Survey posted 2005-05-16 19:04 by Jon.

Contribute ideas to surveys by sending a mail to survey@ticalc.org.

I think the next survey should be, "Who DIDN'T use a calculator to figure out the answer to the last poll?"

Not a bad idea...I'll consider it, though I'll probably try to take the surveys down a new path.

I was kind of kidding, but whatever.

Guess who's in the June 2005 issue of Popular Science... ME! Woooooooooooooooooooooooooooot!

Guess who just calculated Pi to the 1,024,000,000th decimal place... ME! Woooooooooooooooooooooooooooot!

*starts to go to 2,048,000,000th place*

Yes!!!

How is it possible to calculate pi out to that far a degree?? You obviously used some sort of computer/calculator but what method is used?

I used a program called Pifast [available at http:// numbers. computation. free. fr/ Constants /PiProgram /pifast .zip ...with no spaces of course] It uses one of 2 methods: The Chudnovsky method, and the Ramanujan method. It doesn't give a formula for either of the 2, but with some googling, you could probably find an email address to find them out.

*added plus*: This program can also compute e to any number of decimal places using the exponential series exp(1) or 1/exp(-1) [I have no idea what this means, but it is displayed in the program.] The website for reference is: http:// numbers. computation. free. fr/ Constants/ PiProgram/ pifast.html [again, with no spaces]

:-D I have 2,048,000,000 now!!! WooHOOO!!

To represent that many digits in ASCII would require about 2 gigabytes!

It takes up 266 MB on my computer..

It takes up 266 MB on my computer...

Grr.. stupid double post thingie.. it didnt show up for 30 minutes! then it pops up when I came to the conclusion that it had not posted the first time...

Well, it's not in ASCII inside the program. It's much larger only if you dump the answer to a text file.

Try that right now!

That is the answer in a text file.. [C:\Documents and Settings\Owner\My Documents\Calculator Games & Stuff\pifast\ pi.txt to be exact :-P..] It has pi to 2,048,000,000 decimal places and takes up 266 MB.

The program itself only takes up 328 KB.

P.S: If you go to My Computer folder and right-click on C:\ and go to Properties, you can click to compress the files in it.. I have done that so everything is compressed to save space.. That could be why the output is not as huge as it should be..

Yeah, it's probably the compression at work. The answer is made up of just the digits 0-9 and a period (and possibly newlines), so it should compress that much.

Hmm...I had a 48MB file (a bitmap of all the 24-bit RGB colors) compress to 2.13MB using 7Zip twice, but Windows XP's built-in compression did not reduce it at all. It is obviously optimized for speed, & considering how long 7Zip took, for good reason.

Ha! I found out what was makin it so small! I missed a "0" when typing it in, so it only went to the 204,800,000th place.. now I guess Imma have to go over again.. take 10x as long.. [11.7 hours is my estimate.. ouch!]

Did you make that in Paint? :P

You probably wrote a program to produce it, didn't you?

I will just say...Visual Basic is my friend. :)

I like to write C programs to generate a bunch of different images and even sounds from time to time. I usually write to a more standard and portable image format like Portable GrayMap or Portable PixMap, depending on if the image has colour or just greys. The primary reason I use PGM or PPM is their simplicity; for PPM it's just the following in ASCII:

P6
<width> <height>
<maxcolor>

followed by the raw image data. It's much simpler than dealing with MS bitmaps in a plain and simple C program, and I can convert it to many other formats using the Netpbm filters (like ppmtobmp or ppmtopict). They're pretty nice.

Have you tried PNGing that file? If anything is gonna compress generated image data, it's PNG.

Wow! I just tried it & it came out to 126KB. I did not expect it to beat 7Zip by that much...I guess the PNG format must compress each color plane separately. That just shows that certain data with obviously low information content does not compress well by generic methods.

I made an "all colours" image just now too, and I compressed it to a PNG. It's 122K and took 13.078 seconds real time or 9.270 seconds user time. I used compression level 9.

What are the dimensions of your image, and how are the colours arranged? That might make a difference in the effectiveness of compression.

Mine is 256 x 65536 pixels, and I just looped over each R, G, and B (in that order). B increments each pixel, G increments every time B rolls over, and R increments every time G rolls over.

