538
\$\begingroup\$

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

Sandbox FAQ

Posting

To post to the sandbox, scroll to the bottom of this page 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, though you can optionally add a title at the top. 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, and replace the post here with a link to the challenge and delete the sandbox post.

Discussion

The purpose of the sandbox is to give and receive feedback on posts. If you want to, feel free to give feedback to any posts you see here. Important things to comment about can include:

  • Parts of the challenge you found unclear
  • Comments addressing specific points mentioned in the proposal
  • Problems that could make the challenge uninteresting or unfit for the site

You don't need any qualifications to review sandbox posts. The target audience of most of these challenges is code golfers like you, so anything you find unclear will probably be unclear to others.

If you think one of your posts needs more feedback, but it's been ignored, you can ask for feedback in The Nineteenth Byte. It's not only allowed, but highly recommended!

It is recommended to leave your posts in the sandbox for at least several days, and until it receives upvotes and any feedback has been addressed.

Other

Search the sandbox / Browse your pending proposals

The sandbox works best if you sort posts by active.

To add an inline tag to a proposal use shortcut link syntax with a prefix: [tag:king-of-the-hill]. To search for posts with a certain tag, include the name in quotes: "king-of-the-hill".

Get the Sandbox Viewer to view the sandbox more easily!

\$\endgroup\$
0

3548 Answers 3548

1
59 60
61
62 63
119
1
\$\begingroup\$

Gödel numbering of a string

Posted!

\$\endgroup\$
1
\$\begingroup\$

Compute the generalised XOR (posted)

\$\endgroup\$
2
  • \$\begingroup\$ \$(b - ((p + q) \bmod b)) \bmod b\$ can be simplified to \$-(p+q)\bmod b\$. \$\endgroup\$
    – Bubbler
    May 28 at 4:00
  • \$\begingroup\$ Yes, I was a bit worried about people getting confused about negative modulo, but I'll make that edit, it's probably easier to understand that way. \$\endgroup\$
    – xavc
    May 28 at 4:02
1
\$\begingroup\$

Construct a Modulo Multiplication Table

\$\endgroup\$
5
  • \$\begingroup\$ You could change it to be "consider a (n-1) x (n-1) ... and modulo n" (e.g. 5x5, taken modulo 6) so you don't have to ignore the last row and column \$\endgroup\$
    – pxeger
    Jun 2 at 13:47
  • \$\begingroup\$ @pxeger Thanks. Any suggestions for the title? \$\endgroup\$ Jun 2 at 13:48
  • \$\begingroup\$ short and simple: "Modulo tables"? \$\endgroup\$
    – pxeger
    Jun 2 at 14:00
  • \$\begingroup\$ Usually I think of quadrants as axis-aligned.Maybe "Modulo multiplication table", maybe throw in the word partial. \$\endgroup\$
    – qwr
    Jun 2 at 14:26
  • \$\begingroup\$ Also more test cases just for good measure \$\endgroup\$
    – qwr
    Jun 2 at 14:26
1
\$\begingroup\$

Salvage My Chemistry Homework

Introduction

I pulled an all-nighter yesterday doing my Chemistry homework. However, I now realize that was a terrible idea, because my handwriting was too sloppy and now I can’t distinguish the capital letters from the lowercase letters in my chemical formulas! This is a problem, because now the formulas can mean completely different things - for example, co could be cobalt or carbon monoxide. I want to know which of the formulas in my homework are ambiguous in this manner.

Since there are many formulas in the homework, I would like a program to help me so I don’t have to go over them all manually. Also, because the homework is due in an hour, I would like to have the code be as short as possible so I don't have to type as much.

Background

A chemical formula consists of one or more 'units'. Each 'unit' consists of a symbol of an element (the list of valid symbols is given in the input) optionally followed by a number greater than 1. For example, c, co and br2 are valid units (assuming that "c", "co" and "br" are in the list of symbols), while c0 and br2r are not valid units.

Input

The input consists of

  • The list of elements (all letters are lowercase)
  • A chemical formula with all letters in lowercase

Output

Output an integer, the number of ways there to separate the formula into individual units.

