lr1970

42 ms·

481.
by lr1970 5y ago
To clarify, abbreviation ZFC stays for Zermelo–Fraenkel set theory with the Axiom of Choice included [0]. [0] https://en.wikipedia.org/wiki/Zermelo–Fraenkel_set_theory
482.
by lr1970 5y ago
Whatever they do TIOBE Index methodology is flawed. MATLAB at #13 ahead of Ruby #16 and Swift #17 makes no sense. MATLAB is a niche language popular mostly among Electrical and Mechanical Engineers while Ruby and Swift are general purpose a
483.
by lr1970 5y ago
> It's just about what happens when those concerns conflict with other concerns, such as maximizing user engagement. They care about their own kids more than they care about abstract faceless millions of other kids on FB. Moral bank
484.
by lr1970 5y ago
> It's interesting that BO tries to tar the author by claiming she was terminated for violations of federal export control regulations - but she was an employee communications manager. How does such a role even intersect with that?
485.
by lr1970 5y ago
> Textadept[1] is also quite nice, and configured (but not written) in lua. It seems to work fine in MacOS, but I haven't used it extensively. [1] https://orbitalquark.github.io/textadept/ Straight from your li
486.
by lr1970 5y ago
> Att was destroyed as soon as they sold to cingular. The story of AT&T and Cingular is much weirder, see [0]. Once there was AT&T Wireless that became Cingular (with the help of SBC). Then SBC bought AT&T (landline) then the
487.
by lr1970 5y ago
Before you put it sleep activate Airplane mode. Hopefully it will disable wake ups.
488.
by lr1970 5y ago
> How is the "modular port" concept any different than a universal port with dongles ... At least the "modular port" adapters are not dangling from the side of a laptop as the dongles do. Dongles totally ruin the esth
489.
by lr1970 5y ago
> Most data is not 300 years old or in the distance future, in fact ranges 1970+-292 years are very common. In what domains? Astronomy, geology, history call for larger time range. Laser and High Energy physics need femtosecond rather th
490.
by lr1970 5y ago
The problem is not with the Wes' original decision but with the fact that it was never revisited even when pandas took off at much larger scope. Should had been fixed before 1.0 release.
491.
by lr1970 5y ago
An illustrative example of how reasonable short-term and narrow-scope considerations can be really bad in long-term and/or at a larger scope.
492.
by lr1970 5y ago
Early on, pandas made some unfortunate design decisions that are still biting hard. For example, the choice of datetime (pandas.Timestamp) represented by a 64-bit int with a fixed nanosecond resolution. This choice gives dynamic range of +-
493.
by lr1970 5y ago
Just to clarify, scikit-learn 1.0 has not been released yet. The latest tag in the github repo is 1.0.rc2 https://github.com/scikit-learn/scikit-learn/releases/tag/1....
494.
by lr1970 5y ago
A recent (9 days ago) related Hacker News thread about ASML and its $150 million machine https://news.ycombinator.com/item?id=28266708
495.
by lr1970 5y ago
> 400 applications?! I've heard about people on cscareerquestions sending hundreds of CVs, but this is insane! Yep, looks like "resume spam". No wonder recruiters do not read submitted resumes anymore. Kind of DoS attack :
496.
by lr1970 5y ago
> Problem is you’re fighting path loss which is exponential with distance, whereas the power of your emitter is scaling linearly. No, EM radiation power decays quadratic with the distance, not exponentially.
497.
by lr1970 5y ago
>> with lots of money simply going to waste > What does this look like? Is there money going into the system that’s not going to anyone / any company at all? It means that money efficiency and productivity are low. A surgeon i
498.
by lr1970 5y ago
I am a bit biased here, for I am married to an American Physician. She works longer hours than I do. Her work is much more stressful (literally, life and death decisions almost every day). She can be sued for malpractice while I can only be
499.
by lr1970 5y ago
In the USA a title company that validates the ownership provides a title insurance for exactly this case. If the true owner of the house is determined (typically in court) not to be the one the title company claims, then the title insurance
500.
by lr1970 5y ago
All luggage is supposed to undergo a security check in an X-ray machine. This is true for private jets as well. I am curious how it played out in this particular case. Did it go through an X-ray machine or not? If yes, how did they mistook
501.
by lr1970 5y ago
> Bonus fits into the category of "supplemental wages" which Google suggests has a special 22% tax rate. However that rate also apparently applies to RSUs most of the time. This 22% tax rate is the default rate applied by your
502.
by lr1970 5y ago
> If only we had thermocouples with solar panel like efficiency... Thermocuples are still heat engines converting thermal energy (heat) into zero-entropy energy (electricity in this case). Second Law of Thermodynamics limits efficiency
503.
by lr1970 5y ago
> Conda has been a lifesaver for me in the past, but it got so slow in ~2019 This is why mamba [0] was created. It is a C++ reimplementation of conda for much better performance. mamba is a drop-in replacement of conda and can operate on
504.
by lr1970 5y ago
IMHO, the main issues of USSR economy in addition to central planning were total lack of economic and social incentives for productive work. Career growth and personal income were almost independent of how a person performed in their daily
505.
by lr1970 5y ago
When people talk about climate change they mostly focus on rising mean temperature. They fail to realize that variance of the temperature distribution is increasing as well making extreme weather events (in both directions) more likely. Hen
506.
by lr1970 5y ago
> Can you elaborate a bit? Prophet can use MCMC sampling and includes uncertainty in its forecasts. Prophet is a GAM (Generalised Additive Model). It decomposes time series in additive components: trend, seasonality, holidays and noise.
507.
by lr1970 5y ago
> Do you have any favourite libraries for producing such? For modern Deep Learning based probabilistic forecasting you can try DeepAR with parametric likelihood function [0] or Multi-Horizon Quantile RNN (non-parametric) [1]. The impleme
508.
by lr1970 5y ago
Prophet with MCMC can produce probabilistic forecasts. But how to choose MCMC priors and measure accuracy of the probabilistic forecast are open questions.
509.
by lr1970 5y ago
As someone who spent a good part of my professional career in forecasting and time-series analysis I would like to point out that "point-forecasts" are mostly useless in many import practical applications such as FinTech, e-commer
510.
by lr1970 5y ago
> Even with a cursory understanding of history, the extreme right wing regimes are considered fascist, aka extremely top down. IMHO the political spectrum has a circular topology where extreme left and extreme right merge into one point.

More ›