Tilted Forum Project Discussion Community  

Go Back   Tilted Forum Project Discussion Community > The Academy > Tilted Knowledge and How-To


 
 
LinkBack Thread Tools
Old 10-25-2004, 01:09 PM   #1 (permalink)
 
KnifeMissile's Avatar
 
Location: Waterloo, Ontario
Queer properties of infinite sums...

Actually, this is a precursor to a post that I really want to make and, actually, I've wanted to make this post for some time now but haven't gotten around to it and now it looks like the best days of this forum are gone. I don't recognize all the interesting characters who could appreciate what I'd like to explain. Oh well, maybe they're still lurking. Here goes...

...the precursor, that is. Are we all aware of what an infinite sum is?

The harmonic series is a series of the form: 1 + 1/2 +1/3 + ... + 1/n. Surprisingly, even though the terms limit to zero, this series doesn't converge. It will sum indefinitely high. However, if you alternate the terms, like so: 1 - 1/2 + 1/3 - ... + (-1)^n/-n (order of operations apply), then this does converge! In fact, it can be shown that any series, where the term sign alternates and the absolute values of the terms monotonically decrease, will converge to a finite value. Can everyone see why this is true? Can a stronger statement be made?

The claim I'd like to make will probably leave most physicists incredulous. I've tried to explain it to some before but have only been met with skepticism. Hopefully, many math enthusiasts will undestand me. For now, I just want to know what the board is currently like...
KnifeMissile is offline  
Old 10-25-2004, 01:37 PM   #2 (permalink)
Junkie
 
filtherton's Avatar
 
Location: In the land of ice and snow.
We are studying infinite sums right now in my calc class. They seem like a nice change of pace from the standard integration problems we had been doing. I can't yet see why your alternating and decreasing monotonic sequence converges when summed, but give me a week more of class and i will probably be able to.

Let's see:

1 - 1/2 = 1/2 = .5
1 - 1/2 + 1/3 = 5/6 = 0.83333
1 - 1/2 + 1/3 - 1/4 = 7/12 = 0.58333
1 - 1/2 + 1/3 - 1/4 + 1/5 = 47/60 = 0.78333
1 - 1/2 + 1/3 - 1/4 + 1/5 - 1/6 = 37/60 = 0.61666

I don't have a general explanation as to why it converges, but based on its initial behavior it seems to be approaching a specific number.
filtherton is offline  
Old 10-25-2004, 02:43 PM   #3 (permalink)
Tilted
 
Location: Sydney, Australia
A stanger fact is that there are infinite sums such that if you change the order of summation then the value they converge to changes. I'm also doing infinite sums right now in my calc class and should probably understand how to set one up- but don't.

I'll hunt around for an example later.
molloby is offline  
Old 10-25-2004, 04:35 PM   #4 (permalink)
Addict
 
Enough with the suspense! Let's have your claim! =) Though I'm not a physicist, I'll still enjoy reading it.
phukraut is offline  
Old 10-25-2004, 04:48 PM   #5 (permalink)
Addict
 
An example of molloby's idea appears on page 77 of Goldberg's Methods of Real Analysis (second edition):

We have that

(1) \sum_{n=1}^\infty (-1)^{n+1}/n = L = \log 2. (where L is the sum of the series).

Now, divide L by 2, giving

L/2 = 1/2 - 1/4 + 1/6 - 1/8 + ..., and thus

(2) L/2 = 0 + 1/2 - 0 - 1/4 + 0 + 1/6 - 0 - 1/8 + ... .

Adding (2) and (1) gives

(3) 3/2 L = 1 + 1/3 - 1/2 + 1/5 + 1/7 - 1/4 + ... .

You will notice that (3) is just a rearrangement of (1), yet it sums to 3/2 L = 3/2 \log 2.

The book goes on to give the following theorem:

Let L be a conditionally convergent series of real numbers. Then for any real number x, there is a rearrangement of L which converges to x. However, if the series is absolutely convergent, then all rearrangements sum to the same value.
phukraut is offline  
Old 10-25-2004, 05:28 PM   #6 (permalink)
Junkie
 
It's been a while since I've worked with infinite sums, but I'd be interested in this discussion.
__________________
"Fuck these chains
No goddamn slave
I will be different"
~ Machine Head
spectre is offline  
Old 10-25-2004, 06:09 PM   #7 (permalink)
 
KnifeMissile's Avatar
 
Location: Waterloo, Ontario
Quote:
Originally Posted by phukraut
Enough with the suspense! Let's have your claim! =) Though I'm not a physicist, I'll still enjoy reading it.
Quote:
Originally Posted by phukraut
Enough with the suspense! Let's have your claim! =) Though I'm not a physicist, I'll still enjoy reading it.
Oh, whatever! You all seem to already know what a conditionally convergent series is, anyways. Yes, that was what I was trying to build up to. No physicist I have ever talked to has taken my word that such a thing exists which is why I thought you'd all find it fascinating. They all believe that addition is necessarily commutative. Oh well, I can still explain how such a thing exists. It won't be a formal proof but it won't need to be. As long as I can describe the idea in detail, you can all formlize it for yourselves (and you all sound like you can!).

