ED Astrometrics: Maps and Visualizations

[Orvidius]

So it was visited and reported with data like the main star, then later purged from EDSM after you downloaded your data? I wonder what criteria EDSM uses to purge data, did that happen when they moved everything to the new servers I wonder, because there was a bit of hold up there, would have been a huge task at the time and they might have been trying to minimise the amount of data they were moving around.

EDSM was doing it for a few months before the move to the new server. Anthor was trying to prevent the server from completely running out of space. Eventually he did that server move, and and was able to stop purging data, but quite a bit was already purged at that point. Also, during the server move there were a few days where EDSM was down and not receiving data, so I probably have some data from those days via EDDN that never even made it into EDSM at all. That would be a small amount though. The vast majority of the "missing" data on the EDSM side is just due to the purging, and it's still a small minority of the data compared to the whole.

Ok done, a carrier jump with a CMDR on board will record the stars and asteroid belts and a first visited tag to EDSM, so systems with no data can't be attributed to FC's, an empty FC won't report any data, and a CMDR on board will record at least the main star and a first visited tag!

Right, these would have to be one of two things-- systems that were merely jonked in the pre-FSS days (system exists in the data, but no body data submitted), or cases where EDSM purged the body data, but not the system. I was under the impression that it would all have been purged, but maybe it was just the bodies, since that's where the bulk of the data storage footprint is located. I wouldn't be surprised if EDSM prioritized keeping the systems even if it was purging the bodies.

 

[Deleted member 115407]

The hell are you nerds talking about?

My little bell thing keeps popping off.

 

[HeIlrider]

Nerd stuff is the best stuff

 

[Orvidius]

What are we talking about? Nerd stuff. Obviously.

 

[Orvidius]

Well, if you insist
(I always fear that you start hating me for reporting too many things.)

-snip-

Same for highest, same for Class II, Class III, ... Guess it's a thing that affects all bodies where you go for overall - moon - planet.
It also affects other stats e.g. gravity, radius, ...

Sorry, forgot to reply to this part. Yep, I'll probably work on a small amount of the discrepancies at a time, particularly since some of the lookups can take a while. But yeah, the high/low values are always at the mercy of outliers in the data that mostly originated in old EDSM data, where the planet type was incorrect, or a duplicate was entered with bad values, or there was some sort of units conversion applied, etc. Most of the main floating point numbers are back to the old EDSM values from the dumps, because I changed how I was storing it (decimal instead of float), and wanted to recover the available precision. But of course that meant getting old broken values back in there as well.

One of the things I'm trying to do is find where the legitimate numbers end. A realistic cutoff value that I can do a lookup against, and re-sync those systems from EDSM via the API. And then whatever doesn't get cleaned up that way, I can set the values to null, or delete the body altogether if it's clearly a duplicate or something.

 

[Eahlstan]

Sorry, forgot to reply to this part.

Your previous answer covered this pretty well.
These things are all caused by your "is a moon", "is a planet" and "I don't know what this is" script, and since you're "not worried about that part" the overall records and the records for "planet" and "moon" will always be erroneous if one of the overall record bodies is part of the "I don't know what this is" category.

 

[Orvidius]

Right, those records will always apply to the "definitely a moon" or "definitely a planet", but the ones that have handcrafted names or whatever will only count toward the entire body type's category. So it's really just a caveat to have in mind when looking at the records. The moon/planet categories will cover procedurally generated systems and catalog systems that followed the internal naming conventions, but not the hand-crafted stuff from Sol, for example.

 

[Eahlstan]

Sorry to bring this up again, but I just wondered how you determine if it's a "planet" or a "moon".
I get that it's nearly impossible to tell for named bodies like "Earth" or "Tira Flirble", but for every other body - no matter if it's a star or not - looking at the last character (or last two to see if there's a space) should be sufficient to tell. If it's an upper-case letter (A, B, C, ...) it's 100% a star, if it's a number (1, 2, 3, ...) it's what you chose to call "planet" and if it's a lower-case letter (a, b, c, ...) it's what you call a "moon" - "moon-moons" are "a a" or "a a a", so nesting is no problem for the "last character"-approach.
If there's a way to recode your script to do that bodies like "2MASS J05120458+1640267 A 4" (class I giant with the highest mass) wouldn't vanish in the "I have no idea"-abyss and only named bodies will spend their life there.

 

[Orvidius]

Yes, that's exactly how it works.

EDIT: Actually, apparently I did have it look for normal proc-gen formatting, and skipping the catalog systems.. Doh! Yes, I can probably improve that.

EDIT 2: This is when I need to do a forehead slap. I thought for sure it was including the catalog systems, but I forgot that I had made those regex patterns more restrictive. LOL

 

[Orvidius]

OK, I hacked up those regexes, and kicked off another run. It'll complete overnight.

