Parent must have mistaken GMP for OMP, and wanted to shoot a quickie — no having to read the article! some research should be done— is it because HNers browse HN just before bed?
To be honest i made the same mistake at glance… fortunately i am not mentally exhausted…
Given the author mentions multiple cores being available, I'd guess you could use any method, including MPI, to distribute the computation. But whether you used 1 core or 10k cores, it would be nice to have a 20x speedup on each core via this arithmetic/fixed size optimization. Since that's the focus of the article, communication technologies feel pretty unrelated.
The amount of effort and extra hardware required to maintain this reproducibility across different hardware (cpu generations, GPU hardware/drivers, etc) feels insane to me. For example, people seem extremely interested in performance, but insist on using the much slower modes in MKL to maintain the bottom few bits of their golden data.
Do you find this genuinely important? If not, have you had any success encouraging e.g. validating against golden data in a non-bitwise comparison?
I personally don't find it important. And I have seen that the demand of having numbers match exactly comes from a misunderstanding of how floating point calculations work.
Sometimes this is understood by the banks and the solution is to implement a tolerance in the comparisons. In other cases other solutions may have to be used, such as forcing certain reports to be computed on certain software/hardware.
That's not what I'm talking about. Floating point computations happen when doing risk calculations, for example computing risk numbers in a Black-Sholes calculation.
A difference in the 15'th decimal has no impact on the risk number.
If I recall, the zstd binary has a mode that is basically "adapt the compression ratio so that compression isn't the bottleneck", but I didn't have a chance to try it. Have you used that?
tldr: zstd seems to be somewhat more flexible, generally provides slightly better performance across almost all metrics, and decompresses much more quickly than Brotli.
I understand cause of death statistics can be calculated, but do you have an article on what happened with a decapitation being counted as a covid death?
There's at least one widely known case specifically a motorcycle accident listed there, which due to being widely known was corrected.
This comes down to how you fill in unknowns in your model of the world, many seem to be just trusting authoritative reports. I would like to trust official reporting, but direct knowledge via peers involved in various industries has proven to me or at least created a confident opinion that trusting things at face value is absurd and a near guarantee of inaccurate conclusions. Its amazing how many seem to experience Gell-Mann amnesia in regards to this sort of thing.
Since there's only one public report of this which easily is found on google (which is what I linked), I'd assume such extreme (decapitation) false labeling of cause of death is not so common. However cases of elderly who had multiple conditions being reported as covid deaths appears likely as a fairly massive distortion of the reported death count. The motorcycle case just serves to show how extreme that can get, you've got to ask just how a report like that even occurred, and also note that it seems to only have been corrected after public scrutiny. (I can't be sure it was corrected due to public scrutiny, but I'm leaning towards that since explanation of how it was corrected is dismissive and vague, no clear suggestion that it would have been corrected if not for publicity of it, seems an easy extra few words to add if it were the truth)
I can understand if this way if thinking isn't yours and if accepting official numbers is preferred, but what I can't understand is the absolute certainty that any other perspective is wrong.
Public health is complex, and it's a bit tough for us to "agree" that it's "okay" to have different, looser perspectives on how to manage a pandemic. If 50% of the population decides that COVID-19 is fake and that nobody is actually affected at all, then we would have no power as a society to do anything about pandemics.
Agreed that it is complex and I understand it is a bit tough, which is part of why I think the authoritative perspective should not disappear or be entirely disrespected, but would best be toned down. I would trust the CDC or WHO far more if they demonstrated themselves to be organizations I could trust. To me, that starts with being honest instead of carefully chosen lies expected to be "most effective" at controlling people to do the right thing. I have faith in the long term that this will be learned by these organizations and they will get better, but for now I don't see it as rational to accept at face value anything coming out of them that has any wiggle room of interpretation.
The rest of the population that believes things you believe are crazy, are not crazy, but just working with different information. There's cases of some beliefs which pretty much exclusively crazy people believe (mass gangstalking for example) but that's pretty easy to identify and has very direct correlation with mental health conditions. The way to fix this is to treat people as the rational agents they are and allow them to make their own choices. For now it seems myself and others who question authoritative sources on what reality is, have to almost default to "what the authority says is probably lies" because that seems to be true.
My understanding is that the Texas state requirements used to be borderline insane, but that they have improved over time. Skimming http://ritter.tea.state.tx.us/rules/tac/chapter113/index.htm... , I actually found much less objectionable than I expected to.
There's no way to get the raw data from this site, is there? I have long been interested in doing some analysis of how syllabi have changed over time, but that doesn't seem possible with this interface.
I don't think this article provides strong evidence that they are well calibrated on the presidential election specifically (sample size N=3), or that they are correctly accounting for rare black swan events, but it does seem to imply that the criticisms about "538 claims victory no matter what because they always have non-zero probabilities" are oversimplified.
2016 wasn't a black swan event. It was a polling error, which do happen, if rarely. It was not unforseeable, and 538 included a probability of that happening which is why they gave Trump higher chances than most others did.
And 538 does do backtesting on elections back to 1972. That's not particularly trustworthy since it invites over-fitting, but internally they do have a little bit more than N=3 to work from.
(I have fairly minor quibbles with some of Nate's modeling ideas, but I broadly mean to be defending him).
I don't mean to imply 2016 was a black swan event- I agree that ~30% was probably as accurate a take as could be achieved (most evidence that seems reasonable to use indicated a lead for Clinton, but that it wouldn't be that surprising for that lead to be overcome). I just mean that the model assumes a fairly normal election environment, without like a huge attack on Election day or something on election day.
The N=3 comment was meant specifically for evaluating their calibration, not the data they use for their model.
