidius said:
I've regenerated the list, and the hydrogen/helium averages should look better now. I've also added the body total/average/min/max columns, and mass code column.

Click to expand...

Ok, looking better. There's a problem with the terrestrial body counts, though. Using Oephail LG-Y e as a test case, you've got 36 total planets, and I've got 408 (including gas giants.. 401 without them). Since I scanned that one myself, I can be confident about my numbers. Our star tallies DO match, however. The most populated system has 23 bodies.

Edit: Checking another boxel. Scaulua EC-U e3- has 373 planets on mine, 229 on yours.
 
Last edited:

Orvidius

Orvidius
Ah, I know what's happening. It's due to the table join with atmospheres, so it's skipping planets where atmosphere data is absent (probably due to being airless). Shouldn't be hard to fix.
 

Taen

Taen
Looking better and better! One more.. It looks like Max Bodies is showing an inaccurate result. Using Oephail LG-Y e as my usual example, it's showing max bodies as 7. It should be 23. (There are several systems above 7)
 

Taen

Taen
Marvelous! And now I get to try out something I've wanted to check with a larger dataset.. Boxel Helium level vs. the bodycount of the system with the most objects per boxel.

138688


Edit: Sorry for earlier lack of context. Parenting was... something of an adventure today. I'm not good at graphs when I spend a decent amount of time at them, which I obviously didn't here.

So this is from @Orvidius 's data as posted earlier today that we've been ironing out the wrinkles on. Each dot represents a boxel with mass code E-H Vertical (ordinate) represents the highest number of bodies of the most populated system in a given boxel. Horizontal (abscissa.. I get to learn a new word today) represents the helium level.

The assumption is that each boxel has a helium range, which shows pretty consistently in the data. The higher the helium, the wider the range.(Eg. if you find a system with 23.9% helium, the range is going to be somewhere around +/- .02%, whereas at 30% it'll be around -2% to +3%... meaning if you know nothing about a boxel except for the helium percentage of the system you are in, another system might be as low as 28% (if your system represents a high level for the boxel) or as high as 33% (if your system represents a low level). As you find more systems with gas giants, the range can be more closely worked out. Systems without gas giants are assumed to be within the helium range, but cannot be used to calculate it for obvious reasons.

The body count includes all bodies in the system (excluding asteroid fields / rings) including stars and the helium containing body.

The point of all this.. Anecdotally I've noticed that systems with higher numbers of bodies tend to be in boxels with higher helium levels. I've noticed this anecdotally for some time, but didn't have access / patience to parse the EDSM dataset to see if the existing data supports this or not. I'm not sure that this completely supports it either, since there's an awful lot of incomplete systems in EDSM. But I've been scanning full boxels to flesh out the data somewhat, and so far it appears to... at least on the low helium levels. I have yet to find a system below 24% that had more than 23 bodies. (I've fully scanned a bit over 800 so far)

I suspect that the 'preferred band' of 0-20 bodies is more to do with people preferring to scan high value objects.

I am once again out of time, but hopefully I'm at least somewhat coherent here.
 
Last edited:

schlowi123

S
Sorry for complaining, but come on, you can do better than just throwing a graph at us.

Which attribute is on which axis? (I guess Helium content (?) is on the abscissa).
What are the units of the axis?
Is the body count _ including _ the helium containing body?
What did you expect or what does this show us?
Are the highest values relevant or the distribution?
I guess the structures in the beginning are just the usual "it's something in the code" and not real. But there seems to be a preferred "band" between 0 an 20 bodies (IF the ordinate represents the body count). Did you expect that? If yes, why?
The maximum of the "main" distribution seems to be around 26 Helium content (unit???). But the maximum of the "halo" seems to be around 27. But that may be wrong. I didn't fit that that was just from eyeballing.
The distribution is not symmetrical. Is there a reason for that?

For other things I've seen that many of the answers come down to "it's something in the code". Which is rather boring in the end. But it seemed you had a hypothesis and I'm rather curious about that :)
 

Taen

Taen
Yes, I can. I had all of a few minutes to play with the data before I had to return my attention to the kids (much like now). I'll elaborate later.
 

schlowi123

S
That's a great explanation for the graph :)

And I understand the parenting thing. Mine is old enough now to not needing "constant" attention any longer. That goes very well with gaming (unless he wants the PS4 for himself).

So you are saying that the helium content is (somewhat) quantized (plus/minus around the quanitzed values and the plus/minus range is dependent on the value). That is interesting indeed and may be a good explanation for the "pillars" at low helium contents.

Low helium content and low body count makes kind of sense. All the "metals" (and the helium) need to be produced in first, second, third etc. generation stars. So lower helium content would mean earlier generation, thus less "metal" at all.
 

Maolagin

M
Probably a silly suggestion, since the locations won't line up all that well, but I kind of want to have the Gaia "topographic" map as an overlay layer. It does neatly capture the major features of our corner of the galaxy as known to actual science.

www.cosmos.esa.int

IoW_20180614 - Gaia - Cosmos

www.cosmos.esa.int www.cosmos.esa.int

Vaguely related, it would be useful to have traffic heatmaps with different transfer functions on the intensity, so it's possible to make out structure in the saturated blob near the bubble.
 

Orvidius

Orvidius
Neutrons already have a separate map, and the white dwarfs are in the "Dwarf Stars" map too... but it's true, there's no separate one for WDs. I could probably add that.
 
Last edited:

Orvidius

Orvidius
I have the white dwarf map up. Also working on the alternate heat map color scale, using indexed color ranges rather than a purely logarithmic scale. It's already bringing out a lot more structure, but I'm still making tweaks. Test runs take 1-2 hours when using the full data, so it's slow going. :)
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom