# What is the Sandbox?

This "Sandbox" is a place where Code Golf users can get feedback on prospective challenges they wish to post to the main page. This is useful because writing a clear and fully specified challenge on the first try can be difficult. There is a much better chance of your challenge being well received if you post it in the Sandbox first.

To post to the Sandbox, scroll to the bottom of this page or click on the "Add Proposal" link below, and click "Answer This Question". Click "OK" when it asks if you really want to add another answer. Write your challenge just as you would when actually posting it. You may also add some notes about specific things you would like to clarify before posting it. Other users will help you improve your challenge by rating and discussing it. When you think your challenge is ready for the public, go ahead and post it, replace the post here with a link to the challenge and delete it.

See the Sandbox FAQ for more information on how to use the Sandbox.

## Get the Sandbox Viewer to view the sandbox more easily

To add an inline tag to a proposal use shortcut link syntax with a prefix: [tag:king-of-the-hill]

• How are tags added to questions? – guest271314 Jan 9 '19 at 7:51
• @guest271314 You can use this markup to create a tag in a draft: [tag:code-golf] – James Aug 29 '19 at 15:19
• Why no featured anymore? Can't we have it auto-added or something? – S.S. Anne Sep 26 '19 at 15:57
• @JL2210 We now have a permanent info box that links to the Sandbox, so the featured tag isn't necessary – caird coinheringaahing Sep 29 '19 at 13:43

## Solve the Trolley Problem code-golfmachine-ethics

Philosophers have long pondered the Trolley problem. Unfortunately, this no human has solved this problem yet. Luckily, as programmers we can use computers to solve the problem for us!

Your program will take as input a (finite) directed graph (with at most one edge from x to y, for any x and y), with a designated node, and a nonnegative integer attached to each edge (representing the number of people tied to that track). In addition, every node has at least one exit edge.

The trolley starts at the designated node. Each turn, if the trolley is at node x, the utilitarian selects an edge (x,y). The people on that edge die, and trolley is now at edge y. This process continues forever.

Note that people can only die once, so if the edge (x,y) has n people tied to it, and the trolley runs over them, say, 100 times, it will still only result in n deaths.

The utilitarian makes his choices in such a way as to minimize the number of people that die (which is guaranteed to be finite, since there are only finite people). Your program will output this number.

You may take the input graph in any reasonable way you like. For example, you could take it as a matrix, and count the designated node as the one labeled 0. Or you could use something like x1,y1,n1;x2,y2,n2;.... For example 0,a,0;a,b,5;a,c,1;b,b,0;c,c,0 to represent the standard trolley problem (with loops at the end).

Testcases:

• 0,a,0;a,b,5;a,c,1;b,b,0;c,c,0 -> 1 (Go from 0 to a, a to c (killing one person), and then keep looping the trolley from c to c).
• 0,0,1;0,a,5;a,a,0 -> 1 (Keep going from 0 to 0, running over 1 person for all eternity),
• 0,a,5;0,b,1;a,a,1;b,b,6 -> 6 (0 -> a -> a -> a -> a -> ... (note that the greedy solution of going to b would be incorrect))
• 0,a,1;0,b,5;a,b,1;b,a,1 -> 3 (0 -> a -> b -> a -> b -> ...)
• 0,a,1;0,b,1;a,a,0;b,b,0 -> 1 (Note that there are two different options that the utilitarian might take that both kill only one person)

This is , so the shortest answer wins! Good luck.

Notes: There will be no sick loop de loops and multitrack drifting is banned.

• So basically: find the cycle of lowest weight? That's pretty similar to finding the cycle of greatest weight, which is equivalent (for integer weights) to codegolf.stackexchange.com/q/107274/194 – Peter Taylor Jul 22 '17 at 9:36
• @Peter Taylor no, since you may have to kill lots of people to get to that cycle, or it might not be reachable at all. – PyRulez Jul 22 '17 at 12:26
• Aha, so it's finding a rho of lowest weight. Still potentially quite similar in implementation, but certainly admits alternative approaches. It might improve clarity to explicitly describe the rho form of the paths to consider. – Peter Taylor Jul 22 '17 at 12:33
• @Peter Taylor what is rho? – PyRulez Jul 22 '17 at 15:44
• ρ , as in Pollard's rho which uses paths which eventually loop back on themselves to factor composite numbers. – Peter Taylor Jul 22 '17 at 16:08
• You should delete this as it has been posted. – Erik the Outgolfer Jul 28 '17 at 19:41

# An anagram chain quinepermutationscode-golf

Your challenge is to create a program, P_0, that outputs P_1, that outputs P_2, that outputs... P_n, which finally outputs P_0. Every program in that chain has to be a permutation of every other program, and every program has to be distinct.

