raptop changed the topic of #kspacademia to: https://gist.github.com/pdn4kd/164b9b85435d87afbec0c3a7e69d3e6d | Dogs are cats. Spiders are cat interferometers. | Космизм сегодня! | Document well, for tomorrow you may get mauled by a ネコバス. | <UmbralRaptor> egg|nomz|egg: generally if your eyes are dewing over, that's not the weather. | <ferram4> I shall beat my problems to death with an engineer. | We can haz pdf
UmbralRaptor is now known as GeodesicRaptor
e_14159 has quit [Ping timeout: 190 seconds]
e_14159 has joined #kspacademia
<egg|zzz|egg>
!u ;·
<galois>
;: U+037e GREEK QUESTION MARK
<galois>
·: U+0387 GREEK ANO TELEIA
<egg|zzz|egg>
!u ;·
<galois>
;: U+037e GREEK QUESTION MARK
<galois>
·: U+0387 GREEK ANO TELEIA
<egg|zzz|egg>
!u ;·
<galois>
;: U+003b SEMICOLON
<galois>
·: U+00b7 MIDDLE DOT
<egg|laptop|egg>
population of 10, +4.89604291557538573e+01 m at generation 256
<egg|laptop|egg>
mofh: deмcmc is nice
<egg|laptop|egg>
i should really templatize it so we can have an optimizer,
<egg|laptop|egg>
population of 20, +5.34505716919805209e+01 m at generation 128
<egg|laptop|egg>
and +4.86118847201334034e+01 m at generation 256
* egg|laptop|egg
tweaks the initial population a bit
<mofh>
Isn't that similar to I Can't Believe It's Not Gradient Descent, which gave you residuals of ~50m?
<GeodesicRaptor>
… why are you calculating -∇margarine?
<egg|laptop|egg>
mofh: it's faster i think
<egg|laptop|egg>
mofh: or maybe it's only faster because of the parallelism,
<egg|laptop|egg>
probably actually faster though, my laptop doesn't have that much concurrency
<egg|laptop|egg>
but yeah, I have yet to get anything below 47 m
<egg|laptop|egg>
on G01
<egg|laptop|egg>
also the Bc and Bs coefficients look somewhat absurd
<mofh>
Oh?
<egg|laptop|egg>
my burn-in phase might be too long
<egg|laptop|egg>
no that's not it
<egg|laptop|egg>
it just wants to explore that part of the parameter space
<egg|laptop|egg>
hm
<egg|laptop|egg>
actually are they absurd or not
<egg|laptop|egg>
mofh: can you find papers with reduced ECOM parameters other than D0 for GPS satellites
<egg|laptop|egg>
specifically whichever kind of satellite is at PRN G01 these days
<egg|laptop|egg>
mofh (in a sense it's a lot easier to get posteriors from an МCMC algorithm than it is to get fitting residuals) << no i am not trying to *get* residuals
<egg|laptop|egg>
I have them, because that's what I'm optimizing, and I need posteriors, otherwise I have no мcmc to speak of
<mofh>
Aren't you getting your fits *from* the МCMC sampling of your parameter space?
<egg|laptop|egg>
what's the acceptance
<egg|laptop|egg>
the acceptance involves the posterior for the trial state
<egg|laptop|egg>
i need to conjure that up
<mofh>
Ahh. Hmm.
<mofh>
I see what you're getting at now.
<egg|laptop|egg>
for the TRAPPIST optimization iirc least squares coefficiented by the standard uncertainties of the observations gave some χ2 from which the log of the PDF was -χ² / 2.0
<egg|laptop|egg>
(the posterior probability is unnormalized, that's fine)
<egg|laptop|egg>
I guess I could do something similar
<egg|laptop|egg>
how the hell is it 4:52
<egg|laptop|egg>
i should sleep
<mofh>
Like I had assumed you did something similar from the standard uncertainties here, hence my confusion above.
<egg|laptop|egg>
what standard uncertainties
<mofh>
...oh, right, these are well-calculated orbits b/c we're dealing with earth sats, not eggsoplanets.
<mofh>
OKAY.
<egg|laptop|egg>
well
<egg|laptop|egg>
they could have standard uncertainties
<egg|laptop|egg>
there are no EP records
<egg|laptop|egg>
so they don't give any standard uncertainties (nor correlations)
<mofh>
That... kinda doesn't surprise me tbh.