In fact, since it's pretty short, here's the whole program (I could easily make it shorter):

#include <stdio.h>

int main()
{
int r, g, b;

printf("P6\n");
printf("256 65536\n");
printf("255\n");

for (r = 0; r < 256; ++r)
for (g = 0; g < 256; ++g)
for (b = 0; b < 256; ++b)
printf("%c%c%c", (char)r, (char)g, (char)b);

return 0;
}

Yes, indeed, the colour layout makes a difference.

I made the image 4096 x 4096, but left the order of the colours the same. The PNG is 128K in this case.

Mine is also 4096x4096, with green, then low nibble of red horizontally & blue, then high nibble of red going vertically. It is available on the main page of my website (see the link) as a PNG further compressed with 7-Zip down to 646 bytes. I would like to see someone beat that.

Well, my program is only 234 bytes, and it's enough to entirely reproduce the image. :)

You might get it even smaller if you run it through PNG Gauntlet (see link). Be warned though, that's probably gonna take even longer than your 7Zip approach.
Well, PNG is designed specifically for image files, so it's not all that weird it beats a generic compressor. (Note though that PNG doesn't have a clue how to compress photographic images. Then again, I know of no lossless compression that works well on photos.)

PNGGauntlet cut the file size down to about 56KB, but then 7-Zip could only reduce the result to about 2KB, which is significantly more than 646 bytes, so if one is only trying to store the file & not view it while stored, the other method is better.

Guess what me Iq is 197 and your is what

Guess what: you still sound like a [censored] when you write with no attention to grammar and convention. Besides, the internet IQ test where you probably generated that value most likely has a significant amount of error (like 99% of those internet tests).

Thank you. I was waiting for someone to say that :).

Yeah, I remember taking one (but not taking it seriously) because a friend told me too, and even the questions were incredibly stupid - nothing you would probably find on a REAL IQ test.

Ok i lied and i used bad grammer on purpose my Iq is only 137

Then did you also incorrectly spell "grammar" as well?

Honestly, I don't care how smart an IQ test says you are. I've been tested in real life and on the internet both and my scores have ranged from below 100 to over 150. But I really don't think IQ is an accurate way of measuring your intellect.

There isn't *any* good way of measuring your intellect.

Sure there is...hang the person over a volcano inside a wooden cage (high enough that it does not burn) held up by some Kevlar & give them only a pocketknife, a flashlight, & a pack of chewing gum (in addition to whatever they carry in their pockets normally), & see how long it takes them to fall or escape. :)

OK, there's no good way of representing intellect *with a number*. :-)

Which article?

It is you! I was reading the letters yesterday and I came across your name which looked familiar...that was a good argument, and I agreed with the statement about the difference between videogamers and real racers, but I didn't think the article was an (almost) complete advertisement for the Forza, but more that it kind of leaned toward it a little...

Now we all know where you live!!!! (not that that's going to do us any good...)

That wouldn't have been too hard to figure out... my website used to have a link to my high school website, which has a complete address.

You know, I read that letter the other day... That is a good point, but I liked that article. I didn't get it as being completely and ad, but it was a bit more than objective...

If you wouldn't have posted that I would never have even realized it, I rarely look at the names on those letters.

So anyway, how many of you guys get PopSci then? Quite a few it seems like. :D

Me! (Sort of intermittently, though, since I get a neighbor's copy whenever he's finished)

Me! (I'm a third-year subscriber, what a great magazine)

I used to get it, but I lost interest. I couldn't get past all of the advertisements for expensive things that make you look geeky (as opposed to actually making you geeky), nor could I get past the focus of the magazine--theoretical ideas that would take too much money to implement and have little practical application.

Wow, you really are! Nice letter, btw. But you should have mentioned something about how photorealistic calc games are starting to get...

Well, that list would start with the people who didn't vote for 0...

Actually, it would start with the people who answer the surveys first, which would be those who are always checking the site (or happen to check at the right time). I would wager that the first responders are unlikely to have used a calculator, because they would probably be people who like mathematics, & would therefore have heard the answer. At least, I know I would be in that group. Then again, I might not be a representative sample.

Wow...that is a rather large value for Pi. :)

Although I know you're kidding, the exclamation point is not a factoral symbol. The phrase is from the Simpons...Pro. Frinck yells this to get a group of scientists to shut up.

Oh...I do not watch the Simpsons.

I didn't!