Federal Reserve Bank of NY converts major economic model to Julia • Hacker News Board

Re: Federal Reserve Bank of NY converts major economic model to Julia

#71

Post by bachmeier » Fri, Dec 04, 2015, 1:51 AM UTC

grayclhn wrote: post #59

bachmeier wrote: post #19

The real news here is not the move to Julia (although that is likely to be the reason for most of the attention). The important thing is the move to open - an open language, movement to Github, nice explanation of the details.

No. A central bank moving at all to Julia is absolutely fucking huge.

I suppose it depends on whether you think it's important for the popularity of Julia (something I don't care much about, since it's just a programming language) or because you think open science and policymaking is important (which I do).

[I see you're from Iowa State. I'm posting this from the house I bought from John Crespi a few months ago.]

Re: Federal Reserve Bank of NY converts major economic model to Julia

#72

Post by grayclhn » Fri, Dec 04, 2015, 1:57 AM UTC

pzone wrote: post #62

grayclhn wrote: post #59

Earlier quoted context omitted.

No. A central bank moving at all to Julia is absolutely fucking huge.

Agreed. I worked on the FRBNY research floor, this model has real mindshare with policy makers, porting it is a huge investment in the language. They may be training scads of RAs in writing and using Julia, many of them will go off and do PhDs later.

This also provides a lot of justification/support for those of us who'd much rather teach (say) macroeconometrics in julia instead of matlab.

Re: Federal Reserve Bank of NY converts major economic model to Julia

#73

Post by aswanson » Fri, Dec 04, 2015, 1:57 AM UTC

grayclhn wrote: post #58

aswanson wrote: post #22

Earlier quoted context omitted.

Why? Because professors keep assigning students problems with it, immune to its high cost and suspect coding idioms because they aren't professional programmers.

As (potentially) one of those professors, I'm going to suggest that if you're posting on HN, you have no idea how little computer experience some of the Econ, engineering, etc. students have. Matlab and other programs like it try hard to smooth over some of the rough edges that can trip up complete novices. As long as the main focus of the class is teaching "something other than programming" that's going to be pretty…

I was one of those clueless engineering undergrads back in ancient times...I know the mentality of thinking since you know BASIC and can code up a brutally ugly solution in MATLAB that you know programming, as long as the right number is spit out on the command line. I also know, in my old age, that if the purpose of an engineering education is is about building higher order structures that can communicate intent and be extensible artifacts, that this is not acceptable. We need to put in place tools and methodologies that enforce good practice with code in the same way that we enforce good spelling, grammar, sentence, and paragraph structure in language, from the beginning.

Re: Federal Reserve Bank of NY converts major economic model to Julia

#74

Post by bachmeier » Fri, Dec 04, 2015, 1:59 AM UTC

GFK_of_xmaspast wrote: post #16

wyager wrote: post #8

Earlier quoted context omitted.

> Matlab will always have its proponents I honestly can't understand why. I'm currently in a computer vision class that uses Matlab, and it's literally the worst language I've ever used in my life. The design is nonsensical, bordering on malicious I've always thought that the biggest problem with python was that as programs get bigger, you're likely to run into some confusion due to the lack of static typing and subs…

It works well enough and there's an incredible amount of infrastructure behind it. It's an academic's job to produce research, not code.

That was the advice I got from my advisor as a grad student. 15 years later, I disagree strongly. I wasted a lot of time writing code that I couldn't read later, that took forever to get right, and that couldn't be reused. I've made a significant investment in programming tools and it has been worth it. Git alone has probably made me twice as productive.

The logic you state (which is dominant) is wrong. Optimizing the speed of writing the first version of a particular piece of code is a bad idea.

Re: Federal Reserve Bank of NY converts major economic model to Julia

#75

Post by grayclhn » Fri, Dec 04, 2015, 2:10 AM UTC

aswanson wrote: post #73

grayclhn wrote: post #58

Earlier quoted context omitted.

As (potentially) one of those professors, I'm going to suggest that if you're posting on HN, you have no idea how little computer experience some of the Econ, engineering, etc. students have. Matlab and other programs like it try hard to smooth over some of the rough edges that can trip up complete novices. As long as the main focus of the class is teaching "something other than programming" that's going to be pretty…

I was one of those clueless engineering undergrads back in ancient times...I know the mentality of thinking since you know BASIC and can code up a brutally ugly solution in MATLAB that you know programming, as long as the right number is spit out on the command line. I also know, in my old age, that if the purpose of an engineering education is is about building higher order structures that can communicate intent and…

I agree with you, but that's not a decision that can be made unilaterally by professors at the course level without sacrificing _a lot_ of the other content.

You're going to cry if you see the current standards of undergraduate spelling, grammar, and sentence structure, btw. :)

Re: Federal Reserve Bank of NY converts major economic model to Julia

#76

Post by grayclhn » Fri, Dec 04, 2015, 2:18 AM UTC

bachmeier wrote: post #71

grayclhn wrote: post #59

Earlier quoted context omitted.

No. A central bank moving at all to Julia is absolutely fucking huge.

I suppose it depends on whether you think it's important for the popularity of Julia (something I don't care much about, since it's just a programming language) or because you think open science and policymaking is important (which I do). [I see you're from Iowa State. I'm posting this from the house I bought from John Crespi a few months ago.]

Small world, his office is the floor below mine :)

I think we probably agree then --- but this announcement isn't exactly a sea change in open science and policymaking. This model (or one very close to it) has been available on various people's webpages for a while [1], the research on estimating and using DGSE models for forecasting and policy-making has been well documented and openly discussed, etc. And the NY Fed has put out a lot of "accessible" information about features of the model and interpretations of its results.

