## Search found 952 matches

- Fri Jan 25, 2013 2:38 pm UTC
- Forum: Mathematics
- Topic: Stats measure to describe difference between two data sets?
- Replies:
**4** - Views:
**1233**

### Re: Stats measure to describe difference between two data se

If you look at the fourier transformation, the first one will have amplitudes which drop (with increasing frequency) quicker than those of the second graph, if you rescale the x-axis to have the same range for both (which is another way to say "ignore the current labels, just look at the graph&...

- Fri Jan 25, 2013 2:34 pm UTC
- Forum: Mathematics
- Topic: function powers for identity function
- Replies:
**11** - Views:
**3232**

### Re: function powers for identity function

Have you tried f(x)=(ax+b)/(cx+d)? Looking at the existing solutions for f^2(x)=x and f^3(x)=x, this would be my first approach. If you don't require that the function is continuous, it is easier, of course. Just group the real numbers in groups of n, and let the function be a rotation in some arbit...

- Fri Jan 25, 2013 2:29 pm UTC
- Forum: What If?
- Topic: What-If 0029: "Spent Fuel Pool"
- Replies:
**167** - Views:
**47263**

### Re: What-If 0029: Spent Fuel Pool

Well, but nuclear power plants are kind of dangerous. Wouldn't it make sense to put our nuclear power plants in orbit and then just beam down the energy? That way if anything goes wrong it will be harmlessly out in space where it doesn't matter, and not down here with us. You do not have access to ...

- Fri Jan 25, 2013 1:42 pm UTC
- Forum: Fictional Science
- Topic: If you could change the laws of physics...
- Replies:
**50** - Views:
**32722**

### Re: If you could change the laws of physics...

It would make everybody dead. So nobody would be sad anymore? Oh, the fun of ethics. Do we even know if room temp superconductors are impossible? Could be possible. If not, make it possible! I'd lower the speed of light to 200mph. Problably bad for the internet. Signals would need hours to days to ...

- Thu Jan 03, 2013 6:18 pm UTC
- Forum: Mathematics
- Topic: Pokemon probability: Catching Feebas
- Replies:
**8** - Views:
**3978**

### Re: Pokemon probability: Catching Feebas

If they are in exactly the same proportion to each other as they are in the other tiles, you can get no information from getting things you don't want. But if they are in some different proportion, you can get more information to organize your search in the long run. While it might be possible to i...

- Thu Jan 03, 2013 5:52 pm UTC
- Forum: Logic Puzzles
- Topic: New game : Intar
- Replies:
**10** - Views:
**4248**

### Re: New game : Intar

The scoring system is tricky - choice of the first 3 pentominos is relevant for the possible size. And I don't understand how many points challenger / initiator get in the two cases (initiator can/cannot improve size). Here is a blow-up version of your example, by the way, with 37 internal tiles. 37...

- Tue Dec 25, 2012 11:21 pm UTC
- Forum: Mathematics
- Topic: Fun (?) office-related probability question
- Replies:
**10** - Views:
**2210**

### Re: Fun (?) office-related probability question

“is more likely given the following assumptions than with other assumptions”.

- Sun Dec 23, 2012 5:47 pm UTC
- Forum: Mathematics
- Topic: Fun (?) office-related probability question
- Replies:
**10** - Views:
**2210**

### Re: Fun (?) office-related probability question

If all floors have much more than 1 person working there, you can neglect the actual numbers and work with relative numbers only. "3 others in the elevator going to 3 different floors" has the highest probability if all 3 floors have the same number of persons working there and noone works...

- Fri Dec 21, 2012 4:56 pm UTC
- Forum: Logic Puzzles
- Topic: Create a perfect lossless compression algorithm
- Replies:
**21** - Views:
**5299**

### Re: Create a perfect lossless compression algorithm

Really, I need to pull out calculus and model this to see where the optimal point is. Did that. An ideal encoding can be described by the probabilities of e (end of message), n (no bit), 0 and 1. The number of bits (data) per bit (sent) is simply the entropy of the m...

- Tue Dec 18, 2012 10:29 pm UTC
- Forum: Logic Puzzles
- Topic: Create a perfect lossless compression algorithm
- Replies:
**21** - Views:
**5299**

### Re: Create a perfect lossless compression algorithm

Based on Yakk's last post: If we can send "end of message" for free, it is possible to improve the bit/cost-ratio with more "end of message" inside: Use 3 bits to decode a number n between 0 and 7, send n times "end of message", send the 4th bit afterwards (without ...

