michaelmas

Graduate statistics programs emphasize R. Graduate applied math programs emphasize Matlab. Graduate quantitive finance programs emphasize Matlab/C++. Graduate computer science programs emphasize Java/C/C++. Just my observations, but I do not remember having seen a graduate course where Fortran/Python are taught specifically. I'm sure there are instances, but Fortran/Python as well as other languages to me seem to exist primarily out of historical reasons and not due to programming feasibility. For numerical problems, the leaders seem to be R, Matlab, and C++. I at least wouldn't bother to learn any other than R/Matlab/C/C++/Java.

While both syntaxes are related, R is primarily a statistical software and Matlab is primarily a numerical optimization software. Each has its comparative advantage. Doing Macro in R is a waste of time, while doing statistics/econometrics in Matlab is (as noted above) masochistic.

1) Obtain a short tutorial to get you started if you've never encountered Matlab/Octave, and work through it. 2) Find a suitable problem set to work on. 3) If you get stuck, Google it.

Do grades really matter all that much when adcoms have LORs and various of your rankings among the, say, 100 UG students? All I'm saying, and feel free to disagree, is that (absolute) grades are the part of your application that is least relevant in practice. Grades just don't work. And don't get me started on grade conversion.

If using matlab is only a matter of preferences and not absolutely required for the work you're performing, I strongly suggest you familiarise yourself with R. R is a specialised statistical software and a programming language. Its syntax is very similar to matlab's and the number of its pre-programmed functions is enormous.

I'm going as Sargent Bilko.

Just in case I was being ambiguous, I want to iterate that I am not here to degrade anyone's academic work or beliefs.

You're misunderstanding my point on a grand scale. Say you raise the interest rates from 0.25% to 5% today, with the effect that unemployment becomes 25%. If you would instead have done it in 1990, would you have gotten the same effect? 1960? 1925? Hardly. Why? Massive structural changes plus the unpredictable human factor plus black swan events, plus perhaps something more. So what have you learned? That most likely, the unemployment rate will increase substantially when you raise the interest rates. That's a qualitative result, not a quantitative one. The more quantitative you want your result to be, the more conditional and non-universal it will have to be at the same time. And to me (not necessarily you) less valuable. No, not difficult, but complex and unpredictable to the point of the theories being unusable.

I'm pretty sure there are one or two posts from myself on TM from the past, quoting this bit. But I have not laid out any proof in this thread, not so much as evidence, that the economy behaves in a mathematical fashion, and neither does Sargent to the contrasting view. Maybe there's someone here with an undergraduate degree in physics and can correct me, but is there really a divide between physicists as there is between macroeconomists? I'm sure there are qualified physicists that have theirs to say against string theory, for example, but the empirical testing of string theory is as of yet at its infant stage. String theory is what it is currently, just a theory. It's still science because it is falsifiable. They are even testing some of its implications at CERN. We can argue all we like about the theory, but when a theory is properly rejected in physics, that's usually it, and scientists need the intellectual capacity of Newton, Einstein or Feynman to claim otherwise. If the economy really behaved in a mathematically exact way, would we really see a divide between the hardcore Keynesians on the one hand, and the equilibrium theorists on the other?

I'm sure many physicists dislike something about their field, but my discontent with economics concerns its very foundation. Far more fundamental. But I think you're right. If I quit now, I'll probably lose a precious LOR I might need in the nearest future, if I decide to pursue another masters degree in a real discipline. I hate citing Austrian papers because it's something I so not identify with, but "On Mathematical Thinking in Economics" (1977) by Leoni and Frola, as well as "Comments about the Mathematical Treatment of Economic Problems" (1977) by Mises have helped my realisation. Also throw in Taleb's very convincing "The Black Swan". Last but not least, I have become interested in non-technical physics for non-physicists by reading books and listening to lectures and interviews on YouTube of the authors discuss what physics is all about. I think economists should look at what real scientists do, and what allows them to mathematize the real world around us. Their equations are not conditional on time, place or opinions like ours. They are more or less exact, have always applied and will always apply.

I have ceased to believe that economic agents, be it individuals or collective, behave in a manner easily representable in mathematical terms. I have ceased to believe that we can predict the behaviour of humans. I have ceased to believe that we can mathematically model expectations quantitatively. I have ceased to believe that black swan events are inconsequential. I have ceased to believe that economics is science. I have ceased to believe that neoclassical economic theory is anything more than BS. I have ceased to believe that my final MA thesis will will be anything other than BS. I feel like I've wasted the last four years of my life, only to acquire working knowledge of tools that engineers and proper physical scientists have mastered. When I should have taken more college courses in physics and engineering, and eventually apply to physics, engineering or statistics, I blindly chose neoclassical economic theory in the belief that I was going to study something that had a point. Don't get me wrong. I am not a follower of the Austrian school and I am not arrogant enough claiming to be holding the truth in my hands. This has to do with my beliefs alone, based on evidence and my limited inferential abilities. To me, the bulk of the Austrian material (at least the loudest one) seems to be nothing more than pure dissing of neoclassical economic theory. I am relatively close to getting my MA degree in economics. What should I do? I have already chosen a thesis subject, and I have already concluded that it would be nothing more than pure academic dishonesty of me to go ahead with it. It has no meaning to me, and I have no belief that it will matter at all in even the remotest reality. Should I just quit, work for a couple of years, and then apply to computer science or statistics to get some real education? (You don't have to answer any of these questions. It's just a rant.) Or maybe you could prove me wrong. I'd love that to happen, honestly.

We need to rethink how we treat the extreme cases like perfect competition and monopoly, not whether to treat them at all. We need to understand all aspects of the extreme cases, because reality combines different aspects of different poles. We wouldn't have monopolistic competition in our toolboxes if we hadn't have recognized monopoly and perfect competition first. Undergraduate students are, in my view, not taught what and understanding of these extreme cases is good for.

Given how TM is obsessed with math courses as signals, I'd say doing well in Complex Analysis sends a pretty good one. Take my advice with a grain of salt, though.

Should be: I don't want to be 'that guy' and I'll probably get negative reviews for this, but is there really any sane applicant thinking "I'm in at Princeton, but the rejection from Harvard is simply to much for me to handle"?

You would not believe the stuff I would do the be good enough for, say, Georgetown or Columbia.