Page 1 of 2 12 LastLast
Results 1 to 15 of 29

Thread: [The StellarForge Analysis] An inside look at a procedurally generated galaxy

  1. #1

    [The StellarForge Analysis] An inside look at a procedurally generated galaxy



    An Inside Look at a Procedurally Generated Galaxy


    "In less than a hundred years, we have found a new way to think of ourselves. From sitting at the center of the universe, we now find ourselves orbiting an average-sized sun, which is just one of millions of stars in our own Milky Way galaxy." - Stephen Hawking

    "Estimates are that at least 70 percent of all stars are accompanied by planets, and since the latter can occur in systems rather than as individuals (think of our own solar system), the number of planets in the Milky Way galaxy is of order one trillion." - Seth Shostak






    Greetings Commanders !


    After nearly two months of hard work on this project, thinking about the how and why, I'm happy to share with the community a project that was born during Distant Worlds.

    _____________________________________________________________________


    PROJECT OVERVIEW

    Most of you guys may be aware that except around 160 000 systems, the nearly-absolute entirety of the galaxy as it is presented to us ingame is procedurally generated. To be accurate, this simulation needed to at least include the general knowledge and physics that makes and rules the actual Milky Way as we know it. During Distant Worlds, and thanks to some fellow CMDRs, I've started to wonder how accurate this galaxy is to the real one. Thus, this project mainly focus on the analysis of such a generated heavenly body in a realtime environment.
    I'm gonna develop the whole analysis process, primary goals and scheduled journey of this project in a minute, but right before that, I'd like to raise awareness to whoever reads more that, in essence, the project is a continuous science-oriented analysis of the generated galaxy inside of the Cobra Engine, and that there is no clear end, neither "practical" purpose as a "gamer". Look at it as an ongoing scientific project.

    There is also an expedition being organized regarding this project, with the added purpose of helping with Erimus's Galactic Mapping Project, explore the 1st and 2nd quadrants, and break some records .


    This thread is also a hidden call for help ! The crazy amount of data (You'll quickly see what it really means soon enough ...) needed makes the project immensely long and time consuming. Done all alone, I've roughly determined that it would take me 10 years to achieve it, so really, DON'T HESITATE ! Even if it's for a few days, I'll be more than happy :3


    _____________________________________________________________________


    WORKING MATERIAL, PROCESS & PRIMARY GOAL


    It took me a lot of time to define an established goal to this project, as everything simply started for the "fun". As of today here it is :
    • Gather Data for any procedurally generated star system.
    • Extract some statistics from the data.
    • Gather enough data so the real Milky Way can be compared later on. Yes, you read right. Compare that with the Milky Way. To conduct a moderately accurate comparison toward what we know, this needs 160 000 stars (The amount that has been implemented in-engine). Yep. Crazy, I told you
    • After some discussion with Ziljan, get to convert a decent amount of coordinates for the zero origin being Sag A* (not Sol), and get to generate a decently working external GalMap.


    /!\ The third point raised is the real finality of the project, I'm already starting to gather intel, star catalogs and astrophysics references here, more to follow in the future. /!\


    To develop this a bit, getting in touch with Michael Brooks about the engine and espetially how those 160k stars have been implemented, from which catalogs exactly and how their physical details have been generated is running in my head. The more I think about it, the more it looks like being the best option to greatly help in the process, disregarding the fact that it may also be highly likely that this kind of infos is to be kept strictly under the NDA hat.


    As of today, EVERYTHING happens here :


    https://docs.google.com/spreadsheets...OB4tNK9C-XhEk/

    Worksheet : https://docs.google.com/document/d/1...-ZjDDpVL4CUSvo


    LOGGING PROCESS & SHEET STRUCTURE :

    To anyone willing to participate in the project, or even crazy enough to start the whole journey, just ask for edit access.
    Due to the need of exactitude in every data logged, there's ONE rule that needs to be followed. This may sound a bit military-like and really isn't meant to, I apologize in advance
    -> When logging a system, all the data ABSOLUTELY needs to be entered in the 'Systems' and 'Planets' sheet, with all it's implication. If there's something missing, don't log anything at all.


    1/ Systems Log

    This is where most of the data is entered, and also where a part of the science is hapenning. Consider it as the "core" of the spreadsheet. The only data manually entered here is in the 10 first rows, all the other ones, formatted in bright grey are automatically generated.
    Anyone familiar with the Monster Sheet ? Same thing, overall !
    I've included a "Barycenter" at 'Star #' row for one reason : Bodies orbitting more than one star. This is due to to the structure of the second sheet and how things are generated here, and as for now the only way I've found to easily record such things while auto-generating systems names + Star # in the Planets log sheet, so if anyone has other ideas, feel free to share !

