HNHacker News
TopNewBestAskShowJobs

astrophysician

521 karma · joined May 3, 2019

astrophysicianhn@gmail.com
submissionscomments
astrophysician··on Infected people without symptoms might be driving the spread of coronavirus
Yea I agree that it’s still useful like you say as just a way to screen for “higher risk” individuals and have them isolate, but still food for thought.
astrophysician··on Infected people without symptoms might be driving the spread of coronavirus
Just FYI, if you don’t have symptoms or direct contact with someone with COVID (basically if you don’t have a high probability of having COVID before you have the test) a positive test still probably means you don’t have COVID:

Sample numbers:

* Population of US: 3e8

* People that have COVID right now in the US: 1e5 (30x the confirmed cases right now)

* Disease rate: 3e-4

* FALSE POSITIVE RATE for COVID test: 1% (likely higher)

* TRUE POSITIVE RATE for COVID: 100% (definitely lower)

If you take the test, don’t have symptoms or direct contact, and then the test comes back positive for the disease, then under these pretty reasonable assumptions, the likelihood you have the disease is

P(actually have COVID| positive test) = 3e-4 /(1e-2(1-3e-4) +3e-4) = 2.9%

astrophysician··on Potential false-positive rate among the 'asymptomatic infected individuals'
Right, so it's a variable, but if we can independently nail down a false negative/positive rate, then we could try to infer the true rate of infection, but problem is (assuming US numbers):

* assume false positive rate of 10%, assume true positive rate of 100% (it's not but let's be generous)

* maybe 1/10 of the population (30 million) gets tested

* lets say 100,000 people have COVID right now (50x the official number)

number of positive tests = 30 million * 0.1 + 100k * 1 = 3,100,000

fraction of postive tests that actually have the disease = 1 / 31 = 3.2%

problem is FPR (1) depends on the population tested (i.e. p(covid | positive test) != p(covid|positive test, some symptoms) != ...), (2) we need very accurate measurements of FPR because:

lets say we constrain FPR to 10% +/- 1% ==> 10% uncertainty in FPR -- that means our inference of the number of infected people is:

n_infected = (n_positive_tests - FPR * n_tested)

which is: -200,000 to 400,000

so...not very useful.

astrophysician··on NCAA cancels March Madness and other championships
I mean, I don't know the OP's intention here but I think they're probably right, all of these actions will have an economic impact. It's not a statement about whether or not we should or shouldn't do something, I think we need to take lots of big proactive measures, but big proactive measures have lots of collateral damage that isn't negligible. Economic impact will be felt by everyone, from the top to the bottom of the income ladder, so that's also a legitimate pain point...

> suspending evictions, delaying mortgage payments, nationalizing healthcare

I mean sure, I don't think I would disagree with any of that at all (good luck with the last part though, we've been trying that for 40 years), but to say that this wouldn't have a detrimental impact on the economy is a bit confusing to me...it will and that will hurt, but the idea is that it will hurt less than something like COVID running rampant or without measures to protect workers who will be hurt most by the economic fallout.

astrophysician··on US admits 'failing' on testing, says Fauci
What is the true/false negative/positive rate for these tests? I can't find that online and I have tried.

If FPR is 10% and we test 100,000,000 people in the US but only 100,000 have the disease right now, then assuming a TPR of 100% (and a FNR of 0%) we will have 10,100,000 positive results = 101 x the number of people with COVID? How is that useful?

Disclaimer: I am maybe an idiot.

astrophysician··on Freeman Dyson Has Died
NDT came to Thunch several times, but unfortunately not while I was there -- we exchanged e-mails (I was responsible for booking the Thunch guests) and he was excited to come, made tentative plans, but it ended up not working out. The guy is pretty busy these days. He was a post-doc at Princeton back in his research days, so he still keeps in touch with the department every now and then.
astrophysician··on Freeman Dyson Has Died
Thanks for sharing — yea completely agree, it’s always so much fun to be around people like Dyson. There’s more like him at Princeton (and everywhere else, I was at university if Illinois for undergrad and the story is the same there), they just get a kick out of being curious, and always are genuinely interested to learn and hear new things, no matter how old they get. Such a pleasure to be around — they aren’t bitter or insecure, they don’t have anything to prove, they just think that some things are just so gosh darn cool, and that fun energy is contagious.
astrophysician··on Freeman Dyson Has Died
About 4 years ago, I was a grad student in astrophysics at Princeton. Dyson worked at the Institute for Advanced Study which, while not officially part of the university was essentially an extension of the various departments (shared colloquiums, any researcher there was a valid thesis adviser, etc.).

