ED Astrometrics: Maps and Visualizations

The good news is that I think I was able to detect the ones that were broken in this specific way-- Off by a factor of 1000 and uncorrected with EDSM data. I'm running another records update (which will probably take all day to complete).

Yeah, visiting sites with bad data to re-upload them is extremely useful. Unfortunately it's hard for me to know what's not correct until someone reports it. I have a spreadsheet of things that were suspicious, based on a few specific searches (like landables over 15G, which of course got legitimately surpassed later). I think people have mostly cleaned those up. I'd love to add more searches to it, but again it takes people pointing things out to know what needs to be added.

Here's the spreadsheet as it currently stands:

https://edastro.com/mapcharts/files/suspicious-data.csv
 
The good news is that I think I was able to detect the ones that were broken in this specific way-- Off by a factor of 1000 and uncorrected with EDSM data. I'm running another records update (which will probably take all day to complete).

Yeah, visiting sites with bad data to re-upload them is extremely useful. Unfortunately it's hard for me to know what's not correct until someone reports it. I have a spreadsheet of things that were suspicious, based on a few specific searches (like landables over 15G, which of course got legitimately surpassed later). I think people have mostly cleaned those up. I'd love to add more searches to it, but again it takes people pointing things out to know what needs to be added.

Here's the spreadsheet as it currently stands:

https://edastro.com/mapcharts/files/suspicious-data.csv

Anything with a radius less than 137klm is suspect, are any of those left?

Oh btw I decided to clear up some local bad data systems that were between me and the bubble, the first one I reached was Pru Eurk EM-C d13-21, which was listed as a problem with surface pressure on body 1, lol, before jumping in I checked on EDSM to see what the data looked like so I could tell if it was corrected, this is what I found as the listed data;

1656682047862.png


Yep two planets with gravity, mass and radius and 0k surface temp, it's all fixed now!


I just don't know how they managed to just report that data and nothing else, I suspect maybe a program error with whatever they were reporting with at the time, it may have crashed halfway through the report and wouldn't update because it thought it was already done. Someone else who didn't report the data got the first discovery data in game but that doesn't matter, it's fixed now! Quite a few to do near the bubble so that will keep me busy for a while!
 
Last edited:
I have noticed that a lot of the bodies that have problems with atmospheric pressure in the dodgy data list also have 0k temperature, in theory it should be impossible for anything to have 0k temperature, is that also a sign of dodgy data or can some bodies in the Elite galaxy actually have 0k temperature?
 
I have noticed that a lot of the bodies that have problems with atmospheric pressure in the dodgy data list also have 0k temperature, in theory it should be impossible for anything to have 0k temperature, is that also a sign of dodgy data or can some bodies in the Elite galaxy actually have 0k temperature?
No, there's a minimum surface temperature >0K that Stellar Forge will spit out. Maybe 10K IIRC? (Planetary bodies only, stellar objects can have some very weird values, and I distinctly remember someone complaining about zero temperature stars before)
 
You're way off, the coldest planet in my personal DB is this one:

JSON:
{
    "timestamp":"2021-02-06T14:48:14Z",
    "event":"Scan",
    "ScanType":"Detailed",
    "BodyName":"Scheau Byoe AA-A h121 A 4",
    "BodyID":18,
    "Parents":[{"Null":15},{"Star":1},{"Null":0}],
    "StarSystem":"Scheau Byoe AA-A h121",
    "SystemAddress":1017281911,
    "DistanceFromArrivalLS":4839.617607,
    "TidalLock":true,"TerraformState":"",
    "PlanetClass":"Sudarsky class I gas giant",
    "Atmosphere":"",
    "AtmosphereComposition":[{"Name":"Hydrogen","Percent":73.423195},{"Name":"Helium","Percent":26.576815}],
    "Volcanism":"",
    "MassEM":1.51972,
    "Radius":10329252.0,
    "SurfaceGravity":5.677212,
    "SurfaceTemperature":0.048973,
    "SurfacePressure":0.0,
    "Landable":false,
    "SemiMajorAxis":22959216237.0682,
    "Eccentricity":0.070747,
    "OrbitalInclination":7.287722,
    "Periapsis":285.502484,
    "OrbitalPeriod":16664814.949036,
    "RotationPeriod":16667746.928975,
    "AxialTilt":-0.33865,
    "Rings":[{"Name":"Scheau Byoe AA-A h121 A 4 A Ring","RingClass":"eRingClass_MetalRich","MassMT":1701400000.0,"InnerRad":20116000.0,"OuterRad":21425000.0},{"Name":"Scheau Byoe AA-A h121 A 4 B Ring","RingClass":"eRingClass_Icy","MassMT":35756000000.0,"InnerRad":21525000.0,"OuterRad":40025000.0}],
    "ReserveLevel":"PristineResources",
    "WasDiscovered":false,
    "WasMapped":false
}
 
You're way off, the coldest planet in my personal DB is this one:

