Re: One more bit for Veterans Day
-- Original message from Dave Land [EMAIL PROTECTED]: -- On Nov 12, 2008, at 9:22 AM, Olin Elliott wrote: As one of the comments says, Not all heroes have two legs I say I know how an old war dog feels a lots like the homeless veterans wandering the streets and by the waysides! http://kink9570.wordpress.com Pvt. PV On Nov 12, 2008, at 9:22 AM, Olin Elliott wrote: As one of the comments says, Not all heroes have two legs Thanks for this, Ronn. I have worked for a group in the past trying to get a national war dog memorial and the DOD has steadfastly opposed it. A very different look at the dogs of war. Its funny when you think about it, dogs serve in our military, our police forces, they care for the sick and disabled, they work on farms, they are our companions and members of our family -- in fact, they've been a part of our civilization for about 15,000 years by the most convervative estimate. They helped build our civilization, and in every way that counts, they are true citizens. Maybe one day, that will be recognized legally. Only, they won't be allowed to marry in certain states. Dave ___ http://www.mccmedia.com/mailman/listinfo/brin-l ___ http://www.mccmedia.com/mailman/listinfo/brin-l
One more bit for Veterans Day
As one of the comments says, Not all heroes have two legs . . . http://punditkitchen.com/2008/11/11/political-pictures-veterans-thank-you/ . . . ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit for Veterans Day
As one of the comments says, Not all heroes have two legs Thanks for this, Ronn. I have worked for a group in the past trying to get a national war dog memorial and the DOD has steadfastly opposed it. There are memorials around the country, mostly created by veterans to remember the canines who served with them. The government doesn't consider war dogs to be personel, but equipment. We abandoned thousand of service dogs in Vietnam when we left. Recently, there has been some progress -- first President Clinton and then President Bush created policies to make it easier for military personel to adopt the dogs that served with them. Its funny when you think about it, dogs serve in our military, our police forces, they care for the sick and disabled, they work on farms, they are our companions and members of our family -- in fact, they've been a part of our civilization for about 15,000 years by the most convervative estimate. They helped build our civilization, and in every way that counts, they are true citizens. Maybe one day, that will be recognized legally. Olin - Original Message - From: Ronn! Blankenshipmailto:[EMAIL PROTECTED] To: Killer Bs Discussionmailto:brin-l@mccmedia.com Sent: Wednesday, November 12, 2008 8:52 AM Subject: One more bit for Veterans Day As one of the comments says, Not all heroes have two legs . . . http://punditkitchen.com/2008/11/11/political-pictures-veterans-thank-you/http://punditkitchen.com/2008/11/11/political-pictures-veterans-thank-you/ . . . ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-lhttp://www.mccmedia.com/mailman/listinfo/brin-l ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit for Veterans Day
On Nov 12, 2008, at 8:52 AM, Ronn! Blankenship wrote: As one of the comments says, Not all heroes have two legs . . . http://punditkitchen.com/2008/11/11/political-pictures-veterans-thank-you/ Wow -- exactly the opposite from what I was thinking: Not all heroes have _only_ two legs, rather than not all heroes _still_ have two legs. I will withhold my usual comment about dog people. Dave ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit for Veterans Day
On Nov 12, 2008, at 9:22 AM, Olin Elliott wrote: As one of the comments says, Not all heroes have two legs Thanks for this, Ronn. I have worked for a group in the past trying to get a national war dog memorial and the DOD has steadfastly opposed it. A very different look at the dogs of war. Its funny when you think about it, dogs serve in our military, our police forces, they care for the sick and disabled, they work on farms, they are our companions and members of our family -- in fact, they've been a part of our civilization for about 15,000 years by the most convervative estimate. They helped build our civilization, and in every way that counts, they are true citizens. Maybe one day, that will be recognized legally. Only, they won't be allowed to marry in certain states. Dave ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit for Veterans Day