<egg|laptop|egg>
nobody does
<egg|laptop|egg>
they do give accuracies for the whole satellite
<egg|laptop|egg>
okay that actually does give a scale "A satellite's accuracy exponent appears in the same columns within these lines as the identifier on the satellite ID lines. The accuracy is computed from the exponent as in the following example. If the accuracy exponent is 13, the accuracy is 2**13 mm or ~ 8 m. The quoted orbital error should represent one standard deviation and be based on the orbital error in the entir
<egg|laptop|egg>
oh also
<egg|laptop|egg>
"On Line twenty-three, columns 4-13 would hold the floating-point base number used for computing the standard deviations for the components of the satellite position and velocity. Instead of using 2**nn as is done in the orbit accuracy lines 12-20 in the header, better resolution can be attained Page16 using a number like 1.25**nn. The units for position and velocity are mm and 10**-4 mm/sec, respectively."
<egg|laptop|egg>
why have significands when you can vary the base instead
<egg|laptop|egg>
mofh: but since it's orbit-wide (unless there are EP records, which there are not) it doesn't really matter for demcmc purposes
<egg|laptop|egg>
since I only need unnormalized posterior probabilities
<egg|laptop|egg>
mofh: actually since you have them all, are there any EP records in the SP3 files of MGEX?
<egg|laptop|egg>
(it would be more interesting to look in the whole of IGS but I don't think you have that yet)
<egg|laptop|egg>
!choose fix the deмцмц|ιτ's five in the morning sleep you idiot
<galois>
egg|laptop|egg: Your options: fix the deмцмц, ιτ's five in the morning sleep you idiot. My choice: ιτ's five in the morning sleep you idiot
<egg|zz|egg>
03:08 <egg|laptop|egg> mofh: actually since you have them all, are there any EP records in the SP3 files of MGEX? 03:08 <egg|laptop|egg> (it would be more interesting to look in the whole of IGS but I don't think you have that yet)
<egg|zz|egg>
mofh: that would be EP in columns 1..2 of an SP3 file
<egg|zz|egg>
mofh: alternatively are there files whose P records (P in column 1) are more than 62 columns long
egg|zz|egg has quit [Ping timeout: 180 seconds]
egg|cell|egg has quit [Ping timeout: 198 seconds]
egg|cell|egg has joined #kspacademia
<egg|cell|egg>
I think there are three ways to give standard deviations on orbit positions...
<egg|cell|egg>
ГЛОНАСС?
<galois>
ГЛОНАСС: ГЛОбальная Навигационная Спутниковая Система
<egg|cell|egg>
« annexe : le retard dans la programmation des missions spatiales »
egg|cell|egg has quit [Ping timeout: 189 seconds]
egg|cell|egg has joined #kspacademia
egg|cell|egg has quit [Ping timeout: 198 seconds]
egg|cell|egg has joined #kspacademia
egg|cell|egg has quit [Ping timeout: 183 seconds]
egg|cell|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
<egg|work|egg>
!seen mofh
<galois>
egg|work|egg: I last saw mofh at 2019-04-15 - 02:54:55 in here, saying That... kinda doesn't surprise me tbh.
<egg|work|egg>
!acr -add:GINS Global Indian Navigational System
<galois>
Definition added!
UmbralRaptop has joined #kspacademia
GeodesicRaptor has quit [Ping timeout: 202 seconds]
<egg|work|egg>
mofh: meow
<_whitenotifier-3d18>
[Principia] MorePortal closed issue #2134: Random Crashes - https://git.io/fjmoy
<mofh>
the back roof is entirely gone and the spire is completely engulfed in flames
<mofh>
merde indeed :/
<mofh>
and the spire collapsed
<mofh>
I'm going to step away from the internet before I start crying in my office, gah.
<egg|zz|egg>
mofh: happens to things that are under construction works *a lot*
<egg|zz|egg>
mofh: see also la Fenice
<egg|zz|egg>
huh, they concluded that la Fenice was arson eventually
<mofh>
egg|zz|egg: I mean, I'm not surprised, flammable materials are often used in restoration and things are often very haphazardly done while reconstructing things
<egg|zz|egg>
mofh: also can impede the intervention of the fire brigade
<mofh>
egg|zz|egg: plus often these construction works on monuments like this happen *because* they are badly needed to fix long-standing safety issues
<mofh>
and yeah, that too.