JSON:
{
    "timestamp":"2021-02-06T14:48:14Z",
    "event":"Scan",
    "ScanType":"Detailed",
    "BodyName":"Scheau Byoe AA-A h121 A 4",
    "BodyID":18,
    "Parents":[{"Null":15},{"Star":1},{"Null":0}],
    "StarSystem":"Scheau Byoe AA-A h121",
    "SystemAddress":1017281911,
    "DistanceFromArrivalLS":4839.617607,
    "TidalLock":true,"TerraformState":"",
    "PlanetClass":"Sudarsky class I gas giant",
    "Atmosphere":"",
    "AtmosphereComposition":[{"Name":"Hydrogen","Percent":73.423195},{"Name":"Helium","Percent":26.576815}],
    "Volcanism":"",
    "MassEM":1.51972,
    "Radius":10329252.0,
    "SurfaceGravity":5.677212,
    "SurfaceTemperature":0.048973,
    "SurfacePressure":0.0,
    "Landable":false,
    "SemiMajorAxis":22959216237.0682,
    "Eccentricity":0.070747,
    "OrbitalInclination":7.287722,
    "Periapsis":285.502484,
    "OrbitalPeriod":16664814.949036,
    "RotationPeriod":16667746.928975,
    "AxialTilt":-0.33865,
    "Rings":[{"Name":"Scheau Byoe AA-A h121 A 4 A Ring","RingClass":"eRingClass_MetalRich","MassMT":1701400000.0,"InnerRad":20116000.0,"OuterRad":21425000.0},{"Name":"Scheau Byoe AA-A h121 A 4 B Ring","RingClass":"eRingClass_Icy","MassMT":35756000000.0,"InnerRad":21525000.0,"OuterRad":40025000.0}],
    "ReserveLevel":"PristineResources",
    "WasDiscovered":false,
    "WasMapped":false
}

During the Alpha I did find a planet with 0k temp, but thought that might have only been the Alpha, really it was probably not actually 0k to 6 decimal places, as in your example, but low enough to show 0k after rounding. I am wondering though if these bodies with the dodgy data are actually showing 0k to 6 decimal places because all the ones with no data for atmospheric pressure show 0k, but after scanning show temps sometimes in the many hundred K.
 
Back in the days it was possible to manually add bodies to the DB, so if bodies from that time have 0K in the data it's most likely due to the game showing 0K in the system map.
There was a time when I corrected some of the bodies in Orvidius' list and from my experience I'd say that if there are multiple entries with "zero" on a body it's definitely wrong, but if it's just one field it's correct about half of the time.

Only for planets though, stars behave differently, black holes for example always have a real surface temperature of 0K.
 
Back in the days it was possible to manually add bodies to the DB, so if bodies from that time have 0K in the data it's most likely due to the game showing 0K in the system map.
There was a time when I corrected some of the bodies in Orvidius' list and from my experience I'd say that if there are multiple entries with "zero" on a body it's definitely wrong, but if it's just one field it's correct about half of the time.

Only for planets though, stars behave differently, black holes for example always have a real surface temperature of 0K.

These bodies have a lot of erroneous data though for example the system Col 285 Sector EY-O b7-3, 4 planets with 0k temp, no recordings of atmospheric pressure or even whether they have atmosphere at all. I can almost guarantee none of these bodies will have 0k temp after scanning and will report atmospheric pressure, or whether there is any, and volcanic info. What I am wondering is whether this is due to the changing data in the journal as the game developed. For instance this system was first reported to EDSM in 2016 and maybe data on planets that were non-landable at the time wasn't all recorded in the journal and therefore not passed on when the systems were scanned. Some of these systems are quite close to the bubble and may bot have been re-visited by a user reporting data since then. Of course some data, like vulcanism, may not have been present in the game at all.

A long time ago I corrected a lot of erronous data that had been manually entered, it's usually pretty obvious when it's manual data entry, for instance I corrected a number gas giants that reported very low radius int he 1000klm area, and some rocky bodies with entries like 20klm radius, those were simply entering 20klm when it should be 200klm or 2000klm. These current ones mostly appear to be data reporting from a program rather than manual because the same error is often there for each body in a system. I am seeing the exact same data error in system after system from multiple different CMDRS (taken by the report name in EDSM not the discovery name in the game), so I suspect it's something to do with the changing way data has been reported in the journal over time. Of course we also can't rule out some early manual data entry mistakes, I am sure some of these still exist in EDSM as well.

 
That's EDSM failing to reject bad data.
Some uploader seems to send bad data that overrides perfectly good data.

E.g. body 5 in that system was sent perfectly fine over EDDN back in 2019 ( https://edgalaxydata.space/eddn-loo...ournal.Scan-2019-07-10.jsonl.bz2&lineno=74506 ), but like the other bad bodies in that system it got falsely updated when "DreadedLurgy" sent false data to EDSM on 2021-12-31. Spansh seems to have a way better rejection detection as in his DB that body still is fine with data from 2019 ( https://spansh.co.uk/body/324266437168538969 ).

I guess Orvidius uses EDSM dumps without further error checking and thus getting too much bad data in his DB.
 
The good news is that I think I was able to detect the ones that were broken in this specific way-- Off by a factor of 1000 and uncorrected with EDSM data. I'm running another records update (which will probably take all day to complete).

