UmbralRaptor 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
<_whitenotifier-cd19>
[Principia] sthalik commented on issue #2324: Suggestion: don't remove the last manœuvre node from the navball - https://git.io/JeZ9b
<kmath>
<CalicoGame> While adding hats to the character creator we asked ourselves how would they work with putting cats on your head? ⏎ W… https://t.co/jKghIvHWfW
UmbralRaptop has quit [Remote host closed the connection]
UmbralRaptop has joined #kspacademia
e_14159_ has joined #kspacademia
e_14159 has quit [Ping timeout: 378 seconds]
<B787_300>
UmbralRaptop: why is DC so muggy at night
<UmbralRaptop>
B787_300: It just wants you to feel at home
<B787_300>
with its warm embrace?
<UmbralRaptop>
yes
<UmbralRaptop>
Alternatively, insert swamp joke here
<UmbralRaptop>
!8 make the obvious joke about the new Principia release?
<galois>
UmbralRaptop: no
<B787_300>
UmbralRaptop: also L'Enfant Station needs better signage
<UmbralRaptop>
Somehow that figures
<B787_300>
and yes the 7000 series trains are much much better than the 3000 series trains
<UmbralRaptop>
This is a 7000 series train. Ｔｈｉｓ ｉｓ ａ ７０００ ｓｅｒｉｅｓ ｔｒａｉｎ． This is a 7000 series train.
<B787_300>
took me like 5 mins to figure out which of the 3? exits i should have used
* UmbralRaptop
has also taken the wrong exit from some stations before
e_14159_ has quit [Ping timeout: 378 seconds]
<UmbralRaptop>
"With the growth in the population in this regime, the driest region in the brown dwarf desert (35−55MJsini) is reforesting and its mass range shrinking."
<UmbralRaptop>
… is there an #arXivfinds hashtag?
<whitequark>
yes
<whitequark>
i've used it
<UmbralRaptop>
ooh
UmbralRaptop has quit [Remote host closed the connection]
UmbralRaptop has joined #kspacademia
UmbralRaptop has quit [Ping timeout: 202 seconds]
UmbralRaptop has joined #kspacademia
e_14159 has joined #kspacademia
<egg>
UmbralRaptop: oh Fibonacci is the רֹאשׁ הַשָּׁנָה release
<egg>
and del Ferro was the رأس السنة الهجرية one
<egg>
!wpn whitequark
* galois
gives whitequark a 𓌪
<egg>
!choose aaaaaaaa|aaaaaaaa
<galois>
egg: Your options: aaaaaaaa, aaaaaaaa. My choice: aaaaaaaa
<e_14159>
So, apparently one or more papers of mine have been accepted to a workshop. I don't know which paper and which workshop, though.
icefire has quit [Quit: icefire]
<SilverFox>
yall have prepared me for this moment, and i triumphed
<SilverFox>
I have learned that Tensors are stuff that flow in TensorFlow
<mofh>
!u 𝐼₁₁
<galois>
𝐼: U+1d43c MATHEMATICAL ITALIC CAPITAL I
<galois>
₁: U+2081 SUBSCRIPT ONE
<galois>
₁: U+2081 SUBSCRIPT ONE
<UmbralRaptop>
. o O (Using a hairdryer to reflow tensors)
<UmbralRaptop>
OH: "a small nuclear reactor that brews coffee using a small core of Am-241
<UmbralRaptop>
we don't recommend drinking the primary-loop coffee, it's too hot"
<mofh>
egg: yeah I am likely going to use that code (or part of it, at least, for K/Kcomp) (with citation).
<egg>
mofh: the Fukushima code or the phl code
<egg>
the Fukushima code is in FORTRAN which might be an advantage in an academic codebase
<egg>
otoh it has mutable global state and does I/Os for no good reason so
<mofh>
egg: the phl code. I can convert C++ back to Fortran, but your C++ is much more readable than Fukushima's Fortran, and also more bug-free.
<egg>
mofh: well, Fukushima updated it recently, hence the attribution
<mofh>
(also I'm using my own code for elliptic-nome anyway).
<mofh>
also like it seems your code for K is just a bunch of Taylor series so I might use my own Carlson Symmetric Form for that instead, or at least compare how much slower that is compared to this.
<mofh>
Like this seems mostly the integrals of the second and third kind whereas I only need the first kind, heh.
<egg>
mofh: carlson is not necessarily best
<egg>
that's kinda the whole point of fukushima's work aiui
<mofh>
egg: I mean I don't think it's *less* accurate than just a bunch of Taylor polynomials.
<mofh>
Like the code does *interesting* things for E. But I don't need E.
<egg>
yeah but is it more accurate
<mofh>
Hrm, fair point. I guess I *do* need to test that, so having the Fukushima/phl code is handy for that.
<egg>
and is it as fast
<egg>
(in particular our code starts estrining at high degrees)
<egg>
(not sure whether fukushima does that)
<egg>
mofh: poke me if you need help building principia
<egg>
this does have dependencies on the wider numerics libs, for the estrin evaluator
<egg>
mofh: also i'm not completely sure it does what you think it does
<egg>
mofh: EllipticF is not the same function as FukushimaEllipticF
<egg>
EllipticF is calculated from EllipticEFΠ
<mofh>
no, I mean EllipticK
<mofh>
since at present I only care about the complete one
<egg>
ah
<egg>
boring
<mofh>
(I will need K incomplete in the future and *that's* when things get interesting)
<mofh>
I mean I agree, like I could just do the incredibly lazy thing of using AM-GM to calculate it. But that has weird accuracy-per-iteration properties and is a bit slow.
<egg>
mofh: that's F, right?
<egg>
not K
<mofh>
yes
<egg>
FEΠ but KEΠ, because sane notation
<egg>
mofh: anyway, those are taylor series going to degree 19 and Estrin-evaluated
<egg>
this does seem nice for a modern machine
<mofh>
And yes you want to use Payne-Hanek for Reduce for large multiples of π
<egg>
mofh: if you run it on a Cray it's a different question
<mofh>
I mean I'm doing this on a modern machine so those are likely very fast, point.
<egg>
mofh: yeah you do, or really you want your angles measured in turns
<mofh>
(Which reminds me, I need to fix a bug in my double-precision Payne-Hanek, test the hell out of it and put it up on github)
<mofh>
(I have a single-precision one which works correctly for every single single precision float, b/c that's a thing you can test. Not so much for doubles).
<egg>
also aaaaaaaaaaaaaaa how do i into decisions aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaa
<egg>
a
<egg>
a?
<egg>
a.
<mofh>
At this point I legit am wondering if flipping a coin may be of any help.
<egg>
;choose should i flip a coin
<kmath>
egg: There wasn't any delimiters in your message. Split your choices by , or |
<egg>
;8ball should i flip a coin
<kmath>
egg: You may rely on it
<mofh>
(I still lean "do apply for phd", for what it's worth. But not strongly).