The grad students have a tradition (or used to) called “Thunch” (Thursday Lunch) where they invite someone to come have lunch and hang out every week. It was by far one of the highlights of my life — maybe 10 people in the room max, eating pizza, shooting shit with some of the most impressive and accomplished people you could hope to meet — Nobel prize laureates, Neil deGrasse Tyson, famous authors, whoever would respond to our invitation (people tended to say yes). Dyson was one of them. At the time he was already like 92 years old, dressed in a suit, occasionally using a handkerchief, and this guy had just published a result in Nature. The guy was SHARP at 92. Nicest, most interesting person I have ever met. Knew Einstein from his early days at IAS. Worked on an early version of the space program that looked into launching people into space with large spring loaded contraptions that were to be propelled with nuclear weapons. Those were the days I guess...

So sorry to see he’s gone, the guy was a living legend.

astrophysician··on Holes in Bayesian Statistics
Replying since it's been too long to edit:

Now that I've taken some time to actually READ the article referenced in this post, I have a couple more things to say; basically: I agree with virtually everything that they're saying in the paper (except that I still don't understand why Bayesian stats failing in QM says absolutely anything about the validity of Bayesian stats), and I think I set up my own straw man here a bit. I thought this was another "Bayesian stats is bad because priors" post, which in a way it SORT of is, but their arguments are a bit more on the side of "you should take 'Bayesian' analyses with a large grain of salt sometimes because...well...priors" <- and THAT I completely agree with.

The only beef I have with these types of articles are when they seem to imply that because priors sometimes have an outsized effect on the results, we should not use Bayesian statistics/analyses. But really, there is no other option. If you have a statistical question, you're stuck with having to answer for your priors. That being said, what IS good advice is to not simply believe every 'Bayesian' conclusion for exactly that reason -- that result is only as good as the (1) prior, (2) model assumptions, and (3) the data. (2) and (3) are easy to fight about, but (1) can be hard to evaluate, and easily can lead all of us astray even when the prior sounds reasonable (flat priors).

astrophysician··on Holes in Bayesian Statistics
That’s awesome, and maybe I misunderstand his argument (very possible, in fact my main reason for posting this comment). Am I not understanding the arguments being made here? I have not been following these developments regarding using Bayesian stats in QM, it’s just that if Bayesian stats doesn’t help us understand QM it doesn’t mean there’s a “hole” because Bayesian stats makes no claim whatsoever that it has any applicability to QM. Am I just not understanding?
astrophysician··on Holes in Bayesian Statistics
> Bayesian statistics doesn’t give a satisfying answer/method for choosing priors

I absolutely agree with you. Priors are the (very difficult) job of the practitioner.

> choosing priors is a part of statistics

Ah so I think I agree, maybe would nitpick and say priors are part of statistics and CHOOSING priors is part of modeling maybe? Not totally comfortable with that characterization either it’s just that “priors” and “choosing priors” are two different things.

> therefore Bayesian statistics has a “hole” in it.

Completely disagree; this is my main understanding of the argument against Bayesian statistics, I just don’t understand how this follows from the last statement. Priors are hard. But they are explicit! And a necessary ingredient of the posterior. It’s not like we can say “oh priors are hard, so let’s not do Bayesian stats but instead let’s do X” what is X?

astrophysician··on Holes in Bayesian Statistics
Every time I see an article like this, I eagerly read it hoping to find a cogent and coherent criticism of Bayesian stats, and it always ends up being a straw man or a very fair critique of somebody not correctly applying or interpreting an application of Bayes’ theorem.

Bayesian stats is nothing more than a rigorous way to transform beliefs + data into a posterior. Yes, flat priors are not always the best choice. That’s not a criticism of Bayesian stats. It’s a statement about how actually formulating a prior is often times the hardest part of a problem. Is Bayesian stats useful for describing or understanding QM? Idk, again, not really it’s job...

Use Bayesian stats, not with an air of suspicion, but a respect for the fact that it will give you the results implied by your data and prior, under your model assumptions. Nothing more, nothing less. By the way, what is the alternative if you find yourself in a situation where your result depends strongly on your prior and you aren’t really sure how to choose your prior? Wave your hands and find an ad hoc frequentist approach? How about just admitting to yourself that your data isn’t enough to make up for the fact that you can’t really quantify your true prior belief?