Only, they won't be allowed to marry in certain states. One commentator on NPR noted the irony of several states rejecting Gay marriage in the same election where Barack Obama became president. When Obama was born, quite a few states wouldn't have allowed his white mother and African father to be legally married. And the arguments against it would have been pretty much what we hear from anti-gay groups today. Olin - Original Message - From: Dave Landmailto:[EMAIL PROTECTED] To: Killer Bs (David Brin et al) Discussionmailto:brin-l@mccmedia.com Sent: Wednesday, November 12, 2008 4:08 PM Subject: Re: One more bit for Veterans Day On Nov 12, 2008, at 9:22 AM, Olin Elliott wrote: As one of the comments says, Not all heroes have two legs Thanks for this, Ronn. I have worked for a group in the past trying to get a national war dog memorial and the DOD has steadfastly opposed it. A very different look at the dogs of war. Its funny when you think about it, dogs serve in our military, our police forces, they care for the sick and disabled, they work on farms, they are our companions and members of our family -- in fact, they've been a part of our civilization for about 15,000 years by the most convervative estimate. They helped build our civilization, and in every way that counts, they are true citizens. Maybe one day, that will be recognized legally. Only, they won't be allowed to marry in certain states. Dave ___ http://www.mccmedia.com/mailman/listinfo/brin-lhttp://www.mccmedia.com/mailman/listinfo/brin-l ___ http://www.mccmedia.com/mailman/listinfo/brin-l
RE: One more bit for Veterans Day
At least according to the local TV news tonight, they now have a first-rate veterinary facility to treat their wounds. There were probably times when they would have just been shot. http://idiotgrrl.livejournal.com/ From: [EMAIL PROTECTED] To: brin-l@mccmedia.com Subject: Re: One more bit for Veterans Day Date: Wed, 12 Nov 2008 17:50:15 -0800 Only, they won't be allowed to marry in certain states. One commentator on NPR noted the irony of several states rejecting Gay marriage in the same election where Barack Obama became president. When Obama was born, quite a few states wouldn't have allowed his white mother and African father to be legally married. And the arguments against it would have been pretty much what we hear from anti-gay groups today. Olin - Original Message - From: Dave Landmailto:[EMAIL PROTECTED] To: Killer Bs (David Brin et al) Discussionmailto:brin-l@mccmedia.com Sent: Wednesday, November 12, 2008 4:08 PM Subject: Re: One more bit for Veterans Day On Nov 12, 2008, at 9:22 AM, Olin Elliott wrote: As one of the comments says, Not all heroes have two legs Thanks for this, Ronn. I have worked for a group in the past trying to get a national war dog memorial and the DOD has steadfastly opposed it. A very different look at the dogs of war. Its funny when you think about it, dogs serve in our military, our police forces, they care for the sick and disabled, they work on farms, they are our companions and members of our family -- in fact, they've been a part of our civilization for about 15,000 years by the most convervative estimate. They helped build our civilization, and in every way that counts, they are true citizens. Maybe one day, that will be recognized legally. Only, they won't be allowed to marry in certain states. Dave ___ http://www.mccmedia.com/mailman/listinfo/brin-lhttp://www.mccmedia.com/mailman/listinfo/brin-l ___ http://www.mccmedia.com/mailman/listinfo/brin-l ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Nick Arnett ([EMAIL PROTECTED]) wrote: On Wed, 20 Apr 2005 19:42:47 -0400, Erik Reuter wrote Yeah, how could you ever in a million years have guessed that the function for calculating a factorial was called factorial? I'm having trouble getting your joke. No doubt. -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 11:04 PM Tuesday 4/19/2005, Warren Ockrassa wrote: On Apr 19, 2005, at 8:34 PM, Ronn!Blankenship wrote: At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: If you can get me a program to run it, http://store.wolfram.com/catalog/ US$900! (Won't bother trying to calculate that.) Ouch! I hadn't realized Mathematica was that significantly priced. Humm. Humm. Googlicious ... OK, here's an interesting package called HartMath, which appears to do symbolic calculations, is Java based (!) and is free: http://sourceforge.net/project/showfiles.php? group_id=5083package_id=9297 I DLed the .jar, the first on the list, 0.8pre2. And when I entered 900! and clicked the Symbolic evaluation button I got a damn big number, 675268022096458415838790613618008142242694278695893843121982687036850916431804 169691324469526983037942260103705786729085931983476998869285919065010315876518 469767596811126095247870938480044286361868933952727844506303540802432176466580 246966590659517937572235202292355775486538336811021709738937460546491264159091 431501728607211566858106557592300114501329921764549832275386963401126104470290 023370048878772663877045860772935854331516125188001477644611826808228670927866 949828318386418009974998193392065794153256497484862652339189110871145924408965 940626759142949258167198621783746792720926375247869390362900359242717822537380 598869339234478777695830030167053633390314130691558375185247610783420526354756 321131696187745492757014801069333629900037325893705935573252994347344592958667 289887407941746543914799260008488466867087297367132072852037127322012724108308 369130526353650828887251716360815871516034682911067546403982321466736273708959 