makeset

Code Complete by Steve McConnell https://www.amazon.com/dp/0735619670

What? All Abrahamic religious texts are notoriously ambiguous and self-contradictory. History of religion is the study of sectarian conflict.

Front-running is an illegal practice that has absolutely nothing to do with order size.

https://en.wikipedia.org/wiki/Front_running

I can think of few things money can buy that have impacted my quality of life as much as where I live, in both location and living space.

Not verbal, but I think of strict adherence to coding and commenting standards in this spirit. A single character out of place can distract the reader from a fatal bug. Also, acknowledging one's fallibility and handling cases that should never happen:

  else throw("BUG: ...")

RightZoom (free): http://www.blazingtools.com/right_zoom_mac.html

Though I also tend to have ShiftIt or SizeUp installed and always use the keyboard shortcut instead.

It's immature to rail on R for things like how its syntax doesn't look like that other language you already know, but yes, it's extremely quirky. And its unchecked flexibility, while great for experts providing DSLs and such, enables non-programmers to produce garbage implementations. The vast majority of R's myriad community packages (R's biggest strength!) seem to be written by a grad student who read half a tutorial.

Also remember that heart attacks in women can exhibit markedly different symptoms, e.g. without any chest pain or arm numbness, often misjudged until too late.

https://www.heart.org/HEARTORG/Conditions/HeartAttack/Warnin...

> Statements like these suggest that another AI Winter is coming (sooner than 3 years I guess)

I wouldn't bet against it. It's very risky to extrapolate progress in AI.

A DARPA project named CHI [1] succeeded in automatically producing programs from a high-level knowledge description language. And when they fed it its own description, CHI rewrote itself, with a better implementation. Do you fear for your programming job yet? Well, that was 36 years ago.

[1] www.dtic.mil/get-tr-doc/pdf?AD=ADA105661

If you rolled your own naive numerical approximation of L1 regularization, you might not have gotten exact zeros. If you use e.g. LARS or cyclical coordinate descent for the L1-regularized parameter cohort, as suited to the problem, you will get exact zeros, as prescribed by the mathematics of L1.

Yes, it does. You might have been thrown off by the fact that L1 regularization is not L0 regularization, i.e. it doesn't explicitly limit the number of nonzero coefficients. Still, the linearity of L1 constraint boundaries creates spikes in directions with zero components, thus forcing constrained solutions to occur where many variables are driven to exactly zero. See here:

https://en.wikipedia.org/wiki/Lasso_(statistics)#Geometric_i...

My thoughts exactly. Sounds like he matched up with a pretty great advisor (good person, good manager, helpful, smart), worked on interesting problems, made progress on some, published, and graduated. To boot, when he decided to move on, his physics skills made him highly employable in the financial industry. I'd call that a success story as far as grad school goes.

True, although conventional metrics like Sharpe ratio or ROI are not very meaningful for HFT models, because they can't scale with any additional capital (you can safely assume they are scaled to the max). Their returns are extremely consistent, but also ultimately limited in magnitude. Rather than magical money-making machines who have cracked the "secret code" of financial markets, HFTs are essentially a fixed-cost utility service for reducing market inefficiency through improved price discovery.

Bear uses CloudKit, so your privacy is subject to Apple's private keys instead of Evernote's. Is that a significant difference?

> rather than saving it, they should invest it into their businesses

"Saving it" here is not hiding it under the mattress. Buying stocks is literally investing it, only into other people's businesses. Buying corporate bonds is lending it to other people's businesses so they can grow. And investments into existing established businesses could easily be argued to have at least as large a potential impact as starting out on your own with a minor project that is more likely to fail than not.

How do you play a tabletop RPG by yourself?

Well, 3 years (at best) to find an employer who will sponsor you for an H-1B visa application. And then you have win the once-a-year lottery to get the actual H-1B visa (~1/3 chance this year). And then you hope that your employer will agree to apply for a Green Card on your behalf right away (which they don't have to; they can keep you on the short leash of H-1B for another six years). The processing of the Green Card application can also take a few years. And if you change jobs while your Green Card application is in process, it's voided -- you have to have your new employer re-apply from scratch. And if, at any point during all of these, your current status (OPT or H-1B) expires, you have to pack up and GTFO of the country.

So no, not nearly the same in the US.