If you read this and disagree, I sincerely implore you to comment — I just don’t understand the “debate” aspect of Bayesian stats. Some people misunderstand what Bayesian stats is sometimes (very understandable) but I have yet to see a legitimate philosophical or mathematical critique of the Bayesian approach that really made any sense to me. I would like to know if I am wrong though...

astrophysician··on Why scientists need to be better at data visualization
Thank you!!! Yes! VisIt is what I was trying to remember. I don’t know VisIt, only that it exists and numerical people (at least in Astro) use it.
astrophysician··on Why scientists need to be better at data visualization
https://yt-project.org/ -- yt for python was used by a few of my old colleagues for visualizing astrophysical simulations. There is yet another program that I've since forgotten but will try to come back here and comment if I remember.

If you can generate a single static figure, you can manually generate animations by generating each frame separately using your favorite plotting software. You can then stitch them together using many different tools (I use matplotlib) to generate an animation (video, gif, etc.).

What language do you like to use? What's your background -- are you a Matlab person?

astrophysician··on Show HN: Making pro weather data free for Hurricane Dorian
Not a bad idea but only if the tracks are independent. Was reading something saying that these tracks aren’t different random realizations of the same simulation (in which case what you’re saying is a good way to estimate the posterior of the hurricane track — it’s just Monte Carlo) but that these tracks come from many different types of simulations, some with higher quality than others and they’re likely all correlated in some weird way. If it’s the latter case there has to be some care and thought put into combining the predictions into a single estimate of the posterior.
astrophysician··on I've reproduced 130 research papers about “predicting the stock market”
> Do you expect a trading strategy that harms a market to be reliably profitable without risking fines, banning, or imprisonment?

Non-trader here, what do you mean by a particular trading strategy "harm[ing] a market," how would you actually measure harm to a market by a given trading strategy, and would whatever metric you decide on for that not just be a dressed-up subjective argument?

astrophysician··on Milky Way’s Black Hole Just Flared, Growing 75 Times as Bright for a Few Hours
Ah great blog theme! I’ll definitely peruse...I just defended last November (variable stars) and now I miss it. Hope life outside academia is going well for you. Best of luck!
astrophysician··on Milky Way’s Black Hole Just Flared, Growing 75 Times as Bright for a Few Hours
YES! Really great explanation! Have you considered writing for astrobites or any Astro blog for that matter?
astrophysician··on Are We Living in a Computer Simulation? Let’s Not Find Out
Monte Carlo can be used to simulate any stochastic part of a process, not just inputs.
astrophysician··on Adobe shambles – Why subscription software should be illegal
Is there a piece of software that doesn’t require even occasional bug fixes or updates in order for it to run reliably on even a single operating system/browser?
astrophysician··on Ask HN:Skills/Tools to Stand Out as a Data Scientist
Cool! Kudos! Also don’t feel too discouraged if it takes you awhile, I’ve been trying to learn Bayesian related things for years now, and there are many aspects with significant learning curves (speaking for myself, maybe you won’t have so much trouble!). It’s not trivial stuff. But it’s helpful so every bit you can understand is a good investment I think.
astrophysician··on Ask HN:Skills/Tools to Stand Out as a Data Scientist
Some others have commented that this isn’t “required” but I can tell you that solid understanding of Bayesian statistics makes everything make way more sense and gives you a great basis for building more understanding of any ML algorithm. I don’t mean just memorizing Bayes’ theorem and answering the canned “99% accurate test is positive for a disease that affects 0.01% of the population” interview questions, but understanding relationships between methods (GP’s, SVM, neural nets, tree based things) and how they relate back to Bayesian statistics either directly or by approximation is key to having true visibility into what you’re doing. Sure you know about loss functions and regularization, but if you can’t look at a loss function and translate it into a precise statistical statement in terms of priors and likelihoods, you don’t truly know what’s going on. You also will have a hard time extending other loss functions/regularization schemes in a way that isn’t problematic, and explaining/justifying the choices you make on that end. Thinking in terms of Bayesian statistics (Bayesian stats IS stats, it’s mathematically equivalent to frequentist statistics but has a different philosophical interpretation) saves you from many many rabbit holes that in hindsight are obviously Bad Ideas when seen from a more Bayesian point of view. Not to undermine people without a strong statistical background, you can still do good work, but your life becomes easier and less ad hoc (in my experience) once you understand things statistically. You should ideally understand the assumptions behind your models, and what the implications of those assumptions are, and it’s hard to do that consistently and rigorously without Bayesian statistics.

That being said, Bayesian statistics is not SUFFICIENT for doing data science well. You need to know pragmatic solutions to intractable problems, what heuristics might be ok to try, how to EXPLAIN things to people, how to code well enough, etc. But boy does knowing Bayesian statistics help.

← PreviousPage 4 of 4