<egg|zz|egg>
(in particular in venice, where they drained adjacent canals and the fire brigade is on boats)
<mofh>
(yeah that sounds particularly awkward)
<mofh>
oh gods the west rose window is gone?! I *really* hope that's not true. :/
<kmath>
<✔corvidresearch> Cassowaries are ~150lbs. Velociraptors were about the size of turkeys. Honestly, humans don’t spend enough time bei… https://t.co/Cg0x8ILzjZ
<kmath>
<✔teatrolafenice> 🎭 We burnt twice but twice we have risen from our ashes stronger. We are at your side, friends, so fear not! 🎭… https://t.co/1PqDJ3hZjS
<kmath>
<Mikeggibbs> Right. ⏎ ⏎ So, in case there was any doubt, historical buildings require major regular investment. ⏎ ⏎ Swallow it with pu… https://t.co/iyuV3kIB6R
<egg|zz|egg>
mofh: tbh afaik it was rather well maintained (see the maintenance work, or the fact that they were alerted by fire alarms early enough to evacuate)
<mofh>
I mean I'd hope that they were alerted early enough, esp. since from what I gather there were people inside and even a mass being held at the time.
<egg|zz|egg>
mofh: there generally is a mass but the bulk is tourists
<mofh>
I mean there is a reason why I actually didn't go in there at any time during my Paris stay, it was congested as fuck always.
<egg|zz|egg>
mofh: but I don't think it's a case of dilapidated building catching fire like the one in brazil
<egg|zz|egg>
(to start with, restauration works tend to be a good sign of maintenance,)
<mofh>
(I went in 2006 with my parents, so it's not like I'm unfamiliar with the inside; Actually I recall it being one of the things I liked from my trip there).
<kmath>
<incunabula> There are still conflicting reports on the fate of the Cathedral's three great Rose Windows. Some reports in the la… https://t.co/qDBrvxL2Z0
<kmath>
<✔teatrolafenice> 🎭 Siamo stati assaliti dal fuoco, ma dal fuoco siamo risorti più forti di prima. Siamo al vostro fianco, amici, non… https://t.co/ZtUlKsVQgB
<mofh>
I don't know any Italian, so :P
<egg|zz|egg>
mofh: oh come on it's transparent wrt french,
<egg|zz|egg>
mofh: and, given that english is basically french, by transitivity,
<kmath>
<incunabula> Journalist @JulieBrafman was able to photograph the drone-feed of the French police. https://t.co/niBwU6ACYu
<egg|zz|egg>
mofh: apparently firefighting is complicated by molten lead falling down (the roof is lead)
<egg|zz|egg>
or was, rather,
* UmbralRaptop
screams internally
* egg|zz|egg
meows at UmbralRaptop
<UmbralRaptop>
vaporized lead…
<UmbralRaptop>
hrm
<UmbralRaptop>
egg|zzz|egg: English is a mixed state of German, French, and probably half a dozen other languages. This also eggsplains the probabilistic nature of the spelling.
<egg|zz|egg>
UmbralRaptop: spelling, from Old French espel(l)er (compare Modern French épeler), from Frankish *spelôn, from Proto-Germanic *spellōną (“to speak”).
<egg|zz|egg>
germanic, but via french,
<egg|zz|egg>
UmbralRaptop: épeler retains the meaning of "read letter by letter", e.g. "S-P-E-double L", but not the meaning of "how a word is written" (that would be orthographe)
<mofh>
egg|zz|egg: rofl
<mofh>
"Germanic, but via French" sounds like a perfect description of English,
<kmath>
<MotsSurannes> "Je ne vais pas attendre 107 ans" est une expression surannée que nous devons à #notredamedeparis ⏎ C'est le temps q… https://t.co/TrZBzEXGNm
<kmath>
<✔RaphaelleBacque> Dans la nef, encombrée par les restes de la flèche en miettes, l’autel est resté intact, avec sa croix#NotreDame https://t.co/yify3ERdN0
* egg|zz|egg
stares at ref_url
<egg|zz|egg>
mofh: and you get a free link to le monde apparently :-p
<kmath>
<mikegalsworthy> This gets you in the heart. It’s sadly ancient and beautiful. ⏎ ⏎ Parisians singing as they watch the Notre Dame cath… https://t.co/TvyrR9mdMY
<mofh>
egg|zz|egg: yeah I was listening to that earlier
<mofh>
also how did you know my main worries were the stained glass and the organ? :P
<mofh>
(answer: I am a stereotype of myself it seems like :P)
<kmath>
<incunabula> The Vicaire général de l’archidiocèse de Paris has just said: "L’orgue a été détruite et le plomb des vitraux a fon… https://t.co/aTxKvPJlfq
<mofh>
Well fuck.
