I love beer but generally dislike lagers (or at least american style lagers), so I'll often ask for a mystery beer and specify "anything except for a lager." I've truly yet to be disappointed. It's hard to imagine you couldn't do the same with food.
I was wondering if you had any advice for hardware startups?
I'm currently exploring the idea of integrated-circuit spectroscopy, and have the physics/simulations down but the capital required to fabricate the chips and iterate on the designs is quite high.
Assuming 7% returns, and 6% employee match of a graduate student's salary in my programme (32k), graduate school costs ~288k in lost retirement savings, as well.
One of the reasons that FCIQMC doesn't have nuclear motion is that the gradients from FCIQMC, and actually most QMC techniques, are really computationally intensive, so this means that creating the ab-initio surface for the nuclei to roll over is really hard. Perhaps you were considering some sort of FCIQMC approximation to the path integral, but it's not entirely obvious to me how this would work.
As for how DFT would work for this... It should work quite well for qualitative predictions. Actually, DFT does remarkably well for metals and functionals like asymptotically corrected PBE0 are providing remarkable physical insight. While I wouldn't trust the numbers that come from any DFT simulation to three decimal points, I'd certainly trust the physics that's captured.
That being said, metallic hydrogen should be a strongly multireference system, so I'd be interested in seeing how a green's function approach based in many-body perturbation theory (see GF2 from Zgid at U Michigan) would do, as it doesn't struggle with issues of references while still giving you coupled cluster level accuracy.
We're getting around this by putting a paragraph at the end of the conclusion with a link that says something to the effect of "Click here for a live version of the paper with all the code and data used in this project." It's a little more cumbersome than having the paper actually be the notebook but it helps.
I'm trying to push my research group to release our papers as iPython notebooks and it looks like my next paper will be. Our text will be rendered as normal markdown and our figures will be live figures with all the code that we used to generate them. This way, if anyone wants to check our work, extend it, or just better understand it, they're more than free to do so. It should be a nice step forward in the ultimate goal of reproducible science.
That's rather presumptuous to assume that he didn't. In fact, he did and as part of our grad student resources, we have access to a whole host of school specific recruiters. While students from our programme regularly go into the postdoc mill, many go to work at startups, at industrial labs, in finance, at consulting firms, etc. Even more important, though, none of these options are denigrated.
When I started my PhD last year, my department head gave us a speech where he told us that college football players have better career prospects in the NFL than we will in academia. While that was probably a bit of hyperbole, it certainly made us realize how unrealistic most of our goals were.
The top physics kids in the US tend to get picked up by financial (or some other industry) firms because they're really really sharp because making 200k sounds much better than making 30k. This means that the kids who go into physics are doing it because they truly love the subject but it also means that we have a much smaller percentage of kids going into graduate studies than otherwise. This, combined with the fact that many countries are rapidly industrializing and newly able to support an academic class means that lots of international students are applying to Harvard/MIT/Stanford/etc where some of the best science in the world is happening. Some of the foreign kids go back, some stay here. It really depends. As for makeup? probably 40%-50% of my entering cohort was foreign and probably 60% of the foreigners were asian. That's not really a bad thing though. They're brilliant scientists and it's truly a pleasure to collaborate with them and everyone else in my programme as well.
Disclaimer, I'm a physics PhD student at Harvard and go to seminars a lot with kids from MIT.
Optimal, at least in this sense, refers to the optimization of a cost function so we can quibble about the cost function used but optimization in general isn't nonsensical. And really, the OP specified that he was using the Patero sense of optimization with a Patero cost function.
Locality in this sense refers to the presence of "hidden variables" as first put forth by bohm and later suggested by other people. It was believed that the various CHSH, HOM, and other bell-inequality inspired experiments would have put a nail in the coffin but critics always came back and say "but you didn't control for X." This experiment is just another in the long line of experiments started by Aspect and continued by others to put bounds on this.
Don't worry, I felt the same when I was starting to read papers as a graduate student but I came to realize that in cases like the one where you mentioned elliptic modular forms being introduced to vastly overcomplicate something, it was usually done to make something true in a far more general set of space.
For me in physics? That came about when I started reading papers where people were doing stuff with differential geometry on manifolds. I couldn't, for the life of me, figure out why people would talk about a wedge product when a cross product would have sufficed until about 6 months later when it clicked that doing something in a coordinate agnostic framework allows you to prove things for any coordinate system and create general formulas that just need a few things "plugged in."
I'm not going to pretend to know that that's why elliptical modular forms are being used in your context but everytime something has seemed needlessly overcomplicated, I've come to realize after some thinking that it's done with a view towards generality.
Speaking as the author of multiple academic articles (theoretical physics), I would fundamentally disagree with your assessment. We usually are trying our absolute hardest to communicate a new advance in the field in anywhere from 4-8 pages, references and introduction included. We're essentially trying to sum up (in my case) 2 years of work in as short of a space as possible so, to do this, we assume that the reader has a working knowledge of the foundations of the field but provide references to this. The references serve to 1) provide evidence for unoriginal claims that you make (every sentence that communications an unoriginal result should have a citation) and 2) allow those who are unfamiliar with the field to pick up the basics as quickly as possible.
Unfortunately, some of our prose is a bit obtuse because it's reasonably common for scientists to blow off their humanities courses because they're not science -- not realizing that most of our career will depend on the quality of our writing. You shouldn't confuse our incompetence with malice though.