Test Cases

Input: ['c', 'o', 'co'], 'co2'
Output: 2 ('c o2', 'co2')

Input: ['c', 'h', 'o', 'ch', 'oh'], 'ch3cooh'
Output: 4 ('c h3 c o o h', 'c h3 c o oh', 'ch3 c o o h', 'ch3 c o oh')

Input: ['c'], 'c1'
Output: 0 (1 is not a valid subscript)

Input: ['a'], 'co2'
Output: 0 ("c" and "co" are not valid symbols)

Input: ['n', 'h', 'o', 'nh', 'no'], 'nh4no3'
Output: 4 ('n h4 n o3', 'n h4 no3', 'nh4 n o3', 'nh4 no3')

Standard loopholes apply. All reasonable input and output methods are allowed. This is , shortest code in each language wins.

Meta

  • Would it be better to output each way to split the formula or just the number of ways?
  • Any general wording improvements?
  • Anything I should be more specific about?
  • Suggested tags?
  • Given that l4m2's question has now been posted, should I remove this? (While the challenge is similar, there are some significant differences).

Credit to l4m2 for original idea

\$\endgroup\$
1
\$\begingroup\$

Closest Binary Fraction


For this challenge, a binary fraction is a fraction where the denominator is a power of two. Your program (or function) should take a rational number as input, and return the nearest binary fraction with a given denominator.

Input:

The first input will consist of a rational number. This can be represented by any of the following:

  • A floating point number, or built-in rational type
  • A fraction, either as a built-in rational or fraction type, or represented as a pair of numbers
  • A decimal string, in any chosen base

The second input is an integer \$d \ge 1\$, representing the denominator of the binary fraction.

Output:

Output must be able to represent the exact fraction. This means any of the input formats for fractions are valid, as long as they can reasonably represent any valid outputs.

You may choose if you want to simplify the output. Given the fraction 9 / 19 and the denominator 8, either 4 / 8 or 1 / 2 would be accepted.

Test cases:

Other:

This is . Unfinished and also a dupe of a newer challenge of mine

\$\endgroup\$
2
  • \$\begingroup\$ the output part seems.. unfinished? \$\endgroup\$
    – Razetime
    Apr 30 at 2:47
  • \$\begingroup\$ @Razetime It is, yeah. I was halfway through drafting this when I had to go work on something else, so I just left it here as a reminder to finish it :p \$\endgroup\$ Apr 30 at 3:21
1
\$\begingroup\$

Shuffle a subsequence

Thanks to rak1507 for the suggestion

"Random" in this challenge always refers to "uniformly random" - of all possible choices, each has an equal chance of being chosen. Uniform shuffling means that every possible permutation of the array has an equal chance of being chosen.

Given an array consisting of positive digits (123456789), select a randomly chosen, non-empty subsequence from the array, shuffle the elements and reinsert them back into the array in the former indices, outputting the result

For example, take L = [5, 1, 2, 7, 4]. We choose a random non-empty sublist, e.g. [5, 2, 4]. These are at indices 1, 3, 5 (1-indexed). Next, we shuffle [5, 2, 4] to give e.g. [2, 5, 4]. We now reinsert these into the list, with 2 at index 1, 5 at index 3 and 4 at index 5 to give [2, 1, 5, 7, 4].

You may also take the input as an integer or a string, and output it as such, or you may mix and match types.

This is , so the shortest code in bytes wins


Meta

\$\endgroup\$
1
\$\begingroup\$

Is it a Fischer random chess starting board?

\$\endgroup\$
2
  • \$\begingroup\$ Scoring? (code-golf?) Tags? (string chess decision-problem?) \$\endgroup\$
    – pxeger
    Jun 2 at 14:11
  • \$\begingroup\$ @pxeger yes. I always forget to put them in these posts \$\endgroup\$
    – qwr
    Jun 2 at 14:12
1
\$\begingroup\$

Task

You are an angry dude who manages the servers, but you are terrible in it. People often text you, and sometimes they ask why are the servers down - you got very annoyed!


Given a non-empty string. If it contains "server" and "down" (capitalization doesn't matter), you infinitely output the >:( face with a newline after. If it doesn't, you only output "Hi" once, because you are busy and angry.