    2/ Planets Log

    Here is the second part of the analysis, and it's purpose is very simple : Record all types of bodies in a registered system. As this data needs to be exact, it implies the nearly-constant ENTIRE detail-scan of a system. Except when extrapolation for a given body can be easily achieved by listening to its sound, or its look on the SysMap, one should scan everything. (Except for those lonely Asteroid Belts hahah)
    The first 4 rows are automatically, but also CONSTANTLY generated. So be sure to always enter numbers here after logging a system, even if there is nothing. With such an analysis, even nothingness is of great importance !

    3/ Planetary Details

    Optional sheet here. I've created it at first with the idea of also analyzing the data of procedurally generated worlds. I did a step back quickly after realizing that the project would become pure madness due to the HUGE amount of time already required to log things.
    I've kept it nonetheless for one main reason : The Monster Sheet. The formatting and structure of this sheet exactly follows its parent in the Prospector Log, and thus is of real help and easeness when prospecting WHILE logging system for the StellarForge analysis : chained copypasta !

    4/ Comments

    This is where you enter your CMDR name after logging data, and leave comments if the system is interesting, or has bodies with stellar properties, weird stuff, whatever ... This is also where one has to credit a CMDR when the source of the data isn't from the one entering it.

    5/ Repartition_Tables & Star_Stats

    This is where all the statistical calculation and data comparison happens.
    The repartition tables are made to sum up the amount of planets entered, and compare those toward things such as the star class/subclass or luminosity of the system they come from, as well as comparing each of their reccurence in %.
    The star stats are what they are : Several statistics about stars, such as temperature, age, mass or even habitable zone. This is also the sheet that is more likely to get new additions in the future.



    Note for those who wants to help but also keep system names safe :
    If you have checked the spreadsheet, some systems are marked as "Redacted". This happens in the case where a data provider wants to keep its data location safe until getting back to the bubble and selling it. Just copy the layout of the 4 first subsheets, and send me that copy separatedely. I will then update the master sheet with the system names all redacted until data can be released.
    _____________________________________________________________________


    Still reading ? Great, you are now as crazy as I am about this right now !



    As if things weren't crazy enough, there's a million light years expedition coming along with this project. Yep.

    EXPEDITION GOALS

    - Feed data into the StellarForge analysis during its entirety. Not any star class should be privilegied. An Y-class is worth as much as a Black Hole here.
    - Chart unexplored areas, and scout for new discoveries and PoIs for the Galactic Mapping Project.
    - Break the record of distance travelled, cash earned and time spent out in the black in a single trip.
    - And the most crazy one : Travel through ALL of the Milky Way's arms in a single trip.

    EXPEDITION MAP


    (Map courtesy of Corbin Moran & Finwen : https://forums.frontier.co.uk/showthread.php?t=220230)



    Color Coding :
    MAIN PATH
    ROUTE HOME
    BADLANDS
    INTERESTING STAR REGION


    Well, not much more to say about this as the map is pretty much self-explanatory.
    There's no set time of departure, meeting points, or timeframe for this.
    Also, all the stellar objects on the route are "waypoint of interest", and here to avoid having a feel of endlessness after so time ... Some goals to reach if you prefer. Between those, then plan is to explore as much as possible, espetially in those empty areas, so the route isn't meant to be strictly followed as per drawn on the map.



    EXPEDITION SCIENTISTS

    Forum Name CMDR Name Ship Type Jump Range Time Zone
    CMDR EfilOne EfilOne ASP Explorer 33.94 GMT+1
    Ziljan Ziljan


    _____________________________________________________________________


    Obviously, as I'm at the very opposite of a scientist IRL, there may be errors or misreads here and there, and also improvements to be made, even about this thread. Everything looks clear and shiny in my mind, but maybe not on your side, if so, please please please, tell me !

    Now it's time for some thanks. This project couldn't have seen any light without the direct or indirect help of these people (If someone is missing, hit me hahah until your name is shown below hahah) :


    - Erimus Kamzel & Dr. Kaii - Let's be honest. Without DW, and, in an extent all of your threads in the forums, I would clearly not be here today. Seriously. The reason why I went back to Elite after an 8-months break is you guys.
    - Katejina - Same goes as well, I was in awe in front of your Cosmic Compendium. I ran Elite straight after reading the whole thing hahah !
    - LordFedora - You are the one that got me interested in spreadsheeting in the first place
    - Nexolek - FOR THE DATA ! Man, I really can't thank you enough for explaining me all the magic about spreadsheets and even taking the time to sort my own troubles !
    - John Rutherford - Thanks for the science, and many more thanks for helping on the surface albedo calculation and making the transition working from the Monster to my sheets !
    - Waylander Sagan - The science talks. Also pointing me out to references and ressources to get into calculus & astrophysics more seriously. (Quick note if you read this : I ended up by getting "Astronomy I & II", by Russell, Dugan & Stewart, edition Ginn and Company, 1926. It's frickin' old and there's some outdated things, but it was the easiest to get here in the city in the end)
    - Mateusz - One of IRL bestest friends. Bruh, thanks again for the corrections regarding the astrophysics constants, and also for introducing me to the amazeness of Wolfram Alpha !
    - Ziljan - Some crazy stuff in the making ! Thanks for pushing me toward something I've dreamt for months (with the coordinates thing), and glad you've been the first one to be keen on jumping it the Mly expedition !