340907778288275495542324361904648279986839271792460299194432510264644523379395 991985282978285911226899606203612382483131580716433958484050472614126800398777 337618498744473238679117126300231717459682784657805585680670350138852750802921 373604918751649477244642216935337550353000653500651374908320395233829637470261 856530503318323809918448425607509235437751885820964874769502544183651989996746 844172862654427866515944047816229469018791663829307141969082274601330276058178 648773777121931421376254303537184482693907326157766452831988286029176802240410 889938926105068021959172478389001069106980570303791905710576058493231133086344 520081798811656164497676483541612250669679612976096987427379233893916152074411 523193928456876733118992470853277034218629728716444954095722599855632154714820 833256532317771132713265799703107556049739697089494773742549744802946524270224 367053801840640088534572145185152709855631954129931452740576886344488124494458 006176311627682431256064248447093720221499084635722549126549077634457585439809 991491229981043789656267818986552214432636014051520731997065850802887350402054 173712772530962432 00 00 , perhaps? which was much more satisfying than the infinity symbol I got when I clicked Numeric evaluation. This might show more promise for calculating 5,565,709! than a basic Java hack. Oh, I could do it the same way I did the others. The only question is how long it would take to run. -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 11:08 PM Tuesday 4/19/2005, Dan Minette wrote: - Original Message - From: Warren Ockrassa [EMAIL PROTECTED] To: Killer Bs Discussion brin-l@mccmedia.com Sent: Tuesday, April 19, 2005 11:04 PM Subject: Re: One more! On Apr 19, 2005, at 8:34 PM, Ronn!Blankenship wrote: At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: If you can get me a program to run it, http://store.wolfram.com/catalog/ US$900! (Won't bother trying to calculate that.) Ouch! I hadn't realized Mathematica was that significantly priced. Humm. Humm. Googlicious ... OK, here's an interesting package called HartMath, which appears to do symbolic calculations, is Java based (!) and is free: What was Microsoft FORTRAN and then became Compaq FORTRAN and may be HP FORTRAN now has IMSL installed. It is a wondrous callable library of mathematical functionsand I got it with FORTRAN for a few hundred dollars for the package. If you want to throw fits at numbers, that's the way to goas well as do a zillion other things. Mathamatica is a toy in comparison. Which one is best depends on what you are trying to do atm. (I have both, tho I haven't gotten around to reinstalling the FORTRAN on this machine yet.) In this particular case, doing it in FORTRAN would be a little more complicated than typing Factorial [100] and pressing Shift+Enter . . . -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Apr 19, 2005, at 11:01 PM, Ronn!Blankenship wrote: At 11:04 PM Tuesday 4/19/2005, Warren Ockrassa wrote: http://sourceforge.net/project/showfiles.php? group_id=5083package_id=9297 I DLed the .jar, the first on the list, 0.8pre2. And when I entered 900! and clicked the Symbolic evaluation button I got a damn big number, 67526802209645841583879061361800814224269427869589384312198268703685091 6431804 [...] , perhaps? Or thereabouts, yeah, though I don't feel too inclined to do a digit-for-digit check... I tried 5,565,709! as well, and the program didn't die, but it set up a wait cursor that tells me nothing at all about how long it would take to do the calculation. After about an hour I shut it down and went on to other things. Java's convenient but its VM can really suck up the silicon, especially when all you have is a 600 MHz G3. :\ -- Warren Ockrassa, Publisher/Editor, nightwares Books http://books.nightwares.com/ Current work in progress The Seven-Year Mirror http://www.nightwares.com/books/ockrassa/Flat_Out.pdf ___ http://www.mccmedia.com/mailman/listinfo/brin-l
computers are evil, why they must be eradicated [was: One more!]
Julia Thompson wrote about 1,000,000!: If you can get me a program to run it, I could do it here Not sure how long it would take, but it would get done, anyway. n! = (n/e)^n sqrt(2 pi n) (1 + 1 / (12 n) + ...) So, using log10, we easily get: 1,000,000! = 8.264 x 10^5,565,708 or something like that Alberto Monteiro ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Julia Thompson ([EMAIL PROTECTED]) wrote:
If you can get me a program to run it, I could do it here Not
sure
5565709! has 35 126 456 digits and took 7 minutes 57 seconds to
calculate and write to disk. Don't ask me to calculate that factorial,
though, because the last calculation took up about 25% of my RAM, and
since the size of the result is going up almost exponentially, the next
one would exceed my RAM and start swapping to virtual memory. As long as
the calculation is in RAM, the time is going up just barely faster than
linearly (1e6! took about 1 minute), but if it starts swapping then I'm
sure the time will go up much faster than linearly.