So, if you are a conditionally convergent series, then the sum of your terms of like sign must diverge (in other words, the sum of all the negative terms must diverge negatively, while all positive terms diverge positively). Otherwise, you must be absolutely convergent. I mean, if only your negative terms diverge, how can you be convergent? Vice versa for your positive terms, they must diverge as well. If both signs of terms converge then how are you not absolutly convergent? So, now we have established that the sum of terms of like signs must diverge.
Now, pick a real number x, any x! (even x factorial!) If x > 0, then you can sum just enough positive terms to so that that sum will be greater than x. Start adding just enough negative terms so that the sum is less than x. You know you need, at least, one negative term to do this. Now, do the same thing you were doing with the positive terms, before. That is, add just enough terms to exceed x. Again, you know you need at least one term to do this. This is very important. Again, sum the negative terms the same way you were, before. Repeat this algorithm so that the sum will waffle between being greater than and less than the target number, x.
Because we know that the absolute value of the terms in the series limits to 0 (or would people like me to prove that?), we know that this order of terms will sum arbitrarily close to the target number. Furthermore, we know that everytime you switch signs you will use, at least, one term of that sign and that you must, repeatedly and endlessly switch signs. Thus, we know that we will use all the terms from both signs and that will make this a true reordering of our series.
Without loss of generality, you can see that this works regardless of what value x is. Thus, a conditionally convergent series can be reordered to sum to any value.
QED

I hope you enjoyed it!
KnifeMissile is offline  
Old 10-25-2004, 06:54 PM   #8 (permalink)
I am Winter Born
 
Pragma's Avatar
 
Location: Alexandria, VA
KnifeMissile, it's things like this that make me cry myself to sleep at night - even though I completely understand the reasoning (and application) of the theory.
Pragma is offline  
Old 11-07-2004, 03:11 AM   #9 (permalink)
Crazy
 
Location: San Diego, CA
KnifeMissile - I'm a little bit confused. You'll have to fogive me, as my infinite sum theory is a bit rusty.

First of all, what do you mean by "just over x." If it's a distinct number from x, then there are an infinite number of real numbers between it and x, so really any number greater than x would do. If it's not a distinct number from x, then it is x, and so you're just adding and subtracting 0 over and over, in which case the only number you can form is x by rearranging.

Also, I do recall that there are infinite sums that do not converge to a number, but also do not diverge to infinity. Think about this one: 2 - 2 + 2 - 2 + 2 - 2 + ... At any one point it's either 2 or 0, but I certaintly wouldn't say it converges to 1.

Could you try and clarify your proof a bit? Maybe write out some equations and examples?

I'm also a bit curious about WHAT you're trying to prove. I had to read to the end just to find any kind of conclusion... if you clearly stated what you were trying to prove at the beginning, it would make reading it much easier.
__________________
"Don't believe everything you read on the internet. Except this. Well, including this, I suppose." -- Douglas Adams
Rangsk is offline  
Old 11-07-2004, 08:22 AM   #10 (permalink)
Insane
 
Location: Ithaca, New York
This is the damndest thing I've ever seen.
fckm is offline  
Old 11-07-2004, 08:25 AM   #11 (permalink)
Rookie
 
cliche's Avatar
 
Location: Oxford, UK
Quote:
Originally Posted by phukraut
An example of molloby's idea appears on page 77 of Goldberg's Methods of Real Analysis (second edition):
[snip]
The book goes on to give the following theorem:

Let L be a conditionally convergent series of real numbers. Then for any real number x, there is a rearrangement of L which converges to x. However, if the series is absolutely convergent, then all rearrangements sum to the same value.
Quote:
Originally Posted by knifemissile
Thus, a conditionally convergent series can be reordered to sum to any value.
Looks like you've been beaten to it!
It's an interesting concept though - but I guess intuitions can be appeased with the idea that you're essentially asking for the difference between two infinite (divergent) series. Their values being undefined it seems reasonable that depending on exactly how you ask the question you can get any answer you like.
__________________
I can't understand why people are frightened of new ideas. I'm frightened of the old ones. -- John Cage (1912 - 1992)
cliche is offline  
 

Tags
infinite, properties, queer, sums

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT -8. The time now is 11:07 AM.

Tilted Forum Project

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0 PL2
© 2002-2012 Tilted Forum Project

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360