 

[Orvidius]

The records have updated on the website. It should reflect a lot of the recent corrections, though I'm sure there's more to fix. (some corrections are still running or are already fixed, such as "NGC 2168 BSB 5203 3 b" having a stupidly high radius for an Icy body).

 

[varonica]

See my new record? 45g landable planet, surface temp of over 5,000k!

 

[Orvidius]

See my new record? 45g landable planet, surface temp of over 5,000k!

Truly amazing! I can't wait to do a splatter test there.

EDIT: I love that its catalog number is 1701, like the Enterprise.

EDIT 2: I'm re-updating the galactic records so that this will show up. It'll probably run overnight and finish sometime tomorrow.

 

[Orvidius]

LOL, the records were just finishing up, but I have to start it again. I just remembered that I had a sanity check in there, and it was ignoring landables over 20g of gravity. Clearly that was no longer a useful test.

 

[varonica]

LOL, the records were just finishing up, but I have to start it again. I just remembered that I had a sanity check in there, and it was ignoring landables over 20g of gravity. Clearly that was no longer a useful test.

You used a sanity test in ED? Was that sane?....sorry, silly question I know!

 

[varonica]

Hmm, that's a little bit annoying!

After my discovery of KOI 1701 1 other players have suddenly jumped in and started visiting KOI catalog stars maybe hoping to find a similar body. My huge list of bookmarks of unvisited KOI catalog stars I had built up to visit are being farmed through and now I have to check all my bookmarks before I set off. That's a lot of work I spent going through all those stars and bookmarks...sigh.

Still I don't own them so I guess that's what I get for discovering stuff!

Not sure what to do now, just resume where I am, race around to get as many as I can or just plod on and hope!

 

[Orvidius]

Heh! Yeah, that's always the danger of finding something. Suddenly everyone wants to jump in. The thing is, it's in a catalog system but I'm pretty sure it's entirely procedurally generated, so the chances of finding something like this in another KOI system aren't any different than exploring anywhere else. Either people aren't realizing this, or it's suddenly drawn their attention to the fact that they can still get their name on catalog systems.

 

[Maolagin]

Heh! Yeah, that's always the danger of finding something. Suddenly everyone wants to jump in. The thing is, it's in a catalog system but I'm pretty sure it's entirely procedurally generated, so the chances of finding something like this in another KOI system aren't any different than exploring anywhere else. Either people aren't realizing this, or it's suddenly drawn their attention to the fact that they can still get their name on catalog systems.

I'd be hard pressed to back this up with data, but I feel like catalog systems are somewhat more likely to have odd anomalies like this. I wonder if their positioning being different from where SF would have put a system throws the procgen off its game somehow.

Of course, if that was true, you'd probably expect a lot more anomalies in catalog clusters, where the galaxy mass function definitely would not have created so many big blue stars in one place. So who knows.

 

[Orvidius]

As far as I know, that's basically it-- the catalog systems are mostly proc-gen, but with predetermined position and mass, and possibly an override for what the primary star type should be and some other features (to match the catalog information, of course). So anomalies are likely to occur in the sense that the mass and star choice might not match what the StellarForge might have selected otherwise, skewing it toward proprortionately more or less massive stars compared to the planetary mass. I think this is why we see a lot of planetless stars, and stars with other stars in planetary orbits, compared to what we see in non-catalog systems... if that makes sense.

Sadly, we can really only speculate based on what details have been released, and from experience in exploring these systems.

 

[varonica]

I'd be hard pressed to back this up with data, but I feel like catalog systems are somewhat more likely to have odd anomalies like this. I wonder if their positioning being different from where SF would have put a system throws the procgen off its game somehow.

Of course, if that was true, you'd probably expect a lot more anomalies in catalog clusters, where the galaxy mass function definitely would not have created so many big blue stars in one place. So who knows.

Most KOI stars are actually not big bright stars, in fact KOI 1701 seems to be an exception to this, very few are in the O B A grouping, mostly F and G range, but that makes sense since it would make it much harder spotting a transiting body with a very bright star! The thing that really interests me about the KOI group is one thing mainly, their almost unique pattern in the galaxy, I can only imagine the observations taken were all aligned along a very small patch of the night sky.

Since people seem to be farming them anyway I guess there's no point in keeping it secret, but here's my rough map of KOI distribution using in game bookmarks!

Spoiler

As you get further away from the bubble the KOI catalog stars angle out of the galaxy plane at a fixed angle, the end point of my KOI hunt was going to be to see how far along that angle I can chase them, maybe even achieve a galactic high point. However now that people have seemingly jumped in and started harvesting all the KOI stars close tot he bubble I have decided to start my chase along that high path right away. I will keep you filled in on the progress!