You really don't need macstl unless you have a strong desire to use valarray in C++...for example, the ATLAS project http://math-atlas.sourceforge.net/ already uses Altivec (and SSE/SSE2, etc) wherever it results in a speedup. So, if your code does linear algebra, use ATLAS and you'll see an automatic speedup in many cases. Other projects such as fftw http://fftw.org/ include Altivec/SSE/SSE2 optimizations as well. ATLAS includes lots of other optimizations such as cache-blocking, loop-unrolling, etc. I don't know of macstl includes such optimizations, but I do know that ATLAS performance approaches the theoretical peak performance on G4/G5 for things like matrix-matrix multiplication.
We really shouldn't even call it an image format. Most people think of image formats as a way to compress and store image data for viewing or printing; things like JPEG or GIF or PNG.
DNG is a format for storing the data recorded by the CCD's in a digital camera. This data can of course be processed and displayed as an image, but DNG really isn't an image format exactly.
One-liner Mathematica solution to billboard puzzle
on
Google's Math Puzzle
·
· Score: 5, Informative
I wrote this in a few minutes in Mathematica, and found the answer to the first puzzle. The second puzzle was annoying so I just searched google for it instead.
I was really trying to believe what you were saying until I got to this:
They also have guaranteed solvers for nonlinear (and/or partial) DE's...
Are you joking? That is like saying "they have guaranteed solvers for any equation of any kind at all." What is meant exactly by a "gauranteed" solver?
Does that mean that the hundreds/thousands of papers published each year on numerical methods for differential equations are all a waste of time? Heck, all those authors could have saved themselves a lot of time and just picked up a copy of Aberth's book!
From what I can tell, it appears to be a method for transforming a polynomial into a differential equation some of whose solutions are roots of the polynomial. From that point I suppose one could use numerical methods for ODE's to find those solutions.
People here have been commenting that Newton's method works just fine for finding roots of polynomials. But, convergence can be quite slow, especially for unidentified multiple roots though, and for highly clustered roots you can run into conditioning problems.
The paper makes no mention of actual numerical algorithms (in particular no discussion of convergence rates or guarantees for solving the ODE numerically) so it is hard to say whether the result is actually useful or just a bunch of manipulation of symbols on paper.
But, there is the delay from the sound traveling from the speaker to your ear (roughly 1 millisecond per foot of distance traveled). So one solution is simply to put the speakers about 60 feet away.
I wondered about that too. Could it be that the panorama sky would show the sun in multiple places? After all, the panorama is made using multiple shots stitched together, and the sun moves across the sky during this process. Maybe that's why they edit out the sky in the panorama shots.
When you look at Mars in the sky, it looks red doesn't it? Also, when you look at it through a telescope, it looks red. Ergo, one might expect it would look red when viewed close up as well.
Here's my issue: "introversion" is not an inherent quality of an individual, the way "body mass" or "cranial volume" or even "eye color" are. The range of human behaviors is incredibly complex -- far too complex to be boiled down to a simple binary quality even when restricting behavior to a specific context. Yet, many people clearly derive great pleasure from determining their own (and others') personality "types" according to various instruments.
The great complexity of human behavior, plus the obvious pleasure people get from categorizing, leads me to view such "types" very suspiciously.
There is clear biological evidence for sexual differentiation. What actual scientific evidence is there that the distinction between "introverts" and "extroverts" is anything more than pop-psychology b.s.?
>There are two types of people in this world. Those who believe that people can be separated into two different types; and those who don't.
How about this: There are two types of people in the world: those who can be categorized, and those who can't.
Introvert vs. extrovert is a made-up distinction
on
The Introvert Advantage
·
· Score: 2, Insightful
The whole idea of splitting people into categories like this is really stupid. Hopefully, 100 years from now people will recognize that this is as ridiculous as phrenology (i.e. measuring bumps on your head to determine personality traits) is recognized to be today.
> It's a fairly simple matter of plugging the in the right >values for m, g, S, C, p, A, v0, and x0, and we'll settle this >right now. If anyone knows what those numbers are I >invite you to share them.