R's Scheme-like lexical scoping, lazy evaluation, and reflective abilities make it very extensible if you know what you are doing.

"It's not the years, honey. It's the mileage."

Because liquidity is finite. Profitable investment opportunities don't scale indefinitely. When you discover that an asset is being offered for cheaper than what you think it's worth, pretty soon either you or others buy up all the cheap supply, and the price catches up with your valuation.

You are correct. From the background notes [1] on the band's page:

"It is not all one speed, but each section is at a constant rate, meaning that time does not “ramp” (accelerate or decelerate). We just toggle from one speed to another. When the guitars explode, we are 200x slower than reality (6,000 frames per second), but Tim and Andy’s short bursts of lip sync (Tim twice and Andy once) are only 3x slower than real life (90 frames per second). The watermelons are around 150x, and the spray paint cans are a little over 60x."

[1] http://okgo.net/2016/11/23/background-notes-and-full-credits...

I'll say this: I switched to MacOS after smashing my last Windows (XP) laptop to pieces on the wall in a fit of rage ten years ago. Since then, I'd honestly completely forgotten how a computer could possibly drive a person to physical violence, until I bought a Windows (10) laptop recently just to play some games, and OH MY GOD. It is so, so unbelievably mindbogglingly fucking terrible.

There's an excellent opportunity to apply machine learning here to automatically let you discover what you like. Starting from a set of varietals and/or random blends, it would generate new blends based on your rating feedback, eventually converging towards your preferred tastes. In addition to arriving at your optimal wine(s), a further analysis of the results (plus qualitative labels from experts) could also generate knowledge useful to you elsewhere, like "You like full-bodied left-bank Bordeaux, and also California Merlots. You dislike light-bodied wines in general."

Cutting out all sources of caffeine (I had to; developed an allergy) and taking fast-release melatonin have lately made a tremendous difference for me in falling asleep.

Unfortunately a lot of R's internal standard library code isn't much better, rife with assumptions of minuscule data, gratuitous copying of input data many times over and pointlessly returning them in results, glacial I/O, grossly inefficient implementations even at the C level, etc. Using R in a production environment is a slow journey through reimplementing R itself.

> Status. Regardless of whether it should be or not, working towards and eventually getting a PhD degree is culturally revered and recognized as an impressive achievement. You also get to be a Doctor; that’s awesome.

Ha, sarcasm, very funny. Seriously though, outside of academia, if you're at an actual cocktail party or something with normal people, don't ever say that you have a PhD. It's like admitting that you're into some really weird shit. If anyone else mentions it about you, quickly change the subject to something less cringeworthy, like herpes.

And for the love of God, never ever call yourself a "doctor." WTF is wrong with you? No.

If you're still a graduate student ("working towards"), this is all moot, because they shouldn't have let you into the party.

A friend of mine participated in an ayahuasca ceremony, and he said that it completely changed him, from a person who wanted to try ayahuasca to a person who would never do it again.

As a competitive powerlifter (and certified personal trainer) I should caution that there is an important difference between general strength training using the squat/bench/deadlift and the actual sport of powerlifting. Powerlifting involves not only doing competitions, but also training specifically for them using maximal weights, which considerably increases the risk of injury with no training benefit other than temporary competition readiness (peaking).

Unfortunately, sexy as it sounds, liberally calling strength training "powerlifting" drives too many newcomers into an ego-lifting mindset, attempting maximal singles for no good reason, or soon hopping on some extreme Russian peaking program they found after Starting Strength.

Learning the powerlifts and training for strength is great for anyone, but unless they actually plan to step on a platform and risk their skeletal integrity for white lights in the near future, I strongly recommend that people keep strictly to weights they can safely lift for 5+ reps per set, and only track the calculated (rather than tested) 1RM.

That's a particularly incompetent and irresponsible view of engineering.

The key to modeling anything correctly is to be aware of exactly what assumptions your choice of model corresponds to and how well they match the real-world processes underlying your data.

So you've fitted some curve, but why that curve? What are the implications of assuming linearity between your input and output domains? What have you assumed about the distribution of noise over your predictions? How about over your input features? How would you expect the bias and variance of your predictions to degrade if any of these assumptions no longer held?

As an engineer, if you don't understand the statistics behind your models enough to answer to such questions, their real-world applications aren't going to go far beyond wishful garbage-in-garbage-out number crunching.