@BrainSteel Let's say you wanna go to Peru. You'll probably fly into Lima and need a hotel. Google 'Lima hotels' and the first hit is Hilton. Their stock symbol is HLT. Half Life Three confirmed.
@steveverrill Sorry if the comment I left was a bit too direct. Like you say it's pretty hard to examine the challenges in such a way that there is no selection bias, but that's exactly why I think it's more useful to break down the analysis and ask when and why one language does better (or even break down the numbers by tags, say).
I just feel that the numbers posted are a bit misleading, that's all
Sierpinski Carpet & friends in ASCII
Overview
Write a program that prints out simple fractal patterns given a bit-pattern encoding the fractal, plus the per-generation scale factor of the fractal and number of generations.
Explanation
Here is an ASCII representation of the Sierpinski Carpet:
...
There's a tag algorithm-advice which is currently being used three times. All three questions were asking for help with solving a programming challenge found elsewhere. We usually consider that to be off-topic. Two of the questions were sort of salvaged and turned into challenges (the third one i...
@Sp3000 I thought the bounty system had broken as my question still shows up in the featured tab - then I realised it was you :) Will you leave the bounty there for the week to see if it attracts any more answers (and upvotes for question and answers...)?
@Sp3000 Lol it doesn't matter. This seems a very friendly competition (like most on PPCG) so I don't see a problem either way.
The bounty is like a big sign saying "Go and upvote orlp's answer", which is why I'd be inclined to make the most of it. I'm biased though, since it may also get the question a few more upvotes...
My only intuition is that at the shortest length the strings will look quite distinct from each other, but I have no reason to believe that's true. It's just a guess that nearby valid strings suggests some redundancy
Do you remember how many shortest strings there were for 2 by 2?
If we knew there would be a palindromic shortest string for 3 by 3 that would mean brute forcing would only have to consider combinations of strings up to length 39. Still unrealistic but a big difference...
I also wonder whether, with enough quick exclusion criteria, it might be worth a blind approach just testing random strings shorter than the shortest known. I suspect you've already got shorter than would be viable for that though
@PeterTaylor Correct me if I'm wrong, but suppose you take two path graphs A and B where the start is at one end of A but not for B, i.e. B's start is somewhere along the path between A's start and exit. If a string solves A but not B, then A needs to overtake B at some point. But if you execute the same instructions on both mazes simultaneously, I don't think that's possible, and the two positions will overlap on the same square at some point, executing the same instructions for the remainder.
What tricks do you know to make CoffeeScript code shorter?
CoffeeScript is language that compiles into JavaScript ("transpiles" into ES3, to be exact). The golden rule is "It's just JavaScript", meaning there's no runtime overhead. Most of JavaScript Tips & Tricks apply as well.
As a side-effec...
@PeterTaylor That's good, thanks :) But yeah, I haven't been able to generalise which mazes are necessary. I'll add the cut vertices part later though, that was a good observation
If you meant how long the program took, then "overnight"
If you meant how long we've spent on this problem... well check the start date :P
@trichoplax I love questions like those though, because they actually take up a lot of time and thinking power. It gets pretty boring without those types of questions.
I'm currently trying to figure out genetic algorithms so I can get the computer to think for me
Join and learn... is all I can really say - I'm sure most people still receive golf tips from others, for example, it's just a nice community where we share and learn
The problem with GA in my experience is that they're freaky good at solving 80% of a problem. To get them to solve the rest of the problem you have to be very clever in a lot of cases.
Yeah, different machine learning algorithms have very apparent strengths and weaknesses. GA is really scary good at some things, and absolutely abysmal at a lot of others.
What is in my pocket?
Synopsis: Find characters that are enclosed by MYPOCKET.
Example Input
MYPHEIF
YFOCKVH
MBNDEIF
TEUFTMY
ESNDUWP
KBOVUVO
CENWFKC
OPYMTEB
Example Output
F
BND
EUF
SNDUW
BOVUV
ENWF
Huh? How did we get that as an output? The "pocket" can sometimes be diffic...
I am a sockpuppet account of Josh, AKA The Unhandled Exception. I am going to help him test the XMPP chat integration system he's building.
Am I allowed to exist? I promise to behave, and never leave Meta or Meta's chat. If so, may I please have enough rep to chat?
EDIT BY JOSH: I have drawn a ...
A few days ago, I answered a question on a meta site which was closed as not-constructive.