Examples

server down -> >:( (infinitely)
server -> Hi
down -> Hi
serverANDLETTERSAFTERTHEWORD down -> >:( (infinitely), as it contains both server and down
Why are your servers down? -> >:( (infinitely)
sErVeR dOwN -> >:( (infinitely)
serverDowN -> >:( (infinitely)
serverd0wn -> Hi

Rules

  • This is , so the answer with shortest bytes wins.
  • These loopholes are, obviously, forbidden.
  • Standard code-golf rules apply.
  • Please specify the language you are using and the amount of bytes.
  • It would be great if you would put a link to a sandbox where your code can be ran in action, such as TIO.
  • Explaining your code is very welcomed.





Please comment down below if this is good enough or not, or if it's a duplicate (hope not)
\$\endgroup\$
3
  • 1
    \$\begingroup\$ I don't want to discourage you, but this seems a little trivial and uninteresting to me. Of course, that's just my opinion, and others may enjoy answering this challenge. For the output, would a an infinite list of characters be okay? \$\endgroup\$
    – rues
    Jun 6 at 18:20
  • \$\begingroup\$ It seems like combine two tasks together: 1. test if a string contains both two words or a not; 2. implement a truth machine \$\endgroup\$
    – tsh
    Jun 7 at 8:21
  • \$\begingroup\$ Damn, that question looks very similar \$\endgroup\$ Jun 7 at 8:23
1
\$\begingroup\$

Ambiguous Chemical Formula

\$\endgroup\$
4
  • 2
    \$\begingroup\$ This looks like an interesting challenge, but in its current form needs a lot more details (as do most of your sandbox posts - I'd recommend using a question template to make sure you've always included everything). For people who don't know chemistry, it's lacking crucial information: what is the exact syntax of a chemical formula, and how does one know it's ambiguous? \$\endgroup\$
    – pxeger
    Apr 22 at 16:28
  • 3
    \$\begingroup\$ I also think it would be more interesting if the element list were provided as input. Otherwise most of the actual golfing part of the challenge boils down to compressing the list (or finding a short built-in to output them) \$\endgroup\$
    – pxeger
    Apr 22 at 16:29
  • \$\begingroup\$ @pxeger Quite unrelated to chemical though, ClI and CLi both don't fell good \$\endgroup\$
    – l4m2
    Apr 22 at 18:23
  • \$\begingroup\$ I've edited this down to a stub now that it's been posted to save space \$\endgroup\$ Jun 7 at 1:01
1
\$\begingroup\$

Dr. Lamport's Unfinished Business

\$\endgroup\$
6
  • \$\begingroup\$ Also, I'd love some advice on how to specify what to do for some of the more "exotic" languages. e.g. C and such will obviously just use function-pointers… but I really don't want to alienate any Jelly golfers; those solutions are always my favorite! \$\endgroup\$ May 20 at 15:49
  • 1
    \$\begingroup\$ Even function-pointers in C wouldn't help afaik, because you can't dynamically construct the function. Maybe you should allowed output the source of G, or alternatively make R take m and pk and just function directly as G. \$\endgroup\$ May 21 at 7:48
  • \$\begingroup\$ @CommandMaster Thanks for the feedback! Would defining R(digest, pk) work? Such that it gets composed: R(φ(m), pk) -- (this would also align better with the way the original paper defined them.) \$\endgroup\$ May 21 at 14:08
  • 1
    \$\begingroup\$ I think it would work, and it's also good because it lets programming languages without the concept of a "function" participate \$\endgroup\$ May 21 at 14:15
  • \$\begingroup\$ @CommandMaster How's the current setup look? I'm certainly worried about accomodating languages without functions. Especially Jelly; I'm not really sure how it works, but I always enjoy those answers a lot! \$\endgroup\$ May 21 at 14:39
  • \$\begingroup\$ I've edited this down to a stub now that it's been posted to save space \$\endgroup\$ Jun 7 at 1:02
1
\$\begingroup\$

Write a netquine

There are a lot of online interpreters, for a lot of languages. For example, there's the ubiquitous TIO, but there are also smaller, more specific ones that others have made, such as interpreters for Vyxal, ngn/k, or Grok.