As might be expected, the resulting number is not really compressible.
Using gzip, I compressed the resulting ASCII file of digits [0-9] to
15 827 771 bytes, a factor of 2.22 compression. Since log2(10)=3.32
bits, we would expect about 8/3.32 = 2.41 compression just by coding the
digits efficiently.
kernel: linux 2.6.9-1-686-smp
language: C++
library: GiNaC http://www.ginac.de/
cat /proc/cpuinfo:
stepping: 9
cpu MHz : 2606.436
cache size : 512 KB
physical id : 0
siblings: 2
fdiv_bug: no
hlt_bug : no
f00f_bug: no
coma_bug: no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe cid xtpr
bogomips: 5160.96
c++ factorial1.cc -o fc1 -lcln -lginac
date; ./fc1 fout5565709; date
Wed Apr 20 06:46:24 EDT 2005
Wed Apr 20 06:54:21 EDT 2005
wc -c fout5565709
35126452 fout5565709
Program:
#include iostream
#include ginac/ginac.h
using namespace std;
using namespace GiNaC;
int main()
{
ex poly;
poly = factorial( 5565709 );
cout poly ;
return 0;
}
--
Erik Reuter http://www.erikreuter.net/
___
http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Erik Reuter ([EMAIL PROTECTED]) wrote: 5565709! has 35 126 456 digits and took 7 minutes 57 seconds to Oops, that's what I get for trying to type instead of copying. As you see below, it is actually: 35126452 digits. -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
By the way, there are a bunch of free tools out there that can be used for this type of problem. I somewhat arbitrarily chose GiNaC because it looked robust (being implemented as a C++ library), but there are many more options that could have done the calculation for free. If you aren't running Linux but want to play around with some of the free mathematical and scientific tool sets out there, a good way to do it is with the Quantian live-CD linux distribution (based on Knoppix). http://dirk.eddelbuettel.com/quantian.html If you don't know what a live CD is : it means you just boot from the CD and you are running the OS from the CD, without having to install the OS on your hard drive. When you are done, just take the CD out and reboot and you will be back running your usual OS on your hard drive (or whatever). -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Julia Thompson ([EMAIL PROTECTED]) wrote: I'm drooling. And of course you *would* have to put this up in the week leading up to the science fiction convention we drop the most money at every year head bang on keyboard That costs a lot more than all the components of my expensive big dream project. (Which is a lot more affordable than I thought, now that I check out pricing on *that*) Don't waste your money! Mathematica is highly polished, but there is free software that can do just about everything Mathematica can. One possibility is Maxima: http://maxima.sourceforge.net/screenshots.shtml There are several other free programs that may be better depending on what you are trying to do, but Maxima is the most general purpose free math system that I know of. -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 01:14 AM Wednesday 4/20/2005, Warren Ockrassa wrote: On Apr 19, 2005, at 11:01 PM, Ronn!Blankenship wrote: At 11:04 PM Tuesday 4/19/2005, Warren Ockrassa wrote: http://sourceforge.net/project/showfiles.php? group_id=5083package_id=9297 I DLed the .jar, the first on the list, 0.8pre2. And when I entered 900! and clicked the Symbolic evaluation button I got a damn big number, 67526802209645841583879061361800814224269427869589384312198268703685091 6431804 [...] , perhaps? Or thereabouts, yeah, though I don't feel too inclined to do a digit-for-digit check... I tried 5,565,709! as well, and the program didn't die, but it set up a wait cursor that tells me nothing at all about how long it would take to do the calculation. After about an hour I shut it down and went on to other things. Java's convenient but its VM can really suck up the silicon, especially when all you have is a 600 MHz G3. :\ As I said, Mathematica running on this 3GHz processor took overnight to come up with 1,000,000! (Admittedly I was doing e-mail and some other things which were more IO-intensive than CPU-intensive while it was running.) You may find yourself in the place I was on that old IBM 1130 that I struggled to get to do 2500! when I was trying something else and figured based on the preliminary runs that the finished program would take on the order of two months to run . . . -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