- Mon Dec 10, 2012 7:45 pm UTC
- Forum: Logic Puzzles
- Topic: What isn't the answer to this puzzle?
- Replies:
**13** - Views:
**3667**

### Re: What isn't the answer to this puzzle?

**Spoiler:**

I like that:

Qaanol wrote:Spoiler:

- Mon Dec 10, 2012 7:40 pm UTC
- Forum: Logic Puzzles
- Topic: New abstract game : Invasion
- Replies:
**7** - Views:
**3263**

### Re: New abstract game : Invasion

Player 1 will win, if he can "forget" moves - occupy a position and do not care about it afterwards. This is not possible in the game here (player 1 could occupy a tile he needs to fill later), so the simple symmetry argument does not work. Even worse: If you remove the central tile, playe...

- Mon Dec 10, 2012 3:03 pm UTC
- Forum: Mathematics
- Topic: Happy Numbers
- Replies:
**4** - Views:
**2459**

### Re: Happy Numbers

Very nice. Now if we pretend that a number is happy at random with some probability p, mostly independent of the happiness of its neighbors As you already noted, this is just an approximation, and p itself is fluctuating. p for N-digit numbers influences p' for numbers with \frac{10^N}{285} to \frac...

- Mon Nov 26, 2012 8:58 pm UTC
- Forum: Mathematics
- Topic: Happy Numbers
- Replies:
**4** - Views:
**2459**

### Re: Happy Numbers

Edit: Oh, my mistake. You have to check all (<= 81*N) H(N,X) individually afterwards to get the number of happy numbers. I see that your formula was added at OEIS (with the assumption that Bryan Wolf = BryanWolf :p). Did you check the situation for different bases? 2 and 4 have a fraction of 100%, b...

- Mon Nov 26, 2012 8:47 pm UTC
- Forum: Science
- Topic: Nuclear Physics in the US- Dead?
- Replies:
**15** - Views:
**4689**

### Re: Nuclear Physics in the US- Dead?

SU3SU2U1 wrote:Instead of the starving artist period, or the touring-Europe period of your young adult life, you'll have the (longer) doing-physics period.

You can combine the Europe tour and a PhD in nuclear physics .

FAIR, XFEL, ELI, CERN, PSI, ...

- Mon Nov 26, 2012 6:12 pm UTC
- Forum: Science
- Topic: Immortality
- Replies:
**87** - Views:
**12207**

### Re: Immortality

Computational power really depends on the required level of simulation. Recently, IBM simulated a neuronal network where the number of neurons and synapses was comparable to a human brain (see this article , for example) - just a factor 1500 slower than real time. Using the development of supercompu...

- Sun Nov 25, 2012 11:56 pm UTC
- Forum: Individual XKCD Comic Threads
- Topic: 1138: "Heatmap"
- Replies:
**59** - Views:
**22053**

### Re: 1138: "Heatmap"

There are also a lot of global versions of this map showing traffic to English-language websites which are indistinguishable from maps of the location of internet users who are native English speakers.

Something we should change!

- Sat Nov 24, 2012 9:49 pm UTC
- Forum: Mathematics
- Topic: XKCD #10 was right?!?
- Replies:
**5** - Views:
**1809**

### Re: XKCD #10 was right?!?

You could fix this by using bigger numbers.

I am surprised that the denominator is not a factor of 36^n-1 (similar to arbitrary expansions in base 10: 532436/999999=0.532436532436...).

I am surprised that the denominator is not a factor of 36^n-1 (similar to arbitrary expansions in base 10: 532436/999999=0.532436532436...).

- Wed Nov 21, 2012 4:51 pm UTC
- Forum: Logic Puzzles
- Topic: Blocking game
- Replies:
**8** - Views:
**3304**

### Re: Blocking game

Who is the winner? Player 1 or player 2. You cannot get blocked on the board with 3 pawns, of course, so you try to escape on one board and try to block on the other one. Blocking itself is just a matter of time, therefore you can analyse both boards individually: I would expect that the advantage ...

- Wed Nov 21, 2012 4:34 pm UTC
- Forum: Mathematics
- Topic: Matrices
- Replies:
**7** - Views:
**1424**

### Re: Matrices

Your matrices follow a specific pattern, and this pattern generates linear dependence in some way. Did you check the coefficients to write (0,0,0,0) as sum of 4 row-vectors or (0,0,0,0)^T as sum of 4 column-vectors? This might give some insight in the mechanism.

- Wed Nov 21, 2012 4:24 pm UTC
- Forum: Science
- Topic: Effects on Quantum Physics Interpretation (theory)
- Replies:
**9** - Views:
**1521**

### Re: Effects on Quantum Physics Interpretation (theory)

ArXiv version of the paper We have shown that if the non-local effects that we observe in Bell experiments were due to hidden influences propagating at any finite speed, then non-locality could be exploited for superluminal communication. While it is nice to have a proof, I don't think models with...