Yeah, visiting sites with bad data to re-upload them is extremely useful. Unfortunately it's hard for me to know what's not correct until someone reports it. I have a spreadsheet of things that were suspicious, based on a few specific searches (like landables over 15G, which of course got legitimately surpassed later). I think people have mostly cleaned those up. I'd love to add more searches to it, but again it takes people pointing things out to know what needs to be added.

Here's the spreadsheet as it currently stands:

https://edastro.com/mapcharts/files/suspicious-data.csv

With Eahlstan's input above about players overriding good data with bad data, I bookmarked all the systems from that spreadsheet into the galaxy map. Assuming it was from a lot of different players I would expect a fairly random distribution of bad data, that's not what happened, it's likely most of that data was from one player, with just a couple of outliers that may be from someone else, here's a screenshot!

vPmUqLN.jpg


As you can see, probably from 3 different exploration trips from one player!
 
With Eahlstan's input above about players overriding good data with bad data, I bookmarked all the systems from that spreadsheet into the galaxy map. Assuming it was from a lot of different players I would expect a fairly random distribution of bad data, that's not what happened, it's likely most of that data was from one player, with just a couple of outliers that may be from someone else, here's a screenshot!

vPmUqLN.jpg


As you can see, probably from 3 different exploration trips from one player!
Good thinking. That's actually pretty much the result I'd expect. This really seems like a client somehow corrupting the uploaded data, and none of the major clients do that, such a bug would be noticed in short order. So it's highly likely to be a small number of players with bad clients - either some weird setup that breaks things in an unusual way, or a buggy niche client, or someone doing a bad job of rolling their own scripts.
 
Good thinking. That's actually pretty much the result I'd expect. This really seems like a client somehow corrupting the uploaded data, and none of the major clients do that, such a bug would be noticed in short order. So it's highly likely to be a small number of players with bad clients - either some weird setup that breaks things in an unusual way, or a buggy niche client, or someone doing a bad job of rolling their own scripts.

Yeah that's what most of it is, but there are a couple of really funny ones around that generate their own false data as I have just recently discovered, 1 system name, 3 systems, 2 entries in EDSM, you can imagine what problems that causes, I will write that one up afterwards, I don't know how it can be corrected unless EDSM adds a third entry for the system name. So there are indeed a few.....funny issues that aren't related to bad data being reported but simply the galaxy map being a bad boy!
 
I have started trying to visit some of the systems in the systems without main star scans list.
50 -100 systems per week.

Just doing full system scans and mapping anything worthwhile, plus a bit of mining on the side.

a couple of quick questions,
Roughly how big is a sector?
How many sectors are there?

Although the numbers of systems are dropping on the weekly list I am noticing an increase in the number of systems that have only got "ID64 SystemAdress" and no other details listed, any idea what is causing this?
 
Roughly how big is a sector?

Exactly 1280 lys x 1280 lys x 1280 lys

How many sectors are there?

11340 are known, but there are still a few that are in no DB.

Although the numbers of systems are dropping on the weekly list I am noticing an increase in the number of systems that have only got "ID64 SystemAdress" and no other details listed, any idea what is causing this?
Plotting a route creates a list of systems (and their ID64, main star class and coordinates) the route uses, some 3rd party apps send these routes to EDDN and some sites listening to EDDN store these infos in their DB.
 
Thank you

Plotting a route creates a list of systems (and their ID64, main star and coordinates) the route uses, some 3rd party apps send these routes to EDDN and some sites listening to EDDN store these infos in their DB.
So sometimes a plotted route even though not flown can generate these entries, I will stick to the easy to find systems:)
 
Sorry for replying late, I've been away from the computer a lot for the last couple of weeks. But yeah, my system pretty blindly trusts the EDSM data, and generally assumes that new data is probably better than old data. Obviously that's not always going to be right though. I haven't built up an exhaustive list of acceptable ranges for things, which would probably be needed for completely rejecting bad data. But we also sometimes get legitimate data that is far outside the known norms, such as the 45G landable, with the previous record being about 11. Anything over 10 is highly suspect, but obviously rejecting it would have been the wrong thing in those cases. So it's always going to be a prickly problem.
 
BTW, on the interactive map, I've added URI-hash encoding for the viewport location and zoom level. This allows sharing map links that are already centered on chosen locations, or make click-through links in other apps.

Using this, I've added a clickable arrow next to the sector name in the Sector Viewer, to go directly to that sector on the map. And also, the GEC POI pages will take you to the right location if you click on the POI's map image, and it will also place a pin on the map for you.

I'm planning to add an auto-completing form on the map allowing you to search for a sector name and pan straight to it as well. No ETA on that one yet though.
 
Last edited:
BTW, I have text-searching for sectors working on the interactive map now. It's entirely possible that there might be some bugs, or edges cases that don't work right. Just let me know if you spot anything that doesn't work as expected.

Eventually I might add searching for POIs too, but that will take some trickery with merging the data sets in the search. So... later. :D
 
Back
Top Bottom