- Original Message - From: Ronn!Blankenship [EMAIL PROTECTED] To: Killer Bs Discussion brin-l@mccmedia.com Sent: Wednesday, April 20, 2005 12:56 AM Subject: Re: One more! Which one is best depends on what you are trying to do atm. (I have both, tho I haven't gotten around to reinstalling the FORTRAN on this machine yet.) In this particular case, doing it in FORTRAN would be a little more complicated than typing Factorial [100] and pressing Shift+Enter . . . Fair enough. If you have both, and an overnight computer run to spare, your solution is the best. I was thinking more of cost/benefit for someone on a budget who had to pick something to buy. Dan M. ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Dan Minette ([EMAIL PROTECTED]) wrote: Fair enough. If you have both, and an overnight computer run to spare, your solution is the best. I was thinking more of cost/benefit for someone on a budget who had to pick something to buy. I didn't have either on my computer when I got up this morning. I didn't have GiNaC, either. But overnight to calculate 5M digits sounded way too long to me. So I took a quick look for free software that could do it (about 3 min of reading), decided GiNaC looked good, and typed apt-cache search ginac ginac-tools - Some tools for the GiNaC framework libginac-dev - The GiNaC framework (development files) libginac1.3 - The GiNaC framework (runtime library) apt-get install libginac-dev and then I pasted the GiNaC equivalent of hello, world into my text editor, edited a couple lines, and ran it on 1 000 000! to start. In less than a minute I had the answer. Then I ran it on the requested number which took 8 minutes, and posted the results. Total expenditure: $0, and a few minutes of time (it actually took me longer to create the post summarizing the results than it did to install the software and calculate the results) On the other hand if 8 minutes is too quick for you and you'd rather wait several days for Mathematica to calculate 5M!, then that is of course the best solution if it makes you feel better about all the money you spent on Mathematica... -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 10:30 AM Wednesday 4/20/2005, Dan Minette wrote: - Original Message - From: Ronn!Blankenship [EMAIL PROTECTED] To: Killer Bs Discussion brin-l@mccmedia.com Sent: Wednesday, April 20, 2005 12:56 AM Subject: Re: One more! Which one is best depends on what you are trying to do atm. (I have both, tho I haven't gotten around to reinstalling the FORTRAN on this machine yet.) In this particular case, doing it in FORTRAN would be a little more complicated than typing Factorial [100] and pressing Shift+Enter . . . Fair enough. If you have both, and an overnight computer run to spare, your solution is the best. I was thinking more of cost/benefit for someone on a budget who had to pick something to buy. Oh, I agree. And, as Erik pointed out, there are now much cheaper alternatives to Mathematica (although I'm not sure if they work with the various add-on packages written for Mathematica). At the time I got it, alternatives were fewer if any, and I needed the compatibility. Of course, I wish I'd had both Mathematica and the FORTRAN package back in the good old days . . . though I suppose one would have needed at least a high-end IBM 370 or bigger mainframe in those days just to run them . . . and just think of how slow it would have been . . . -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
As a Pythonist who does a reasonable bit of scientific computing sorts of calculations... I'll say that although I don't have time to see just what it would do with this problem, Scipy, a Python module that uses native libraries, seems to perform quite well at such things once it one muddles through the documentation to figure out the right way to attack the problem. This is especially true if one either obtains binaries for one's specific platform, or, even better, compiles optimized libraries on the very machine on which they'll run. System vendors, such as Sun, also offer highly-optimized standard math libraries for their machines. We haven't invested in such yet. Nick ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Nick Arnett ([EMAIL PROTECTED]) wrote: As a Pythonist who does a reasonable bit of scientific computing sorts of calculations... I'll say that although I don't have time to see just what it would do with this problem, I guess not... Scipy, a Python module that uses native libraries, seems to perform quite well at such things once it one muddles through the documentation to figure out the right way to attack the problem. Huh? factorial(N, exact=1) will calculate it using arbitrary precision integer arithmetic. But it is darn slow. I tried factorial(10, exact=1) and it took more than a minute on my machine. Also, 1! was only a couple seconds, so it looks like it is much worse than linear time. I'd hate to see what happens if you try 5M! -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Wed, 20 Apr 2005 16:29:57 -0400, Erik Reuter wrote Scipy, a Python module that uses native libraries, seems to perform quite well at such things once it one muddles through the documentation to figure out the right way to attack the problem. Huh? factorial(N, exact=1) will calculate it using arbitrary precision integer arithmetic. Right... but Scipy has all sorts of ways to do all sorts of things. I'm usually doing somewhat more complex linear algebra stuff, such as calculating eigenvectors and