For example, if your program (P_0) was

abcd


And that generated another program (P_1)

badc


Which generated another program (P_2)

dabc


Which outputted the first program, you would have a anagram chain of length 3.

Your goal is to make as long a chain as possible, in as short a program as possible.

Your final score will be your chain length - in case there's a tie, the program with the shortest bytecount wins. If there's a tie again, the first poster wins.

## Meta:

• I need a better title. Any suggestions?
• [Related]. Actually, I'm pretty sure this is a dupe of another challenge that I can't find at the moment. – Shaggy Jul 23 '17 at 10:13

# KotH: Atom Bomb Chess

(Draft. I'll flesh it out some more later. It will probably be implemented in JavaScript or maybe C++.)

Atom Bomb Chess is a variant of chess played much the same as regular chess. The only differences are:

1. The game ends when one or both players have no pieces, or if 50 moves have been made without a piece being captured.
2. When a piece is captured, all pieces a king's move away are also "captured", and the piece doing the capturing is also "captured".

For example, let's look at a 4x4:

pbbr
....
....
RBBP


Where r/R is a rook, b/B is a bishop, and p/P is a pawn. Suppose R moves forward 3 spaces. Then, it captures the p and "explodes":

..br
....
....
.BBP


The idea of this KotH is to make a program that plays Atom Bomb Chess. I will have a few programs to test your submissions against.

• This is a really cool idea! Obviously you'll need to flesh out the rules a bit more, but this is a great start. I would also suggest fixing the board size (you seem to suggest it can be played on any size board, but I think it would be easiest to set it as a constant ahead of time). – musicman523 Jul 23 '17 at 22:31
• @musicman523 Yeah, it'd probably be played on an 8x8 – Conor O'Brien Jul 23 '17 at 22:35
• you forgot that pawns are partially nuke proof – Destructible Lemon Jul 23 '17 at 23:15
• @DestructibleLemon Please elaborate ? I wasn't aware there was some standard rulebook. – Conor O'Brien Jul 24 '17 at 1:45
• I'm not entirely sure where it is, but I remember that pawns, while still being eliminated if they are capturing, or are captured, will not get removed if they are adjacent to a capture. maybe use lichess? – Destructible Lemon Jul 24 '17 at 2:51
• @DestructibleLemon I remember playing it when I was younger with no such restriction, perhaps there are multiple variants. Why suggest lichess? Does it have that mode? – Conor O'Brien Jul 24 '17 at 18:02
• Apart from pawns being nuke proof, the game ends as soon as you nuke the enemy king, if you play on lichess. There are videos on YouTube which include an explanation of the rules and basic "opening theory". – Sleafar Jul 28 '17 at 18:17

# Who Won the Chess Game?

Bear with me, this is incredibly bare-bones at the moment; I'll work on this over-time, just wanted to gauge interest on the over-all idea.

Given an input list of moves l, output who won the chess game given that white always goes first, alternating moves from there. Using the following convention for naming the chess squares:

Where the pieces are:

rnbqkbnr
pppppppp
........
........
........
........
PPPPPPPP
RNBQKBNR


The first assumption you will make is that the list of moves passed is a VALID chess game, to completion. The following would be an example input:

[[A2,A4],[A7,A6],...]


The first two moves given would result in:

rnbqkbnr
pppppppp
........
........
P.......
........
.PPPPPPP
RNBQKBNR


Which is white's first move, followed by black's move:

rnbqkbnr
.ppppppp
p.......
........
P.......
........
.PPPPPPP
RNBQKBNR


Then, you would continue parsing moves until a king disappears. Whichever king is left should be declared the winner.

# Waiting to Gauge Interest Before Wasting Time on Full Testcase

• – AdmBorkBork Jul 24 '17 at 18:09

## Nested list unpacking code-golf

In Python 3.5+, the * operator can unpack a list within a list.