People are still arguing over it, leaving comments which end up in my Stack Exchange inbox.
Is it possible for me to unsubscribe from that answer/question? I have no intention of giving it any further tho...
@Dennis I'm trying to use your script (here) to generate strings for testing, but I'm having real trouble (mostly because I don't understand bash very well). I had to change the cipher option because my Mac OS X box doesn't seem to recognize -aes-128-ctr. Also, it seems to be throwing around strings with length 1010-1030-ish on random occasion. Can you help me understand what it is doing?
@Sp3000 If A has found its end, then one turn before doing so, it would have been next to it. Now, B could only have been in one of three places: before A, overlapping A, or after A (at the shared end). Now, since once A and B meet they move together, and neither A nor B can skip vertices or swap their positions (180° turns are impossible in 1 move), B can't have been before A. Thus, either B overlapped A and will find its end in the next turn, or it already did.
Yeah that was roughly what I had in mind :) It was the "swap position" part which I wasn't completely sure about, but I think in order for that to happen they need to be moving in opposite directions on the same move, which is impossible.
@BrainSteel None of this is actually a Bash problem. openssl either was compiled without the cipher or it's named differently in Mac OS X. openssl list-cipher-algorithms should show what's available.
@BrainSteel I've uploaded the first 99 test cases here.
@orlp No, I'm an amateur mathematician who did consider staying in academia but decided in the end to get a "real job" after finishing my undergraduate degree in compsci.
@PeterTaylor I wanted to do a PhD in statistics but my grades were too shitty for that, so I ended up working a "real job" after my honours degree. ;-)
@orlp At the university I went to, there are three types of honours awarded: first class, second class first division, and second class second division. The first two allows you to continue with a PhD. The third doesn't.
Google says it's Douglas Hurd, and I have no idea who the F he is.
Wikipedia says British politician and novelist. Interesting.
@orlp Rhyming slang is like a code language, where you say the first part of the phrase (Douglas, in the case of Douglas Hurd), and people are supposed to know the second part and associate it with a rhyme (Hurd => third).
Not knowing a rhyming slang commonly used in a community automatically marks you as an outsider. ;-)
ok so.. I don't even really understand "Theorem: the maximum independent set of a disjoint union of graphs is the union of their maximum independent sets. So if we break a graph down into unconnected components we can simplify the problem."
this seems trivial
wouldn't any max independent set algorithm take advantage of that?
the interesting part seems to be the first heuristic
@Lembik Give it a few days. I think the general rule of thumb is to wait a week from the first answer, but if there's still activity then wait a bit longer.
@Sp3000 You had a recursive solution that was guaranteed to find the maximum? Mine performed very poorly. I didn't even have the patience to let it finish for n=5.
Actually, I didn't look at intermediate results. It's possible that it found good results long before it finished.
Yes, that one was a lot of work. I like geometry, so it was a good one for me. Must be the most time I ever spent on a SE post. The only one that comes kind of close was that pawns thing.
@Sp3000 Yes, it reaches 76 within about 1 second or so for n=7. Still seems to be stuck at 164 for n=9, though. At least for the couple of minutes I waited so far.
Sierpinski Carpet & friends in ASCII
Overview
Write a program that prints out simple fractal patterns given a bit-pattern encoding the fractal, plus the per-generation scale factor of the fractal and number of generations.
Explanation
Here is an ASCII representation of the Sierpinski Carpet:
...
What tricks do you know to make CoffeeScript code shorter?
CoffeeScript is language that compiles into JavaScript ("transpiles" into ES3, to be exact). The golden rule is "It's just JavaScript", meaning there's no runtime overhead. Most of JavaScript Tips & Tricks apply as well.
As a side-effec...
What is in my pocket?
Synopsis: Find characters that are enclosed by MYPOCKET.
Example Input
MYPHEIF
YFOCKVH
MBNDEIF
TEUFTMY
ESNDUWP
KBOVUVO
CENWFKC
OPYMTEB
Example Output
F
BND
EUF
SNDUW
BOVUV
ENWF
Huh? How did we get that as an output? The "pocket" can sometimes be diffic...
A few days ago, I answered a question on a meta site which was closed as not-constructive.
People are still arguing over it, leaving comments which end up in my Stack Exchange inbox.
Is it possible for me to unsubscribe from that answer/question? I have no intention of giving it any further tho...
