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
* Raptop
sees a job posting, and wonders if I'll be qualified after graduating
<Raptop>
A "the two genders" joke, but for whitequark's cats
<Raptop>
Okay, CARMENES uses probably >1 order of magnitude more compute than we do for their scheduler. I'm… not entirely convinced that they get better results o_O
<egg>
:D
<Raptop>
Also, lolcadence. We're trapped on the surface of a spinning rock next to a gravitationally confined fusion reactor. You're going to have problems even at an ideal site.
<Raptop>
If I mysteriously lose internet access at midnight, blame them.
<Raptop>
Okay, their voice menu works.
<Raptop>
Also, navigating robotic menus is deep in the uncanny valley these days
<mofh>
Raptop: I find that taking a deep breath and then loudly swearing at the voice menu semi-reliably instantly takes you to a human (it varies based on company, IME).
<SnoopJeDi>
Raptop, I missed Mike Brown's talk unfortunately! But my family went and I was able to introduce them to him afterwards and he even remembered chatting with me the last time he came :)
<SnoopJeDi>
he was very receptive to my family cornering him and peppering him with questions for a bit
<kmath>
<✔planet4589> Since I seem to have coaxed some CNES fans out of the woodwork, here is a 1966 primary source document I hadn't see… https://t.co/2VU8clwC6I
<SnoopJeDi>
happily, it was largely the talk I saw anyway
<SnoopJeDi>
modulo "we're done with our Subaru telescope run, but analysis takes a while so..."
ferram4_ has joined #kspacademia
ferram4 has quit [Ping timeout: 183 seconds]
<egg>
mofh: nice
<egg>
mofh: meow
<egg>
mofh: underlying one-step method
Raptop has quit [Remote host closed the connection]
<_whitenotifier-3d18>
[Principia] eggrobin opened pull request #2126: Refer manœuvre times to a time base, not to the current time - https://git.io/fjtE0
<galois>
[WIKIPEDIA] Christoffel symbols#Transformation law under change of variable | "In mathematics and physics, the Christoffel symbols are an array of numbers describing a metric connection. The metric connection is a specialization of the affine connection to surfaces or other manifolds endowed with a metric, allowing distances to be measured on that surface. In differential geometry..."
<kmath>
<eggleroy> 🌑 Fano, the new release of Principia, is out; asynchronous computation of predictions, making long predictions usab… https://t.co/kahED67uNf
<egg|laptop|egg>
(the new moon was friday)
egg|laptop|egg has quit [Ping timeout: 189 seconds]
<UmbralRaptop>
egg: It's probably best to be a bit farther from the blackhole when compiling a release
<egg|cell|egg>
It was released on time
<egg|cell|egg>
Just tweeted late
<UmbralRaptop>
birb undergoing high gravitational redshift
egg|laptop|egg has quit [Ping timeout: 190 seconds]
egg|laptop|egg has joined #kspacademia
<_whitenotifier-3d18>
[Principia] pleroy labeled pull request #2126: Refer manœuvre times to a time base, not to the current time - https://git.io/fjtE0
<_whitenotifier-3d18>
[Principia] eggrobin synchronize pull request #2126: Refer manœuvre times to a time base, not to the current time - https://git.io/fjtE0
<kmath>
<cdmblogs> This ultrasonic radio station broadcasts from inside an actual rat burrow and - the results will surprise you!… https://t.co/lxaa96rJVh
<mlbaker>
no, cos(tA)_{ij} is what you get by taking the power series for the cosine and replacing x^r with (A^r)_{ij}
<mlbaker>
er
<mlbaker>
t^r (A^r)_{ij}, my bad
<SnoopJeDi>
I'm trying to understand what cos(tA) means since cos() is typically defined for scalars?
<mlbaker>
you take the matrix and you slam it into the power series
<SnoopJeDi>
and "apply it to all the entries in this matrix" is the only sensible thing I can think of?
<mlbaker>
right, so it doesn't quite work like that because (A^k)_{ij} =/= (A_{ij})^k
<mlbaker>
except for like, diagonal A
<SnoopJeDi>
mmhmm
<SnoopJeDi>
OH
<SnoopJeDi>
I see what you're saying
<SnoopJeDi>
so cos(tA) is itself a map, got it
<UmbralRaptop>
aaaaaαααα
<SnoopJeDi>
(I feel like this is a Very Obvious Linear Algebra™ thing that I may have known once and forgotten?)
<mlbaker>
btw, it turns out that expression is [e^{-2it}(1+2 e^{3it})]/3
<mlbaker>
which is... much easier to compute...
<mlbaker>
oh sorry, i mean it's the real part of that thing
<SnoopJeDi>
OH YES OKAY
<SnoopJeDi>
I've definitely seen e^A before and naturally e^iA under that understanding lets you understand cos/sin
<SnoopJeDi>
(at least algebraically)
<mlbaker>
yeah
<mlbaker>
I need to figure out a systematic way to rewrite the former type of thing to the latter type of thing whenever possible
<mlbaker>
so that I can actually evaluate this shit >_>
<mlbaker>
I guess I could just approximate the graph's eigenvalues and then analyze the exponential by diagonalizing
<SnoopJeDi>
oo, what sort of graph are you working on?