Your goal today is to write a netquine, or a program whose output, when used as a URL, leads to that program in an interpreter for that language. Essentially, if you have a link to your program, and the program outputs that link, it's valid.


For example, say you write a Vyxal program λλλλ. In the Vyxal interpreter, if you press the button to generate a permalink for this program, you get:

http://lyxal.pythonanywhere.com?flags=&code=%CE%BB%CE%BB%CE%BB%CE%BB&inputs=&header=&footer=

However, due to the way that the interpreter processes links, you could also leave out the blank fields:

http://lyxal.pythonanywhere.com/?code=%CE%BB%CE%BB%CE%BB%CE%BB

You can even replace the HTML character codes with the actual characters:

http://lyxal.pythonanywhere.com/?code=λλλλ

All of these are valid links to the Vyxal program λλλλ, and therefore would all be valid outputs for a netquine.


  • Obviously, Standard loopholes are forbidden.

  • Even though URL shorteners are forbidden by default, shortening a URL, like in the example, is fine, as long as it still leads to the right place.

  • For the purposes of this challenge, a web browser may be considered an "interpreter" for HTML and/or CSS. This means you might be able to do something along these lines.

  • Usual rules apply, as do usual rules.

  • This is , so the shortest answer in bytes wins!


Meta

  • Is this question clear enough?

  • Should I change the scoring to accommodate the potential differences in URL length? (lengths of domain names, different ways of generating links, the ability to shorten links, etc.)

  • Any other feedback?

  • This question is quite similar, but it only allows for TIO languages, and it doesn't allow any sort of shortening, even if the link is still valid (there was a comment asking about that), so I would consider it to be related, but not a dupe.

\$\endgroup\$
2
  • \$\begingroup\$ Interesting idea, but you won't get many answers because most online interpreters use an external compression algorithm to shorten the hash (and then base64 it), which is quite lengthy to implement from scratch. Note that the existing answers on the linked challenge all use a gzip or zlib builtin. \$\endgroup\$
    – Bubbler
    Jun 9 at 23:12
  • \$\begingroup\$ You should provide a list of valid online interpreters, as someone could just make a.com lead to an interpreter for a program in some language that outputs https://www.a.com. Also, please define "in" in the sentence "leads to that program in an interpreter". Does the HTML of that page simply have to contain the program? \$\endgroup\$
    – Makonede
    Jun 9 at 23:40
1
\$\begingroup\$

Encode USB packets

Posted

\$\endgroup\$
2
  • \$\begingroup\$ Is the fifth example correct? Shouldn't 11111111 (8 ones) be JJJJJJKKK? If not, how is 11111111 distinguished from 11111101? (This also applies to your other challenge). \$\endgroup\$ Jun 6 at 19:33
  • \$\begingroup\$ @Spitemaster yep, you're right - fixed. \$\endgroup\$
    – pxeger
    Jun 9 at 14:33
1
\$\begingroup\$

A Cat’s Game to Claim

Posted

\$\endgroup\$
1
  • \$\begingroup\$ Any catchy name suggestions also appreciated! "Proclaim" rhymes as well, but I can't get it to roll of the tongue. \$\endgroup\$
    – AviFS
    Jun 12 at 1:15
1
\$\begingroup\$

Build a raw string


In this challenge you will write code which, given a single line string as input, can output the raw version.

Full spec

The input will be a string like ab\n or ab\u0065e\\ or e\'543. These strings should be converted to a raw version which uses the backslashes.

This probably sounds confusing, so here is a more complete explanation.

The most common form of strings used is simply a quote (one of ', " or the backtick), followed by a sequence of unicode characters, then a closing quote which corresponds the opening one. "abc" or 'THIS is a St%ing', for instance. Note that this is not the case in all languages, for instance most golfing languages do not represent strings in this manner.

However, it's not that simple. Backslashes allow for much more complex strings, and their importance in allowing certain characters means a string like 'ab\' is invalid as the backslash "escapes" the quote. Making a backslash appear in a string requires \\ instead.