[1]: http://sites.sas.upenn.edu/schorf/pages/working-papers

Re: Federal Reserve Bank of NY converts major economic model to Julia

#77

Post by aswanson » Fri, Dec 04, 2015, 2:24 AM UTC

grayclhn wrote: post #75

aswanson wrote: post #73

Earlier quoted context omitted.

I was one of those clueless engineering undergrads back in ancient times...I know the mentality of thinking since you know BASIC and can code up a brutally ugly solution in MATLAB that you know programming, as long as the right number is spit out on the command line. I also know, in my old age, that if the purpose of an engineering education is is about building higher order structures that can communicate intent and…

I agree with you, but that's not a decision that can be made unilaterally by professors at the course level without sacrificing _a lot_ of the other content. You're going to cry if you see the current standards of undergraduate spelling, grammar, and sentence structure, btw. :)

Touche, Prof. Carry on. ;)

Re: Federal Reserve Bank of NY converts major economic model to Julia

#78

Post by RockyMcNuts » Fri, Dec 04, 2015, 2:26 AM UTC

josu wrote: post #43

otoburb wrote: post #36

Earlier quoted context omitted.

>>He also told me to get away from the Sargent and Stachurski course, that they add too much complexity to the explanations just for the sake of looking more Nobel-prize-winner like. That's really good feedback for anybody thinking of attempting that course. Does your professor recommend an accompanying course to Sargent and Stachurski's quantitative economics course? I haven't seen many other courses that provide co…

Nope, I'm sorry, he just told me that I should get into Python. He did recommend this edX course, but it's an introduction to Python, nothing to do with economics. [1] https://www.edx.org/course/mitx/mitx-6-00-1x-introduction-co...

IMF course looks interesting https://www.edx.org/course/macroeconomic-forecasting-imfx-mf...

Sargent course looks pretty intense. IMF course looks maybe more approachable, uses Eviews which is a time-series-oriented statistical modeling package.

Re: Federal Reserve Bank of NY converts major economic model to Julia

#79

Post by grayclhn » Fri, Dec 04, 2015, 2:51 AM UTC

danso wrote: post #13

For those of you in this field...how influential is the NYFed trying out (and apparently being satisfied with) Julia to other government agencies, computational economists, and/or academics? Is it "OS X is Unix/BSD-based" influential, or "Whitehouse.gov uses Drupal [1]" influential? [1] http://buytaert.net/whitehouse-gov-using-drupal

I work on research close to this area -- time series w/ macro applications, but not on these specific models. These specific models are very very important for estimating the effects of monetary policy decisions, forecasting, etc. because they embed a fair bit of economic theory with a decent amount of estimation theory so that the models can address questions about "causality" in a meaningful way: if I intervene by changing the interest rate, what is the distribution of outcomes that will result. (The stuff I work on uses less economic theory so it can't address those questions.)

That's not to say that the NY Fed takes the model's recommendations uncritically and assumes that the models are perfect or the recommendations are the truth, because that would be nuts. Just that this class of models is one of the few ways to use "the data" to provide guidance about policy decisions.

The people who've developed the NY Fed's DSGE model are some of the absolute top people working on these models. It's based on a lot of academic research and a lot of research published in top journals by people at the NY Fed involved in building the model. Other banks have their own models, but I suspect this one is the highest profile. (Again, not precisely my field of research, so I could be missing one or two.)

So, to get around to answering your question. For academic research in economics, this is as "production level" as it gets. Is this the same as using Julia for algorithmic trading? No. The NY Fed's not running this model in real time, and if something crashes then they can restart it. But I think this is huge for Julia's adoption in economic research.

(For a pdf with more about the modeling strategy itself: https://www.newyorkfed.org/medialibrary/media/research/staff... but that is 2 years old now.)

Re: Federal Reserve Bank of NY converts major economic model to Julia

#80

Post by devonkim » Fri, Dec 04, 2015, 3:10 AM UTC

krapht wrote: post #66

mikeskim wrote: post #65

Earlier quoted context omitted.

You can do all of those things in R except maybe Simulink with a GUI (at least not yet, maybe soon?), and you have the choice of working with multiple implementations. https://cran.r-project.org/web/packages/FKF/FKF.pdf https://stat.ethz.ch/R-manual/R-devel/library/stats/html/Kal... http://stackoverflow.com/questions/1738087/what-can-matlab-d... Unless you are doing something very specific and require a specific tool…

I can't disagree more with a blanket statement like that. I would never recommend Python / R to the physicists and scientists that I work with. They don't know how to program well. The code they put out is usually garbage, hard to read, and fragile. Software development is not something they do. Backing up for them is emailing themselves a .zip file of their work and revision control is saving their DocXs and PPTs wi…

After having worked with enough academicians, the code that I had to deal with was too agonizingly crufty for me to deal with that it made a huge influence over my decision to away from academia (and my subsequent drop in grades in school). I figured that if I had enough problems just dealing with the code, I couldn't imagine having to trudge through everything else that led up to that code. But it shattered my preconceived image of academicians being curious about everything and having a great deal of care and pride in the tooling they use - that I should look to them as reference rather than a bunch of people that dropped out of school for a quick buck, and that simply wasn't true whatsoever.

Matlab is going to be the COBOL of academia without some major groundbreaking research that invalidates or supersedes a great deal of research code, and given the community-consensus nature of so much research done I can't imagine quite something so disruptive happening.