[1,*[2,3,4],5] == [1,2,3,4,5]


The unpacked elements are placed directly into the outer list without increasing the list depth. Note that this is different from the nested three-element list

[1,[2,3,4],5]


Your goal is to simplify a nested list expression by resolving all instances of unpacking. So, the output will simply be a nested list without any *.

Specifically, the input will be an expression that consists of

• A non-empty list of expressions [..]
• An non-empty unpacked list of expressions *[..].
• A digit 1 through 9

The outermost layer will always be a non-empty list. The output is such an expression without any unpacking.

Input: A string of characters [],*123456789

Output: Its unpacked analogue. This can be a string or list structure.

This challenge can be done in any language. Python and other languages with the same unpacking format may not use eval or exec or equivalents.

TODO: I/O details, test cases

• This is similar to least Levenshtein distance to a balanced list (replace *[ with { and then remove unbalanced ]). – Peter Taylor Jul 25 '17 at 7:20

# Tell me my vocabulary words! Donated.

When taking textbook notes, I need to write down the vocabulary words and their definition. So your task is to write some code that will give me the vocabulary words and their definitions!

Vocabulary words are detonated with a * on both sides of them, like this: Sentences are .-delimited, meaning that after every . a new sentence starts. An example of a valid sentence would be: This sentence has a *vocab word* in it. The vocab word in the sentence is 'vocab word'.

Input: A string of text with some words marked with asterisks. Only valid inputs will be provided, meaning that only sentences with exactly two asterisks inside of them, and there is something between the asterisks.

Output: A list or delimited string where each item is in the following format: Vocabulary word: sentence. The Vocabulary word is the vocab word found in the sentence. The sentence must be the one with the emboldened vocabulary word in it. If there is any whitespace other than the s separating the words, it needs to be stripped.

Test cases:

Input: *Alan Turing* invented the Turing machine. A *Turing machine* is a machine that follows simple rules, but is capable of any computation.

Output: *Alan Turing*: *Alan Turing* invented the Turing machine.
*Turing machine": A *Turing machine* is a machine that follows simple rules, but is capable of any computation.

Input: What is code-golf? *Code-golf* is the best site on the SE network. But what is SE, you ask? *SE* is a group of Q&A sites, with a system to prevent bad posts.

Output: *Code-golf*: *Code-golf* is the best site on the SE network.
*SE*: *SE* is a group of Q&A sites, with a system to prevent bad posts.

• So split on '.' and then filter to strings containing a *? Or can there be asterisks which don't mark vocabulary words? Should we trim whitespace at the start and end of the sentences? – Peter Taylor Apr 6 '17 at 9:54
• @PeterTaylor A string surrounded by * is a vocab word, like this: *vocab word*. You need to format the outputted sentence correctly, and trim whitespace. I will edit when I have more time. – Comrade SparklePony Apr 6 '17 at 13:52
• Are the following valid inputs or not? a) Unbalanced* asterisk; b) Empty ** vocabulary word; c) Vocabulary* word *has bad spacing. – Peter Taylor Apr 7 '17 at 10:05
• Only the last one is valid. – Comrade SparklePony Apr 7 '17 at 13:25
• The latest wording "only sentences with exactly two asterisks in them" suggests that there's no need to filter: just split on ., trim, and prepend Vocabulary word:  to each sentence. – Peter Taylor Apr 7 '17 at 21:03

# Base 32 RFC 4648 Compliant Alphabet!

When writing my handy-dandy totp/hotp token implementation in Python and Swift (ad: here), I encountered for the first time RFC 4648. There is a nice and long memo about RFC 4648, but I only had to focus on a very specific part of it: Page 8. If you are bored and want some enthralling reading, you can find this memo here, and a useful table here.