- Thu Nov 01, 2012 10:50 pm UTC
- Forum: Gaming
- Topic: Diablo III
- Replies:
**2188** - Views:
**397844**

### Re: Diablo III

The difficulty with no MP is slightly nerfed and the maximum affixes will depend on the monster level, so lvl 61 in Act 1 and 62 in act 2, and even the lvl 63 items won't be able to have the best affixes. They depend on monster level AND item level (more precise, on max(mlvl,ilvl)). You can find th...

- Mon Oct 29, 2012 4:35 pm UTC
- Forum: Gaming
- Topic: Diablo III
- Replies:
**2188** - Views:
**397844**

### Re: Diablo III

With MP0 and in act 1 and 2, most rares will get affixes up to 61 / 62 only, respectively. Just the ilvl63-items will be able to spawn the best affixes.

With MP1+ or in act 3, all monsters are level 63 and all dropped items can get all affixes.

With MP1+ or in act 3, all monsters are level 63 and all dropped items can get all affixes.

- Thu Oct 25, 2012 5:35 pm UTC
- Forum: Mathematics
- Topic: probability of random walk bot getting stuck
- Replies:
**25** - Views:
**3363**

### Re: probability of random walk bot getting stuck

invalid syntax: print N, status, counter invalid syntax: print k, record invalid syntax: print outcome Traceback (most recent call last): File "walk.py", line 1, in <module> from sets import Set ImportError: No module named 'sets' Traceback (most recent call last): ...

- Thu Oct 25, 2012 4:19 pm UTC
- Forum: Individual XKCD Comic Threads
- Topic: 1125: "Objects In Mirror"
- Replies:
**70** - Views:
**25670**

### Re: 1125: "Objects In Mirror"

Mirrors in object are closer than they appear.

- Thu Oct 25, 2012 12:09 pm UTC
- Forum: Mathematics
- Topic: probability of random walk bot getting stuck
- Replies:
**25** - Views:
**3363**

### Re: probability of random walk bot getting stuck

List of x's and approx. probability that cube is stuck by the time x steps have passed: Under the assumption that "gets stuck" is a purely local issue, the ratio P(not stuck after x)/P(not stuck after x-1) should approach a constant. To check this, I calculated -log(1-p) and plotted it ve...

- Wed Oct 24, 2012 10:01 pm UTC
- Forum: Coding
- Topic: Help with RegEx
- Replies:
**4** - Views:
**1374**

### Re: Help with RegEx

If your programming language supports case insensitive patterns with a modifier and you do not care about the case of the other letters, you can simplify it to ^k?sa(v|wery)$ or (what I would prefer as it is easier to read) ^(ksawery|sav)$ The same expression, case sensitive: ^((k|K)sawery|(s|S)av)$...

- Tue Oct 23, 2012 3:15 pm UTC
- Forum: Logic Puzzles
- Topic: Geometry brainbender
- Replies:
**13** - Views:
**3703**

### Re: Geometry brainbender

@Qaanol: Fixed, thanks. I re-wrote that 2-3 times and missed the prefactor somehow. It has no influence on the rest of the solution, as those terms cancel anyway. @ericmn: I think it is a bit misleading to re-use "A" and "C" in your cases - those points are not part of (p_x,p_7,p...

- Sat Oct 20, 2012 3:17 pm UTC
- Forum: Logic Puzzles
- Topic: Geometry brainbender
- Replies:
**13** - Views:
**3703**

### Re: Geometry brainbender

Those two trinangles are not arbitrary - if you fix the distance of one point to P (and fix d and r), the distance of the other point follows. Solution: Equations for the antipodes: Let d be defined as the distance between P and its closest point on the sphere. Let O be the center of the sphere. The...

- Sat Oct 20, 2012 12:00 am UTC
- Forum: Logic Puzzles
- Topic: Connect 4 Puzzle (To infinity!)
- Replies:
**11** - Views:
**4421**

### Re: Connect 4 Puzzle (To infinity!)

Not that we've proved that (n^2 - n - 2)/n^2 is best for evens >= 4, but it's good enough for me. It is not, and the flaw is so obvious I should have seen it: Every solution for n-1 is a solution for n, too. And the perfect (n^2-n)/n^2-solution for n=6k +- 1 is better than the (n^2-n-2&...