<mlbaker>
cycle graphs of arbitrary size
<SnoopJeDi>
also, Waterloo? (just curious since we haven't interacted much)
<mlbaker>
alma mater
<SnoopJeDi>
ah. you and mofh know each other irl maybe?
<mlbaker>
yep
<SnoopJeDi>
neato :)
<SnoopJeDi>
(lucky you!)
<SnoopJeDi>
mlbaker, I was doing a tiny bit of cycle explanation to someone who wasn't quite getting it in a Python channel last week
<mlbaker>
cycle graphs?
<SnoopJeDi>
they were puzzling over the itertools documentation for permutations(), which is a very ugly direct translation of the relevant CPython, which requires some Deep Evil™.
<SnoopJeDi>
yea, if you're computing n-choose-k with the algorithm used in CPython, you're exploring nested cycles
<mlbaker>
hmm
<mlbaker>
isnt n choose k easy to compute by just like...
<SnoopJeDi>
so once you've chosen (k-1) things, you have (n-k) choices remaining
<SnoopJeDi>
and it does this by storing indices in a list and running through the cycle
<SnoopJeDi>
(shuffling indices)
<SnoopJeDi>
mlbaker, to be clear I mean the permutations themselves not the number of them
<SnoopJeDi>
i.e. [(a, b, c), (a, c, b), (b, a, c), (b, c, a), (c, a, b), (c, b, a)]
<mlbaker>
hahahahaha
<mlbaker>
this is the general expression i derived for the transition probabilities of a quantum walk on a cycle
<mlbaker>
after deriving it i then realized it's the most useless expression ever
<SnoopJeDi>
and CPython (and the godawful "roughly equivalent" Python code in the docs) does this by an algorithm that walks cycles in the same way an odometer does, basically
<mofh>
yeah I'm legitimately at a loss for when it would be useful
<SnoopJeDi>
(it gives permutations in lexicographical order)
<mofh>
like Bessel series *can* be useful but what the hell is even the *purpose* of this one?
<mlbaker>
all i did was basically combinatorially reason out the number of length r walks between any two vxs
<mlbaker>
and then yah just plug those counts into the exponential series
<SnoopJeDi>
I am getting nothing done bc I'm listening to it
<mlbaker>
and some nice lil 0F1(...)s pop out
<SnoopJeDi>
mlbaker, what's the elevator-talk for the transition probabilities you're exploring?
<mlbaker>
quantum walks on cycle graphs
<mofh>
"nice lil 0F1(...)s"
<mofh>
that's kinda an oxymoron imho
<mofh>
:p
<mlbaker>
XD
<SnoopJeDi>
I got that part, but why? Looking at some interaction to such-and-such order?
<SnoopJeDi>
(this is just curiosity, I'm trying to get in the habit of practicing scicomm skills)
<mlbaker>
trying to study whether there are any times t where all transition probabilities are 1/N
<mlbaker>
ie 'uniform mixing' occurs
<SnoopJeDi>
ah okay, I've definitely heard that phrase before
<_whitenotifier-3d18>
[Principia] pleroy closed pull request #2126: Refer manœuvre times to a time base, not to the current time - https://git.io/fjtE0
<_whitenotifier-3d18>
[Principia] pleroy pushed 3 commits to master [+0/-0/±6] https://git.io/fjt2q
<_whitenotifier-3d18>
[Principia] eggrobin 436e130 - refer manœuvre times to a time base, not to the current time
<_whitenotifier-3d18>
[Principia] pleroy 44d75ed - Merge pull request #2126 from eggrobin/time-base Refer manœuvre times to a time base, not to the current time
<kmath>
<enf> If anyone was curious what a (1950s) programming language looks like when you start from French instead of English,… https://t.co/hHSTwnskGp
<mofh>
egg|laptop|egg: meow, gotcha, meow, sec.
<egg|laptop|egg>
ALGOL!
<egg|laptop|egg>
whitequark: iirc there was also ru ALGOL?
<_whitenotifier-3d18>
[Principia] pleroy synchronize pull request #2127: Add support for editing burn parameters using a text field - https://git.io/fjtw7
<galois>
[WIKIPEDIA] List of BeiDou satellites | "This is a list of past and present satellites of the BeiDou/Compass navigation satellite system. As of November 2018, 33 satellites are operational: 6 in geostationary orbits (GEO), 6 in 55-degree inclined geosynchronous orbits (IGSO) and 21 in Medium Earth orbits (MEO). Furthermore, 6 satellites (3..."
<mofh>
Wait, they have BeiDou in GEO? That seems weird for a GNSS system
<egg|laptop|egg>
mofh: early satellites, as well as inclined GSO