various sorts of decomposition, including our favorite, singular value decomposition. It's is all a bit taxing our systems, especially when I want to do really large matrices (like 50K x 10K). If it goes into virtual memory, there isn't enough time in the universe to finish some of these things. If anyone is wondering, I'm doing this sort of thing in relation to computational linguistics and so forth. Nick ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Nick Arnett ([EMAIL PROTECTED]) wrote: On Wed, 20 Apr 2005 16:29:57 -0400, Erik Reuter wrote Scipy, a Python module that uses native libraries, seems to perform quite well at such things once it one muddles through the documentation to figure out the right way to attack the problem. Huh? factorial(N, exact=1) will calculate it using arbitrary precision integer arithmetic. Right... but Scipy has all sorts of ways to do all sorts of things. I'm Incredible! Your inaction while wishing for a magical solution to drop from heaven extends even to simple programming tasks! ...on the bright side, at least you are consistent... -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 05:34 PM Wednesday 4/20/2005, Nick Arnett wrote: On Wed, 20 Apr 2005 16:29:57 -0400, Erik Reuter wrote Scipy, a Python module that uses native libraries, seems to perform quite well at such things once it one muddles through the documentation to figure out the right way to attack the problem. Huh? factorial(N, exact=1) will calculate it using arbitrary precision integer arithmetic. Right... but Scipy has all sorts of ways to do all sorts of things. I'm usually doing somewhat more complex linear algebra stuff, such as calculating eigenvectors and various sorts of decomposition, including our favorite, singular value decomposition. It's is all a bit taxing our systems, especially when I want to do really large matrices (like 50K x 10K). If it goes into virtual memory, there isn't enough time in the universe to finish some of these things. FWIW, the same old computer I mentioned previously was limited to doing eigenvalues and eigenvectors of a 24x24 matrix: there wasn't enough memory to work on a 25x25 . . . (That was the first program I wrote, btw.) -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Wed, 20 Apr 2005 18:51:09 -0400, Erik Reuter wrote Incredible! Your inaction while wishing for a magical solution to drop from heaven extends even to simple programming tasks! No, no, no. I'm wishing for magical *documentation*. Python isn't a Perl sort of there's more than one way to do it language... but there often is more than one way. Nick ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
* Nick Arnett ([EMAIL PROTECTED]) wrote: On Wed, 20 Apr 2005 18:51:09 -0400, Erik Reuter wrote Incredible! Your inaction while wishing for a magical solution to drop from heaven extends even to simple programming tasks! No, no, no. I'm wishing for magical *documentation*. Yeah, how could you ever in a million years have guessed that the function for calculating a factorial was called factorial? -- Erik Reuter http://www.erikreuter.net/ ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 06:01 PM Wednesday 4/20/2005, Nick Arnett wrote: On Wed, 20 Apr 2005 18:51:09 -0400, Erik Reuter wrote Incredible! Your inaction while wishing for a magical solution to drop from heaven extends even to simple programming tasks! No, no, no. I'm wishing for magical *documentation*. Python isn't a Perl sort of there's more than one way to do it language... but there often is more than one way. There's almost always more way to do _any_ programming task, even hello, world. -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Wed, 20 Apr 2005 19:42:47 -0400, Erik Reuter wrote Yeah, how could you ever in a million years have guessed that the function for calculating a factorial was called factorial? I'm having trouble getting your joke. It was a joke, wasn't it? Python is completely object-oriented, so it has methods rather than functions. Part of the challenge of such languages is to figure out where in the object hierarchy to find classes and methods, of course, and often, the appropriate way to call them. And then there are lovlies like call-backs and such that make it all so interesting. And now that I've looked, golly, it is darn simple -- it's at the top level, unlike lots of other things. Pretty much where one would expect, which is certainly not always the case. But it's not called factorial, it's called scipy.factorial. Unless, of course, you've imported the namespace instead of the module. But surely you knew that, I suspect, since you seem to have implied knowldge of the Scipy interface...? I still don't get the joke, if that's what it was. If it wasn't, then I guess I don't get the point, unless the point was to belittle the fact that I didn't look up the factorial method before posting, in which case... nothing. Nick ___ http://www.mccmedia.com/mailman/listinfo/brin-l
One more!