- Fri Oct 19, 2012 11:49 pm UTC
- Forum: Logic Puzzles
- Topic: Geometry brainbender
- Replies:
**13** - Views:
**3703**

### Re: Geometry brainbender

Some things I can see: The area of the surface cap is 2 pi r h with the h as defined on wikipedia. This gives the relation pi x^2 = 2 pi r h and therefore x^2 = 2 r h. As h<r, x^2 < 2r^2. Let d be the distance between P and the center of the sphere. h, r and d are connected via similar triangles: r/...

- Fri Oct 19, 2012 5:35 pm UTC
- Forum: Logic Puzzles
- Topic: Connect 4 Puzzle (To infinity!)
- Replies:
**11** - Views:
**4421**

### Re: Connect 4 Puzzle (To infinity!)

Prediction: max for 8x8 tile will be 54/64. Here is an example: O......O ...O.... ......O. ..O..... .....O.. OO...... ....O... ....O... And that should work for 88/100: ........O. ....O..... O......O.. O..O...... ......O... ..O....... .....O.... .O........ .........O .....O.... Conjecture: For even...

- Sun Oct 14, 2012 11:41 am UTC
- Forum: Science
- Topic: Relativity - well, now I'm just depressed...
- Replies:
**4** - Views:
**1784**

### Re: Relativity - well, now I'm just depressed...

Footnote 3 in the article itself is not better... 0.55c + 0.55c = 1c? What? Special relativity (and also this new model) gives 1.1c/(1+0.55^2)=0.84c The paper itself looks a bit dubious, too, but without full access (not on arXiv :(, not even other papers from the authors) it is hard to tell. And ev...

- Fri Sep 28, 2012 8:09 pm UTC
- Forum: Science
- Topic: Does ENCODE mean they can redo all those Genome-wide studies
- Replies:
**9** - Views:
**2030**

### Re: Does ENCODE mean they can redo all those Genome-wide stu

we *know* that we can simply remove them and not affect the organism's phenotype in any noticeable way. I wonder how we can know that. Experiments with humans are out of the question, experiments with other organisms sound quite costly (how many independent places are there to check? 10 000? 100 00...

- Fri Sep 28, 2012 7:58 pm UTC
- Forum: Science
- Topic: The Future!!
- Replies:
**38** - Views:
**5766**

### Re: The Future!!

Xanthir wrote:unless your population is under severe survival stress.

That describes nearly all populations of nearly all species, for nearly all times.

If food is available without relevant limits and predators are not an issue, the population usually expands until one of those becomes an issue.

- Tue Sep 25, 2012 5:43 pm UTC
- Forum: Science
- Topic: The Future!!
- Replies:
**38** - Views:
**5766**

### Re: The Future!!

1912 is also much closer to the beginning of modern science. Right. And the number of scientists increased a lot since then. Additionally, really cheap energy was a feature of that century, now all but vanished. So a lot depends on whether cheap energy is returning somehow. I doubt that energy will...

- Fri Sep 21, 2012 4:17 pm UTC
- Forum: Science
- Topic: The Future!!
- Replies:
**38** - Views:
**5766**

### Re: The Future!!

I don't think the future will be much different than it is now... since most futuristic stuff just isn't economically sound, is outlawed, or goes against human nature. Try to view the current world with the knowledge of 1912. While there were some cars, a car for a significant fraction of the popul...

- Fri Sep 21, 2012 3:01 pm UTC
- Forum: Coding
- Topic: Creating a web-based service
- Replies:
**5** - Views:
**2164**

### Re: Creating a web-based service

especially with all the information that would be attached to each one? I think that will be a very tricky part, especially if you do not want to add 1000000 possible areas of knowledge by hand. There are so many ways to describe the same area, and so many different areas within the same broader to...

- Wed Sep 12, 2012 4:22 pm UTC
- Forum: Science
- Topic: Non-circular gear creation and uses (software now published)
- Replies:
**24** - Views:
**13944**

### Re: Non-circular gear creation and uses

Is it possible to go backwards in some way? ;) Not just variable forward speed, but a bit backwards in between. It would need a "negative radius" somehow - teeth on the inside. I have no idea whether that is possible or not, but it might be interesting. To explore the demand, you could rel...

- Wed Aug 29, 2012 1:46 pm UTC
- Forum: Science
- Topic: Measuring SI Units on a Desert Island
- Replies:
**17** - Views:
**4325**

### Re: Measuring SI Units on a Desert Island

With a frequency comb, as I already posted. You compare the optical transition with an optical frequency of the frequency comb, and measure frequency spacing and offset at radio frequencies. I'm not really sure I understand how it works. You take a large bandwidth laser, use a frequency comb to nar...