HNHacker News
TopNewBestAskShowJobs

ethan_g

129 karma · joined April 29, 2016

submissionscomments
ethan_g··on An alarming number of scientific papers contain Excel errors
Commas are an exception to that. If you paste in "10,100", it will be interpreted as a number, and you lose the comma.
ethan_g··on An alarming number of scientific papers contain Excel errors
This is true for small spreadsheets, and one reason why I love Excel for quick analyses. Debugging is nearly instant because you see the results as you go along. But it doesn't scale very well.

--If you want to re-use a formula, you basically have to copy-paste it. Generally there isn't much modularity. Programmers are well aware of the dangers of this.

--It's also not a friendly platform for writing regression tests (possible with VB, but who does that?).

--The two-dimensional nature of the spreadsheets and lack of loops means you often need to do hacky stuff to emulate multidimensional arrays or loops. Easy to make mistakes here.

ethan_g··on An alarming number of scientific papers contain Excel errors
Java, for example.

Yes, it's true Excel is used for a lot of big important things, but I'd agree with GP it's not very good at them. Moreover, I think most people who use Excel for big important things, would agree it's not good for it--they get stuck in that situation when a small spreadsheet grows big over the years. Lack of version control, limited modularity, and no ability to unit test make large spreadsheets error-prone, and Excel tends to barf on you when you push it too hard (seems to be a number of crash bugs and performance bugs that never get fixed).

Excel is incredibly productive for quick analysis and making charts. But it's really not a development platform and it's really not for heavy lifting.

ethan_g··on Goodbye, Object Oriented Programming
There's some truth to this, but "can be optimized" is not the same as "there exists a compiler that optimizes". Your statement gets repeated a lot (and similarly about the JVM: it has run-time information that compilers don't have, so can optimize better), and yet C++ and Fortran continue to come out on top.
ethan_g··on Choosing 3M points on a 300-dimensional sphere
Yes it is, note in particular that the randoms are normal, not uniform.
ethan_g··on “Give me 15 minutes and I'll change your view of GDB” [video]
Well, "need" is a strong word. The features are definitely helpful though.

With large programs, there can be a really long distance between where something is created/modified and where it's used. If there's a bug you catch that occurs where the object is used, reversible debuggers help you trace back to where it's created, or who modified it last.

If you're at a place that's hard to put a useful breakpoint (e.g. a widely used function), and need to answer a what-if question, being able to manipulate state from the debugger is super useful. You might not be able to easily insert code to the program to create that state without messing up everything else (again, example is a widely used function, where inserting code could mess up other users of the function).

ethan_g··on Genetic evidence for natural selection in humans in the contemporary USA
To be clear, "we" is first person, plural. I take it your comment and gp to mean "I" by "first person" and not "we".

Anyway, it's standard in scientific writing to use "we" instead of "I". https://en.wikipedia.org/wiki/We#The_author.27s_.22we.22

ethan_g··on Feynman on Fermat's Last Theorem
Apologies to rain on the parade, but this sort of argument is very well known and commonly used in number theory. Since these techniques don't lead to proofs, they may not get as much attention in standard math education, but they are still well-known. Many number theory papers use this technique (often with more refinements) to estimate the number of expected solutions. I would be stunned if it wasn't used on Fermat's last theorem long before Feynman did.

For an exposition on some heuristics mathematicians use, https://terrytao.wordpress.com/2015/01/04/254a-supplement-4-... is a good read (although it's aimed at readers already having some background in number theory).

ethan_g··on Integer percentages as fingerprints of electoral falsification
Benford's law applies when the underlying distribution is approximately exponential. Because candidates with an exponentially small chance to win are not likely to run, it wouldn't make much sense for elections to have an exponential distribution. Much more plausible distributions are (truncated) normal or uniform, neither of which satisfies Benford's law.
ethan_g··on Why you shouldn't exercise to lose weight
Building muscle can burn crazy calories though. It's quite common for normal men (I'm not talking about pro athletes, just ordinary guys) on strength training programs to add 500-1000 calories/day to their diet, and not gain fat. I wish the article would have mentioned this, because it only seems to talk about cardio exercise, which indeed makes it difficult to burn a lot of calories.
← PreviousPage 2 of 2