    _____________________________________________________________________

    DATA PROVIDERS LIST :

    - Qohen Leth
    - Arjin Fury
    - Ziljan


    A very special place for CMDR Wolzan, that provided me with the complete data dump of www.elitegalaxyonline.com ! That represent 15523 stellar objects, from which : 3958 stars, 1574 gas giants, and 9717 terrestrial planets ! It also goes along with several extensive data logs such as orbital measurments, atmospheric analysis, ring measurements ... You have my eternal recognition and respect CMDR, both for the amount of work done and providing such an amount of data so quickly.


    _____________________________________________________________________

    Right on, commanders, and ... Remember ! o7

  2. #2
    Hi, my database at elitegalaxyonline.com has over 15,000 entries now which might give you a good chunk of data to work with. Have a look and let me know if it fits with what your trying to achieve.

  3. #3
    Thanks a lot CMDR Wolzan ! I'll dive into it in the next few days

  4. #4
    OMG, just look at that route map!!!???




    I may one day lend a hand to this project. Right now I'm returning from the Western Expedition, so maybe after some rest and recuperation on Risa.....

  5. #5
    This looks.... ummm.... daunting! The scope of this project! I am in awe. My mind boggles.

  6. #6
    You sir, are a true madman, eh, explorer and scientist. I bow before your ambition!
    You can put me on a list of possible contributors. However, I'll be busy the next months with some roleplaying in ED, and trying out the 2.1 update. Once that has become less interesting, I will certainly go out exploring again. Can you send me a reminder when the 2.2 update gets delayed by FDev :-) ?

  7. #7
    @Mengy No worries, the expedition is gonna be in a constant run during at least one full year, plenty of time to hop-in
    @TheTick You might be interested to know that 3 to 4 pages of roleplay history is in the writting for this expedition ... Will be along the ranks for our .. "mission" .. with this cmdr once he's back from the other side of this galaxy. Soon enough the story of Ornas Diehl will be released. Look closely to our internal comms in the upcoming few days

  8. #8
    Utterly Insane! At some point, the shear weight of the data will collapse the entire universe!
    --
    Barking Mad, I tell ya...
    --
    Yet... for reasons I am currently unable to fully understand... Compelling!
    --
    There is only one "Monster" I can support at a time... if that one feeds yours... you have another contributing Mad Man.

  9. #9

  10. #10
    @Michael Darkmoor - You actually are raising a very valid point I hadn't thought about ! I could definitely try to see if data callback from the monster sheet can be made. Thanks

  11. #11
    16h50 update : Added details in the case of data secrecy (first discoveries), and data providers list, with very special thanks toward CMDR Wolzan !

  12. #12
    That is one scary map! it already has me shaking, and calculating, can my Anaconda store enough beer and pizza for the journey......

  13. #13
    Hahah, a capital ship may be needed to bring enough megatons of beer out there ...

    Also, thread updated !

    - A 4th goal has been defined
    - New spreadsheet contributor/data provider, and also .... first CMDR willing to jump in the Mly expedition : Ziljan !
    - New data for the spreadsheet : Approx. [x,y,z] coordinates input, [R, ϴ, Z] coord. system conversion for Saggy A* as being the origin point, Mean Density Calculation and Surface Pressure (Still some work to be done for this one, but getting there).

  14. #14
    This looks like the big daddy of all projects. There are some things I do not understand about the way the data is recorded though I will have another look through. I would love to help on this I will be recording all my data is there any tools or programs you know of that could help me do this?

  15. #15
    For now, all the data is recorded manually as it needs to be 100% exact.

    The closest WORKING tool I'm aware of so far is Wolzan's OCR engine; tried ExplOCR, and EliteOCR more recently, and they aren't reliable enough for the exactitude needed, exception made for EliteGalaxyOnline.
    Now the only barrier to me is that I'm a complete newbie when it comes to coding, and API implementation
    Hope to have this resolved soon enough so anyone willing to participate just has to take screenshots and feed it to some custom app; that will make everything way less painful and accelerate data gathering greatly !

Page 1 of 2 12 LastLast