Basically, I was looking for the alphabet that could be used when base 32 decoding a String. Well, this is it: ABCDEFGHIJKLMNOPQRSTUVWXYZ234567, and padding, =. However, simply printing that is not the challenge –– that would be too simple1. Instead, we are going to print this (the comments are for your reference and do not need to be printed):

__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0x00 - 0x0F or   0 -  15
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0x10 - 0x1F or  16 -  31
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0x20 - 0x2F or  32 -  47
__,__,26,27, 28,29,30,31, __,__,__,__, __, 0,__,__,  // 0x30 - 0x3F or  48 -  63
__, 0, 1, 2,  3, 4, 5, 6,  7, 8, 9,10, 11,12,13,14,  // 0x40 - 0x4F or  64 -  79
15,16,17,18, 19,20,21,22, 23,24,25,__, __,__,__,__,  // 0x50 - 0x5F or  80 -  95
__, 0, 1, 2,  3, 4, 5, 6,  7, 8, 9,10, 11,12,13,14,  // 0x60 - 0x6F or  96 - 111
15,16,17,18, 19,20,21,22, 23,24,25,__, __,__,__,__,  // 0x70 - 0x7F or 112 - 127
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0x80 - 0x8F or 128 - 143
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0x90 - 0x9F or 144 - 159
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xA0 - 0xAF or 160 - 175
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xB0 - 0xBF or 176 - 191
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xC0 - 0xCF or 192 - 207
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xD0 - 0xDF or 208 - 223
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xE0 - 0xEF or 224 - 239
__,__,__,__, __,__,__,__, __,__,__,__, __,__,__,__,  // 0xF0 - 0xFF or 240 - 255


There may be trailing spaces after every line, and trailing newlines after the last one.

Also, the single digit numbers can be written as 0[digit] instead of [space][digit]. However, be sure to include the spaces between the four groups.

By the way, the __ actually represent 255, but the former do not make me drown in a sea of digits.

1 Should the challenge be to just print that string?

Any other suggestions to make this challenge more interesting?

• Perhaps get a list of characters (of any length) and print such a table? – ugoren Jul 29 '17 at 19:14

# Is this 2048 board valid?

Some 2048 boards are impossible to get into. For example,

2 _ _ 2
_ _ _ _
_ _ _ _
2 _ _ 2


will never occur in a 2048 game. Additionally, these are all impossible:

8 _ _ 8    8 _ _ _    2 2 2 2
_ _ _ _    _ 8 _ _    2 2 _ _
_ _ _ _    _ _ _ _    _ _ _ _
8 _ _ _    _ _ _ _    _ _ _ _


Your program needs to accept a 2048 board, and return a truthy value if the board is reachable, else falsy.

//Explanation of 2048 goes here.

I've listed 4 different boards that cover major test cases. Are there any others I'm missing?

• hehe, don't forget [tag:sliding-puzzle] – Stephen Aug 1 '17 at 19:56
• As you've stated yourself the main thing missing here is the bulk of the challenge. – TheLethalCoder Aug 2 '17 at 9:14
• I wonder if it suffices to look one step back, or if you need to check if the previous position can itself be produced. – xnor Aug 4 '17 at 7:35

# Smooth Usage [On hold while alternative scoring is considered]

We've all seen CPU usage graphs like this one:

Doesn't that look ugly? It would look much nicer as a lovely smooth sine wave...

## Challenge

Write a program in the language of your choice that will infinitely produce a regular sine wave in Task Manager's (or Activity Monitor's if that's your thing) CPU usage graph.

You may assume:

• Background CPU usage is constant
• Only a single core must display the pattern
• The system has sufficient cooling to prevent thermal throttling
• Features such as Intel TurboBoost are disabled

This is tagged as to encourage short answers, but ultimately will be a as I suspect perfect solutions will be hard to come by.

• – MTCoster Aug 11 '17 at 15:14
• code-golf and popularity-contest don't mix together. You have to choose one or the other, but making it strictly code-golf would be difficult because you'd have to define what is a good enough sine wave, but on the other hand popularity contests are very risky to do. – dzaima Aug 11 '17 at 15:16
• @dzaima That was exactly my dilemma - which would you suggest fits the challenge best? – MTCoster Aug 11 '17 at 15:17
• I'd say a scoring algorithm of some sort would be best for this, no idea how you'd do it. Like the related one was objective. If you could read the word, it was valid. With this, is a bumpy sin-wave a sin-wave? – Magic Octopus Urn Aug 11 '17 at 20:39
• @MTCoster If you're not sure whether popularity-contest fits, then you can be sure that it does not. It is quite difficult to make an well received one. Go with code-golf or make a code-challenge if you can come up with a good own winning criterion. – flawr Aug 11 '17 at 20:53