<mofh>
I mean I was honestly sadly expecting that to be the case, pipe organs like that are fairly fragile and it was located in one of the more on-fire bits of the cathedral.
<egg|zz|egg>
mofh: not sure what their source is for that quote
<kmath>
<incunabula> There is now an apparently reliable report from @Valdiguie, a French blue-tick journalist on the scene, that the No… https://t.co/YtIvTYwGzR
<kmath>
<ClementSalviani> Il y a une certaine preuve du génie des maçons et architectes médiévaux dans ce cliché : seul le voûtain / le hourd… https://t.co/mf8MBT7TrU
<kmath>
<ClementSalviani> (J'avoue un authentique kink pour le lexique de l'architecture. Archivolte, bandeau, voussoirs et coussinets, douel… https://t.co/73S0adOa5W
<egg|zz|egg>
mofh: communication about the damage to the glass is a bit cahotic...
<egg|zz|egg>
chaotic even
<egg|zz|egg>
mofh: hm, funny that it's chaotic in en but cahotique in fr
<egg|zz|egg>
wait no, both exist in fr,
<B787_300>
huh one interesting thing about the Notre Dame fire will be the renewed interest in gothic architecture
<egg|zz|egg>
B787_300: also if you like the structural engineering stuff, this is reinforced stone: there is an underlying metal structure (goes through the voids where the glass panes are as well) https://twitter.com/ClementSalviani/status/1117934337096286212
<kmath>
<ClementSalviani> Vrai ! L'artisan invisible de la cathédrale gothique est le forgeron, le marteau hydraulique et l'acier qu'ils prod… https://t.co/XuI1vKabiv
<mofh>
B787_300: this is a feature as far as I'm concerned.
<egg|zz|egg>
cc whitequark: civil engineering of sorts ^
<mofh>
23:27:50 <fibmod> what this video of people singing fails to capture is the smell of weed
<mofh>
23:28:08 <fibmod> make me giggle a bit
<mofh>
23:28:17 <fibmod> most surreal place to chill
<egg|zz|egg>
hah
<B787_300>
i wonder if later we will be able to trace an architechural resurgence of gothic and vaults to this fire
<kmath>
<hseckel> Ambiance étonnante sur l’île Saint-Louis, entre les uns qui applaudissent les pompiers, et les autres qui commanden… https://t.co/C9GOMW7Vec
<egg|zz|egg>
(when was the last time i went to Berthillon)
<mofh>
egg|zz|egg: I wandered in there, took one look at the prices, audibly gasped and went to a nearby Picard.
<egg|zz|egg>
hah
<egg|zz|egg>
!acr -add:ABF Architecte des Bâtiments de France
<galois>
Definition added!
<egg|zz|egg>
!acr -add:ACMH Architecte en Chef des Monuments Historiques
<kmath>
<ClementSalviani> @kinkybambou Tmtc, donc, le jeu des poussées est totalement fragilisé si y'a un trou dans le bordel. Y'a des ABF et… https://t.co/TuidgAbgrH
<kmath>
<incunabula> "Sur le sommet de la galerie la plus élevée, plus haut que la rosace centrale, il y avait une grande flamme qui mon… https://t.co/xCVC8UMi6A
<B787_300>
dont you also normally build the vault on a MASSIVE scaffold?
<kmath>
<JBAffre> Mon père me partage son hypothèse sur la cause de l'incendie de #NotreDame de Paris, fréquente selon lui dans les v… https://t.co/IlwcGyELIN
<egg|zz|egg>
yes, i saw the Hugo quote
<B787_300>
egg|zz|egg: what is the theory?
<B787_300>
for those of us that dont speak french
<egg|zz|egg>
B787_300: ah yes, apologies
<egg|zz|egg>
B787_300: in roof restoration, one welds bits of zinc that are on the wooden beams; they can be brought locally to 270 °C; even in the absence of air, pyrolysis starts, and continues because exothermic; it progresses within the wood, until it reaches a part exposed to the air
<egg|zz|egg>
and at this point smoke filled with free radicals catches fire immediately; this explains delay (after the works) and sudden start (no smoke until it erupts)
<B787_300>
hrm one thing that strikes me as odd is that wood normally has a higher pyrolysis temp... although that might be for fresher green wood
<B787_300>
like IIRC it was in the 400 C range
<egg|zz|egg>
B787_300: it has ceased to be fresh a few centuries ago :-p
<kmath>
<631_Archi> @ClementSalviani Sans compter qu’il va être impossible de conforter quoi que ce soit en toute sécurité avant d’avoi… https://t.co/rfQSm9WdeE