1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . Enough Already Maru -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Tue, 19 Apr 2005, Ronn!Blankenship wrote: 1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . OK, then, how many digits does 5,565,709! have? (1,391,423 are trailing zeroes, I can tell you that much) Julia ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 07:00 PM Tuesday 4/19/2005, Julia Thompson wrote: On Tue, 19 Apr 2005, Ronn!Blankenship wrote: 1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . OK, then, how many digits does 5,565,709! have? (1,391,423 are trailing zeroes, I can tell you that much) Since 1,000,000! ran overnight (and had the CPU going at about 50% capacity all that time), I think exploration of anything larger will have to wait until the next time I go out of town, unless someone wants to offer us time on a supercomputer or you want to set up the Brin-L distributed computing network . . . -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
Ronn!Blankenship wrote: At 07:00 PM Tuesday 4/19/2005, Julia Thompson wrote: On Tue, 19 Apr 2005, Ronn!Blankenship wrote: 1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . OK, then, how many digits does 5,565,709! have? (1,391,423 are trailing zeroes, I can tell you that much) Since 1,000,000! ran overnight (and had the CPU going at about 50% capacity all that time), I think exploration of anything larger will have to wait until the next time I go out of town, unless someone wants to offer us time on a supercomputer or you want to set up the Brin-L distributed computing network . . . If you can get me a program to run it, I could do it here Not sure how long it would take, but it would get done, anyway. Julia ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Apr 19, 2005, at 7:23 PM, Julia Thompson wrote: If you can get me a program to run it, I could do it here Not sure how long it would take, but it would get done, anyway. If you feel like hacking and compiling your own Java applet, give this a whirl, though I don't know how well it'll handle any really long numbers: http://www.unix.org.ua/orelly/java-ent/jnut/ch01_03.htm -- Warren Ockrassa, Publisher/Editor, nightwares Books http://books.nightwares.com/ Current work in progress The Seven-Year Mirror http://www.nightwares.com/books/ockrassa/Flat_Out.pdf ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Apr 19, 2005, at 8:07 PM, I wrote: On Apr 19, 2005, at 7:23 PM, Julia Thompson wrote: If you can get me a program to run it, I could do it here Not sure how long it would take, but it would get done, anyway. If you feel like hacking and compiling your own Java applet, give this a whirl, though I don't know how well it'll handle any really long numbers: http://www.unix.org.ua/orelly/java-ent/jnut/ch01_03.htm Huh. Just tried it; it returns 'Infinity' for 5,565,709!. Obviously a factorial on a large number needs a kind of juggling that a quasi-64-bit OS running Java can't hack out of the box... -- Warren Ockrassa, Publisher/Editor, nightwares Books http://books.nightwares.com/ Current work in progress The Seven-Year Mirror http://www.nightwares.com/books/ockrassa/Flat_Out.pdf ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: Ronn!Blankenship wrote: At 07:00 PM Tuesday 4/19/2005, Julia Thompson wrote: On Tue, 19 Apr 2005, Ronn!Blankenship wrote: 1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . OK, then, how many digits does 5,565,709! have? (1,391,423 are trailing zeroes, I can tell you that much) Since 1,000,000! ran overnight (and had the CPU going at about 50% capacity all that time), I think exploration of anything larger will have to wait until the next time I go out of town, unless someone wants to offer us time on a supercomputer or you want to set up the Brin-L distributed computing network . . . If you can get me a program to run it, http://store.wolfram.com/catalog/ I could do it here Not sure how long it would take, but it would get done, anyway. Less time than it will be before I can afford to upgrade to the latest version, no doubt . . . -- Ronn! :) ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
Ronn!Blankenship wrote: At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: Ronn!Blankenship wrote: At 07:00 PM Tuesday 4/19/2005, Julia Thompson wrote: On Tue, 19 Apr 2005, Ronn!Blankenship wrote: 1,000,000! has 5,565,709 digits, 249,988 of which are trailing zeroes . . . OK, then, how many digits does 5,565,709! have? (1,391,423 are trailing zeroes, I can tell you that much) Since 1,000,000! ran overnight (and had the CPU going at about 50% capacity all that time), I think exploration of anything larger will have to wait until the next time I go out of town, unless someone wants to offer us time on a supercomputer or you want to set up the Brin-L distributed computing network . . . If you can get me a program to run it, http://store.wolfram.com/catalog/ I'm drooling. And of course you *would* have to put this up in the week leading up to the science fiction convention we drop the most money at every year head bang on keyboard That costs a lot more than all the components of my expensive big dream project. (Which is a lot more affordable than I thought, now that I check out pricing on *that*) I could do it here Not sure how long it would take, but it would get done, anyway. Less time than it will be before I can afford to upgrade to the latest version, no doubt . . . Probably. :) Julia ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
On Apr 19, 2005, at 8:34 PM, Ronn!Blankenship wrote: At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: If you can get me a program to run it, http://store.wolfram.com/catalog/ US$900! (Won't bother trying to calculate that.) Ouch! I hadn't realized Mathematica was that significantly priced. Humm. Humm. Googlicious ... OK, here's an interesting package called HartMath, which appears to do symbolic calculations, is Java based (!) and is free: http://sourceforge.net/project/showfiles.php? group_id=5083package_id=9297 I DLed the .jar, the first on the list, 0.8pre2. And when I entered 900! and clicked the Symbolic evaluation button I got a damn big number, which was much more satisfying than the infinity symbol I got when I clicked Numeric evaluation. This might show more promise for calculating 5,565,709! than a basic Java hack. -- Warren Ockrassa, Publisher/Editor, nightwares Books http://books.nightwares.com/ Current work in progress The Seven-Year Mirror http://www.nightwares.com/books/ockrassa/Flat_Out.pdf ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more!