# Time bomb ping pong

## Challenge - Both teams

All users are divided into two teams based on their PPCG ID. For example, my ID can be found here, from which you can see that my ID is 34388. To check on which team you are, run the following snippet:

## Challenge

This is a challenge, all submissions should be written in

Your function should accept four variables as follows, obviously within your function you can name these whatever you like
- 1st variable represents the current value of 1 bar of Megatanium
- 2nd variable represents your current bank balance
- 3rd variable represents your current stock holding
- 4th variable represents the iteration number

The return value should be negative if you wish to sell stock, positive if you wish to buy stock, 0 if you wish to neither buy or sell

For example;
If you wish to buy 10 bars of Megatanium at the current price you would return 10
If you wish to sell 5 bars of Megatanium at the current price you would return -5

I will call your function 1000 times. Stock price will always be an integer, chosen at random, with a minimum of 0 and a maximum of 256. Method of selecting the stock price is described in more detail below, it will NOT be an even distribution!

Your bot will be disqualified if it does any of the following at any point

• Try to buy more stock than it can afford
• Try to sell more stock than it holds
• Try to write any value to any of the global variables
• Fail to return a value
• Return a value that is not an integer

Here is the code I will be running, the score output at the end will be your bot's score. It is calculated from your bank balance plus the value of your held stock at the latest value.

function go(bot) {
bank = 1000;
stock = 0;
disqualified = 0;
for (i=1; i<=1000; i++) {
price = prices[i];
trade = window[bot](price, bank, stock, i);
bank = bank - (price * trade);
if (bank < 0) disqualified = "RUN OUT OF MONEY";
if (stock < 0) disqualified = "TRIED TO SELL STOCK YOU DIDN'T OWN";
if (disqualified) break;
}
if (disqualified) {
console.log("Disqualified on iteration " + i + " REASON: " + disqualified);
}
else {
score = bank + (stock * price);
console.log(bot + " scores " + score);
}
}


The function for generating a suitable distribution of random values is a slightly modified version of the function described here
Every bot will be given the same set of prices, but the set will not be generated until immediately before the bots are run.

function randn_bm() {
var u = 0, v = 0;
while(u === 0) u = Math.random();
while(v === 0) v = Math.random();
w = Math.floor(Math.sqrt( -2.0 * Math.log( u ) ) * Math.cos( 2.0 * Math.PI * v ) * 32) + 128;
while (w < 0 || w > 256) w = randm_bm();
return w;
}
prices = [];
for (i=1; i<=101; i++) {
prices.push(randn_bm());
}


## Completion

You may submit as many bots as you like, try to be inventive with your algorithms! You may use a global variable called data, this will be initially set to null and will always be available to your function.

## Example bots

function buyBot(p,b,s,i) {
/* Bot always buys as much as it can */
return Math.floor(b / p);
}

function randomBot(p,b,s,i) {
/* Bot buys and sells randomly */
if (Math.random() > 0.5) {
return Math.floor(Math.random() * (b / p));
}
else {
return -Math.ceil(Math.random() * s);
}
}

function smartBot(p,b,s,i) {
/* Bot always buys at under 100 and sells at over 150 */
if (p < 100) {
return Math.floor(b / p);
}
else if (p > 150) {
return -s;
}
else {
return 0;
}
}

function bankruptBot(p,b,s,i) {
/* Bot always sells at under 100 and buys at over 150 */
if (p > 150) {
return Math.floor(b / p);
}
else if (p < 100) {
return -s;
}
else {
return 0;
}
}

function alternateBot(p,b,s,i) {
/* Bot buys and sells alternately */
if (data == 1) {
data = 0;
return Math.floor(Math.random() * (b / p));
}
else {
data = 1;
return -Math.ceil(Math.random() * s);
}
}


## Winning conditions

All bots will be run on locally by me approximately 1 week after the question is posted (date will be decided when question is posted, no point setting a date in sandbox)
Winner will quite simply be the bot that has the highest final score after the last iteration, as calculated by the function provided above. The array of prices used will be published after a winner has been crowned.
There are no set conditions on the speed of your function, but please be fair and try to avoid anything that will take more than a couple of minutes to execute

Note: Please leave an upvote if you think the challenge idea is good and is clear, a downvote if you think the challenge idea is bad, and comment if you can't understand the challenge. Last time I asked on TNB 2 users told me that they can't understand anything.