This makes a nice toy model but it won't cut it for estimating the relative velocity of the foam (which can be done more easily just by watching the video and using fifth-grade math: velocity = distance traveled/travel time).
First of all, the foam was very likely tumbling. So A isn't constant. Second of all, the foam was probably rougher in some places than others; so C isn't constant either. To solve both for the velocity of the foam and its orientation requires a nonlinear system of 7 differential equations. You'd also need to know the exact shape and surface characteristics -- clearly impossible to know at this point.
I'm an applied mathematician (numerics is my specialty). I use OS X for numerical computation, but I also use it to write papers with TeXShop (great program, by the way) and I use Powerpoint (soon Keynote) for presentations. With OS X I can use the best software for the job, whether it be unix/X11-based, or commercial, more traditionally Mac-like software. With the G5 I'll be able to use the best hardware too!
Actually the G5 has dual floating point units, which means it can do 2 double-precision floating point ops per cycle (maximum)...and technically speaking the altivec can do 4 single-precision floating point ops per second, which would allow for a theoretical peak of 16 GFlops per machine. Of course the actual performance would be lower, especially in a cluster. I don't think you'd need 123 machines though.
You know, sometimes they actually print things that are true. When I was in high school in the 80's, someone at my school got really drunk at a party and was dropped off at 2 a.m. by his "friends" in a snowbank near his house. He nearly froze to death - his body temperature got well below the level where normally you would die. His feet and hands were ice chunks, but he miraculously survived. The Weekly World News picked up the story, and reported it very accurately.
Umm...not sure what is meant by "OS X graphical version", but I'm running Matlab for OS X and its the full version, does everything that it does on other platforms, including fancy graphics. It works really well with Apple's X11 because then all the graphics are hardware accelerated.
Slashdot Mirror
You really don't need macstl unless you have a strong desire to use valarray in C++...for example, the ATLAS project http://math-atlas.sourceforge.net/ already uses Altivec (and SSE/SSE2, etc) wherever it results in a speedup. So, if your code does linear algebra, use ATLAS and you'll see an automatic speedup in many cases. Other projects such as fftw http://fftw.org/ include Altivec/SSE/SSE2 optimizations as well. ATLAS includes lots of other optimizations such as cache-blocking, loop-unrolling, etc. I don't know of macstl includes such optimizations, but I do know that ATLAS performance approaches the theoretical peak performance on G4/G5 for things like matrix-matrix multiplication.
c Lib.html includes ATLAS so you don't even have to download or install anything - it comes with OS X.
Not only that, but Apple's vecLib http://developer.apple.com/ReleaseNotes/MacOSX/ve
We really shouldn't even call it an image format. Most people think of image formats as a way to compress and store image data for viewing or printing; things like JPEG or GIF or PNG.
DNG is a format for storing the data recorded by the CCD's in a digital camera. This data can of course be processed and displayed as an image, but DNG really isn't an image format exactly.
I wrote this in a few minutes in Mathematica, and found the answer to the first puzzle. The second puzzle was annoying so I just searched google for it instead.
en = N[\[ExponentialE], 1000]; Table[x = (Floor[en*(10^k)*10^10] - Floor[en*(
10^k)]*10^10); If[PrimeQ[x], {k, x}, {k, 0}], {k, 0, 100}]
I was really trying to believe what you were saying until I got to this:
They also have guaranteed solvers for nonlinear (and/or partial) DE's...
Are you joking? That is like saying "they have guaranteed solvers for any equation of any kind at all." What is meant exactly by a "gauranteed" solver?
Does that mean that the hundreds/thousands of papers published each year on numerical methods for differential equations are all a waste of time? Heck, all those authors could have saved themselves a lot of time and just picked up a copy of Aberth's book!
From what I can tell, it appears to be a method for transforming a polynomial into a differential equation some of whose solutions are roots of the polynomial. From that point I suppose one could use numerical methods for ODE's to find those solutions.