- Original Message - From: Warren Ockrassa [EMAIL PROTECTED] To: Killer Bs Discussion brin-l@mccmedia.com Sent: Tuesday, April 19, 2005 11:04 PM Subject: Re: One more! On Apr 19, 2005, at 8:34 PM, Ronn!Blankenship wrote: At 09:23 PM Tuesday 4/19/2005, Julia Thompson wrote: If you can get me a program to run it, http://store.wolfram.com/catalog/ US$900! (Won't bother trying to calculate that.) Ouch! I hadn't realized Mathematica was that significantly priced. Humm. Humm. Googlicious ... OK, here's an interesting package called HartMath, which appears to do symbolic calculations, is Java based (!) and is free: What was Microsoft FORTRAN and then became Compaq FORTRAN and may be HP FORTRAN now has IMSL installed. It is a wondrous callable library of mathematical functionsand I got it with FORTRAN for a few hundred dollars for the package. If you want to throw fits at numbers, that's the way to goas well as do a zillion other things. Mathamatica is a toy in comparison. Dan M. ___ http://www.mccmedia.com/mailman/listinfo/brin-l
One more Hoon tee-shirt
Just in case Steve checks his email before heading out to the con Twaphu-anuph in shades and an earring, holding a Pan-galactic Gargle Blaster, wearing ragged blue jeans cutoff at the upper knees and his own tee-shirt proudly proclaiming: I'm not Herbert. Vilyehm Teighlore ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit of ecconomic data
On 20 Jan 2004, at 12:31 am, Erik Reuter wrote: On Sun, Jan 18, 2004 at 09:14:41PM -0600, Dan Minette wrote: OK, first analysis of income by 20% grouping and top 5%. RR + GB^2Clinton 1st 20% 7.6% 15.9% 2nd 20% 8.9% 15.5% 3rd 20% 11.2% 14.6% 4th 20% 14.0% 15.8% 5th 20% 24.8% 28.8% top 5% 40.7% 43.4% The numbers don't exactly match with the GDP numbers for a couple of reasons. There was approximately 2%-3% greater growth in the numbers of households under RR + GB^2 than under Clinton. The share of the GDP growth that went to household income was greater under Clinton. That is fairly strong evidence in support of your contention that Democrats are better for the poor than Republicans. Under Clinton, the bottom 40% had approximately DOUBLE the rate of growth as under the best 8 years of RR + GB^2, as you said. I wonder why JDG hasn't commented. One wouldn't want to let contingent facts get in the way of revealed truth :) -- William T Goodall Mail : [EMAIL PROTECTED] Web : http://www.wtgab.demon.co.uk Blog : http://radio.weblogs.com/0111221/ It is our belief, however, that serious professional users will run out of things they can do with UNIX. - Ken Olsen, President of DEC, 1984. ___ http://www.mccmedia.com/mailman/listinfo/brin-l
Re: One more bit of ecconomic data
OK, first analysis of income by 20% grouping and top 5%. RR + GB^2Clinton 1st 20% 7.6% 15.9% 2nd 20% 8.9% 15.5% 3rd 20% 11.2% 14.6% 4th 20% 14.0% 15.8% 5th 20% 24.8% 28.8% top 5% 40.7% 43.4% The numbers don't exactly match with the GDP numbers for a couple of reasons. There was approximately 2%-3% greater growth in the numbers of households under RR + GB^2 than under Clinton. The share of the GDP growth that went to household income was greater under Clinton. But, we see that, while richer households did approximately as well under 8 cherry picked years from the last 14 under Bush^2 RR as under Clinton, they bottom 40% only did half as well, and the middle 20% did about 20% worse. Dan M. ___ http://www.mccmedia.com/mailman/listinfo/brin-l