# Golf a return-oriented code generator!

## Background

Return-oriented programming (ROP) is a computer security exploit technique that allows an attacker to execute code in the presence of security defenses such as non-executable memory (W xor X technique) and code signing. (from Wikipedia)

## Challenge

In this challenge, you should write a program, that takes the code of the existing program and the required code, and output the stack necessary to execute that program.

## Rules

• Standard loopholes apply, as usual.
• How the machine works
• At first, IP is equal to the top of the stack, and the top of the stack is popped.
• For each clock cycle (whatever it means), the command at the position of IP is executed, and the IP is advanced if the command does not modify IP.
• The behavior if the IP is at the last instruction and that instruction does not modify IP is undefined.
• Assembly instructions
• All commands are case insensitive.
• Note: There is nothing that guarantees commands must be 3 characters long, or limited to some sets. After all, this is not real assembly. However:
• You can assume that all the characters are in the English alphabet, uppercase or lowercase.
• There may be some other commands "similar but not the same" with ret, so checking for the first character or the SHA256 hash won't work.
• The special command ret will pop the value on the top of the stack, and set the instruction pointer IP to that value.
• You may assume that all the other commands won't modify the stack, or the IP.
• The "existing program" and "required code" will be represented as a string, separated by newline characters (you may optionally take a list of strings as input).
• The required code will never contains ret.
• The output should be a stack of line-numbers in appropriate format (list of numbers - may be reversed, array of numbers, etc.)
• The command executed right after the last command in the "required code" must be a ret.
• Because the memory of the machine is limited, you should output the shortest possible stack. If there are multiple shortest stack, output any.
• You may assume that there exists an output.

## Example test cases:

• Existing program:
1:  add eax, ebx
2:  lea eax, [2*eax+4*ecx]
3:  ret
4:  mov eax, ebx
5:  xch ecx, eax
6:  ret

(the line numbers are just for demonstration purposes. They are not included in the input, you can use 0-indexing or 1-indexing)

(disclaimer: this is not real assembly, just for demonstration purposes)

• Required code:
lea eax, [2*eax+4*ecx]
xch ecx, eax
mov eax, ebx
xch ecx, eax

In that case, you should output [2, 5, 4] (2 is at the top of the stack), because if the stack have that value, then:

• First, IP = 2.
• The commands
2:  lea eax, [2*eax+4*ecx]
3:  ret

are executed.

• On executing ret, the IP get the value 5. Then, the commands
5:  xch ecx, eax
6:  ret

are executed.

• Then, similarly, when the next ret is executed, the value 4 is popped from the stack, and commands
4:  mov eax, ebx
5:  xch ecx, eax
6:  ret

are executed.

Therefore, the commands executed (apart from ret) are:

2:  lea eax, [2*eax+4*ecx]
5:  xch ecx, eax
4:  mov eax, ebx
5:  xch ecx, eax


which is equal to the "required code".

## Winning criteria

This is , shortest code in bytes win!

• You should probably include a list of all valid identifiers or if all of them are always 3 characters mention that and whether there are instructions that begin with r other than ret. – ბიმო Jan 28 '18 at 18:57
• @BMO Better now? – user202729 Jan 29 '18 at 12:00
• No idea what the alphabet ØA;Øa¤ is but apart from that I think you made it pretty clear now. Other things that came to mind: What if there are multiple possibilities for a shortest stack (eg. like this)? Will the input always have a solution (eg. unlike this)? And are the line numbers part of the input? If so is it safe to assume that they always start with 1 and have an offset of 1? Maybe you could add some testcases. – ბიმო Jan 29 '18 at 12:52

# Faux Compress a String

Given a string s, perform a faux compression on it. Below is an example with f('hello world').

To faux compress a string, start by taking a frequency count of all letters.

hello world -> [h:1, e:1, l:3, o:2, :1, w:1, r:1, d:1]

Next, sort smallest to largest in count with a tie-breaker of ASCII-code.

[ :1, d:1, e:1, h:1, r:1, w:1, o:2, l:3]

Next, in the original string, replace each letter with it's index in the frequency list.

[3,2,7,7,6,0,5,6,4,7,1]

Next, convert each integer to binary, and join them all with the digit 2, then convert this to base-3.