People here have been commenting that Newton's method works just fine for finding roots of polynomials. But, convergence can be quite slow, especially for unidentified multiple roots though, and for highly clustered roots you can run into conditioning problems.
The paper makes no mention of actual numerical algorithms (in particular no discussion of convergence rates or guarantees for solving the ODE numerically) so it is hard to say whether the result is actually useful or just a bunch of manipulation of symbols on paper.
But, there is the delay from the sound traveling from the speaker to your ear (roughly 1 millisecond per foot of distance traveled). So one solution is simply to put the speakers about 60 feet away.
I wondered about that too. Could it be that the panorama sky would show the sun in multiple places? After all, the panorama is made using multiple shots stitched together, and the sun moves across the sky during this process. Maybe that's why they edit out the sky in the panorama shots.
When you look at Mars in the sky, it looks red doesn't it? Also, when you look at it through a telescope, it looks red. Ergo, one might expect it would look red when viewed close up as well.
for all my novels. Sometimes I even write in rot13 code just for fun.
The great complexity of human behavior, plus the obvious pleasure people get from categorizing, leads me to view such "types" very suspiciously.
There is clear biological evidence for sexual differentiation. What actual scientific evidence is there that the distinction between "introverts" and "extroverts" is anything more than pop-psychology b.s.?
If you'll forgive me, that sounds like something an extrovert would say
Most people who know me would call me an introvert. So there!
All those tests do is tell you how you answer the questions on the test. Everything else is someone's interpretation.
>There are two types of people in this world. Those who believe that people can be separated into two different types; and those who don't.
How about this:
There are two types of people in the world: those who can be categorized, and those who can't.
The whole idea of splitting people into categories like this is really stupid. Hopefully, 100 years from now people will recognize that this is as ridiculous as phrenology (i.e. measuring bumps on your head to determine personality traits) is recognized to be today.
...would make a nice work outfit for my brother the CFO.
> It's a fairly simple matter of plugging the in the right
>values for m, g, S, C, p, A, v0, and x0, and we'll settle this
>right now. If anyone knows what those numbers are I
>invite you to share them.
This makes a nice toy model but it won't cut it for estimating the relative velocity of the foam (which can be done more easily just by watching the video and using fifth-grade math: velocity = distance traveled/travel time).
First of all, the foam was very likely tumbling. So A isn't constant. Second of all, the foam was probably rougher in some places than others; so C isn't constant either. To solve both for the velocity of the foam and its orientation requires a nonlinear system of 7 differential equations. You'd also need to know the exact shape and surface characteristics -- clearly impossible to know at this point.
I'm an applied mathematician (numerics is my specialty). I use OS X for numerical computation, but I also use it to write papers with TeXShop (great program, by the way) and I use Powerpoint (soon Keynote) for presentations. With OS X I can use the best software for the job, whether it be unix/X11-based, or commercial, more traditionally Mac-like software. With the G5 I'll be able to use the best hardware too!
4 single-precision floating point ops PER CYCLE, not per second (oops)
Actually the G5 has dual floating point units, which means it can do 2 double-precision floating point ops per cycle (maximum)...and technically speaking the altivec can do 4 single-precision floating point ops per second, which would allow for a theoretical peak of 16 GFlops per machine. Of course the actual performance would be lower, especially in a cluster. I don't think you'd need 123 machines though.
They're wasting their time, don't they realize that mentalics will soon mess up all the models anyway?
You know, sometimes they actually print things that are true. When I was in high school in the 80's, someone at my school got really drunk at a party and was dropped off at 2 a.m. by his "friends" in a snowbank near his house. He nearly froze to death - his body temperature got well below the level where normally you would die. His feet and hands were ice chunks, but he miraculously survived. The Weekly World News picked up the story, and reported it very accurately.
Two words: Xserve RAID.
Umm...not sure what is meant by "OS X graphical version", but I'm running Matlab for OS X and its the full version, does everything that it does on other platforms, including fancy graphics. It works really well with Apple's X11 because then all the graphics are hardware accelerated.
I only did it because someone else was saying "how dare you attack the statements of a former MIT professor" or something like that.