1121021112111211020210121102100211121 -> 240591504997661290

Lastly, print or return both this number and the sorted keyset of the frequency map.

[240591504997661290, ' dehrwol']

Is the final result.

You now have a "faux compressed" string.

To get a real compressed string you'd figure out the shortest sequence of bits to replace 2 with which would be a unique delimiter and treat the binary digits as they are, bits, instead of bytes.

# More Examples

000000000001 -> [58839486765, '10']
eeEeeeeeeeeeeEEEEEeeeeee -> [16508589985213004629636, 'Ee']


# Rules

• Your code's function may be undefined if the following is not met:
• The string will contain more than one unique character.
• Lowest byte count wins because this is .

# PSA: I have many PENDING PROPOSALS, tell me which to delete.

If you post a comment on any of these challenges with the words "This isn't good in my opinion."

It will be removed.

• The posted one by MDXF. – user202729 Feb 3 '18 at 5:28

# Casinos and Gamblers king-of-the-hillcops-and-robbers

You either play as a casino, or as a gambler. Casinos offer bets, and gamblers choose bets to take. Casinos want to make money, gamblers want to have more money than others. Read that sentence again, it is the core concept here.

1. All casinos simultaneously comes up with a Bet. A Bet consists of three parts:
1. Entry Fee (Positive integer amount a gambler must pay to take)
2. Odds (The chance that a gambler will win between 0 and 1)
3. Reward (Positive integer amount a gambler is given if he wins). This can be any amount, even if it would make the casino go negative.
2. Each gambler optionally chooses a bet to take. They must have enough money to pay for the Entry Fee.
3. We calculate who wins and who doesn't (using a PRNG), and pay out.

At the end of a game, we give points as follows:

1. Casinos receive 1 point for each dollar they have (can be negative)
2. Gamblers receive N^2 points for having more money than N other gamblers.

A tournament consists of many games, and a player's score is their average score across all of the games.

Gamblers and Casinos all have complete information throughout the game.

• Might want to specify that entry fee and reward must be positive (negative/negative games map to positive/positive games but complicate the "can this gambler afford this game" calculation.) – histocrat Jan 31 '18 at 19:34
• I have a hunch that 100 dollars won't give quite enough granularity for casinos to work with. Maybe bump it up to 10,000? – Beefster Feb 2 '18 at 23:09
• Also, how are odds represented? Fraction? Floating point number? Precision? – Beefster Feb 2 '18 at 23:10
• @Beefster Floating point, of course. I think you may be right, but 10,000 seems too high. I think I'll do 1,000 – Nathan Merrill Feb 3 '18 at 0:06

# Busier Beaver cops-and-robbersbusy-beaver

## Cops

You will write two programs in a language of your choice

• A public program: This program must execute in a finite amount of time. So it can not go on indefinitely (although there is no time limit). It also wise if you program it to output a large number of bytes to standard output. This is the what you post in your answer. You must also post the programming language.
• A secret program: This program must output more bytes than the public answer, i.e. it is a busier beaver. It also must not go on indefinitely. It also can not be longer than the public program. You do not post it the answer immediately. After one week, if no robber has cracked the answer, you edit it in.

For these programs, you may assume that numerics types are actually unbounded, i.e. that overflow never occurs. This means that you can store, for example, Graham's number in a variable. It also means that finding the maximum value of storage capacity of a numeric type results in an error.

If no robber has cracked your answer, your score is equal to (length of public program)/(length of secret program), which you are trying to maximize. In case of ties, the longer public program wins.

## Robbers

The cops will post a public program. Your job is to find a busier beaver. That is, you need to find a program that outputs more bytes to standard output than the cops program. It can not go on indefinitely, and can not be longer than the public program (although it can be the same length).

You get points equal to (length of cops public program)/(length of your program) for each post you crack.

• Need to specify that it's the number of bytes output when given no input. – Peter Taylor Feb 16 '18 at 11:38

# Indices of inner join (code-golf)

Given two lists of integers x and y of possibly unequal lengths, return the list of pairs (i, j) such that the ith element of x is equal to the jth element of y.

The output consists of a vector of pairs or a pair of vectors (the zip of the other option, may be useful for languages without pair data structure). The pairs may be given in any order, and using 0-based or 1-based indexing.

In pseudo-code: (this returns two lists of indices, assumes 1-based indexing, and gives lexicographically sorted output)

input vector<int> x, vector<int> y
vector<int> xout, yout;
for (i in 1 to length(x))
for (j in 1 to length(y))
if (x[i] == y[j]) {
add i to end of xout;
add j to end of yout;
}
return xout, yout;


This is code-golf so answers will be scored in bytes with fewer bytes being better.

### Test case 1

x = [5, 1, 1, 2], y = [5, 1]

-> xout = [1, 2, 3]


(5, 1, 1 each appears once in y)

-> yout = [1, 2, 2]


(the first position appears once in x, the second twice)

Reshuffling the pairs is okay:

-> xout = [2, 1, 3]
-> yout = [2, 1, 2]


## Test case 2

x = [2, 2, 2, 3, 3], y = [4, 2, 2, 2]

-> xout = [1, 1, 1, 2, 2, 2, 3, 3, 3]


(each of the three first positions finds three matches in y)

-> yout = [2, 2, 2, 3, 3, 3, 4, 4, 4]


Improvements and clarifications are very welcome!

• It should be ready to post now. The sandbox is not very active (not many people know about it), so even sandboxed posts may not have sufficient feedback. – user202729 Feb 20 '18 at 2:29
• "The output consists of a vector of pairs or a pair of vectors" seems unnecessarily restrictive. The semantically correct return type in languages which support it would be a set of pairs, and while that may not be the golfiest I don't see a reason to prohibit it. – Peter Taylor Feb 20 '18 at 12:15

# Number of Adjacently Divisible Partitions of n

Output the following sequence (OEIS A167865):

1, 0, 1, 1, 1, 1, 2, 1, 2, 2, 2, 1, 4, 1, 3, 3, 3, 1, 5, 1, 5, 4, 3, 1, 6, 2, 5, 4, 5, 1, 9, 1, 6, 4, 4, 4, 8, 1, 6, 6, 7, 1, 11, 1, 8, 8, 4, 1, 10, 3, 10, 5, 8, 1, 11, 4, 10, 7, 6, 1, 13, 1, 10, 11, 7, 6, 15, 1, 9, 5, 11, 1, 14, 1, 9, 12, 8, 5, 15, 1, 16, 9, 8, 1, 18, 5, 12, 7, 10, 1, 21, 7, 13, 11, 5, 7, 12, 1, 14, 12, 15, 1, 20, 1, 14, 17, 9, 1, 18, 1, 21, 10, 17, 1, 19, 5, 14, 14, 7, 6, 22, 3, 14, 9, 12, 8, 25, 1, 18, 13, 17, 1, 24, 8, 16, 21, 11, 1, 23, 1, 22, 6, 12, 7, 19, 7, 15, 19, 11, 1, 28, 1, 21, 17, 18, 11, 27, 1, 16, 10, 18, 6, 28, 1, 18, 25, 9, 1, 25, 5, 29, 19, 18, 1, 27, 14, 20, 8, 11, 1, 30, 1, 31, 15, 21...


Number of partitions of n into distinct parts greater than 1, with each part divisible by the next.

### Definition:

Choose one of the three options:

• Output the sequence indefinitely.

• Take n as input and output the n first elements of the sequence.

• Take n as input and output the nth element of the sequence. Both 0- and 1-based indices are allowed, but please specify which one your answer uses.

Hint: There is a simpler formula than the one used on OEIS and in this post. It may save you some bytes (it saved six bytes on a simple reference implementation I made in Python). I'll add it here later.

This is , so the shortest answer in bytes wins.

• I'm not sure about having three different output methods, when one even changes the input required. Almost looks like 3 different challenges. Maybe you could make it so that if no input is given output sequence indefinetely, and if input is given output option 2 or 3. – Brian H. Jan 25 '18 at 10:03
• @BrianH. I saw a few sequence challenges allow to choose from these three output methods, and it seemed to work. Here's one such challenge: codegolf.stackexchange.com/q/152402/61405 – Steadybox Jan 25 '18 at 10:32
• True, i wonder if there's a Meta question about this, can't seem to find one – Brian H. Jan 25 '18 at 10:54
• @BrianH. I don't think there's a meta question for this particular issue, but I think it at least partly falls under allowing flexible input. – Steadybox Jan 25 '18 at 11:19