Also, quotes can be escaped via \', \" or \`. A sequence like \\" should become \" while \"\ is invalid.

The newline and tab are both important, and so most mainstream languages have the \n and \t which are actually 1 character each, representing the newline and the tab, respectively. So \\t\"\n becomes

\t"

(note the trailing newline)

To add an extra layer of usability for "exotic" characters, it is possible to use a unicode escape: of the form \uxxxx where each x is one of 0123456789abcdef. This transforms into the unicode character whose codepoint is equal to the base-10 equivalent to the hexadecimal number xxxx. For instance \u0031 becomes 1 because 31 in hexadecimal is 49 in base 10, and 1 has a codepoint of 49.

Your input string may contain any of these escapes, and you may assume that \ is followed only by the backtick or any of \'"tnu. The unicode escape sequence is guaranteed to be valid, and escapes as well.

You should output this string after the transformations described.

Worked example

Input string:

ab%\u0035e\n\n\t5\'e\\54c%&\u0095

The first three characters remain identical.

\u0035 becomes 5 because 35 in hex -> 53 in base 10, and String.fromCharCode(53) === '5'.

\n\n\t becomes two newlines followed by a tab.

5 remains 5.

\' becomes ': escaped by backslash

e remains e.

\\ becomes \.

54c%& stays the same.

\u0095 is an unprintable with codepoint 149.

Result:

ab%5e

    5'e\54c%&<unprintable>

(<unprintable> is an actual unprintable character) Note that functions capable of evaluating such strings are disallowed. To be clear, your program must handle the syntax described above, and not any other format.

Other test cases:

Input Output
\'\" '"
\\"\' \"'
\\u0031\u0031\\t \u00311\t
\\\\\\\\\n \\\\<NL>

[<NL> is newline]


Meta:

  • Duplicate? Couldn't find anything.
  • Any clarifications needed?
  • Better title / tags?
\$\endgroup\$
13
  • \$\begingroup\$ A bit unclear to me: Do we need to handle the string syntax of the answer's implementation language, or the syntax you specify here? If that of our own language, then what if our own language has few special characters? (E.g. in the APL family, it is common that the only unusual character in a string is the quote, which must be doubled.) \$\endgroup\$
    – Adám
    Jun 14 at 11:54
  • \$\begingroup\$ "eval functions or similar are not allowed." but what constitutes being similar to eval? What about functions in the JSON object of JavaScript? What about library functions to do this? Does APL's built-in ⎕JSON utility function constitute an eval, though the language's real eval is ? \$\endgroup\$
    – Adám
    Jun 14 at 11:56
  • \$\begingroup\$ @Adám In some calculator questions, the OP (wisely) decides to disallow eval functions. Basically these functions. In essence, from my point of view, if your language is capable of evaluating these strings, then the function that does them is disallowed. However since APL's eval seems different, you can use them. Also in response to the one asking about string syntax, you must handle the one in the question, otherwise it gets a little unfair. \$\endgroup\$
    – user100690
    Jun 14 at 12:06
  • \$\begingroup\$ Will the input include surrounding quotes? \$\endgroup\$
    – Adám
    Jun 14 at 12:56
  • \$\begingroup\$ @Adám Nope. (filler) \$\endgroup\$
    – user100690
    Jun 14 at 13:01
  • \$\begingroup\$ Then do we have to handle unescaped ' and/or "? \$\endgroup\$
    – Adám
    Jun 14 at 13:03
  • \$\begingroup\$ @Adám You still have to handle them, just not surrounding the entire string. \$\endgroup\$
    – user100690
    Jun 14 at 13:08
  • \$\begingroup\$ Suggested test cases: '" and \'\" \$\endgroup\$
    – Adám
    Jun 14 at 13:09
  • \$\begingroup\$ @Adám You know what, the idea of not having surrounding quotes is a little confusing, so I'll remove that rule. That means that your suggested test cases are identical to each other, so I'll add the latter to demonstrate. Also will add about five more cases. \$\endgroup\$
    – user100690
    Jun 14 at 15:28
  • \$\begingroup\$ "if this string were to be enclosed in quotes in your language and parsed as an expression, no errors should show." makes it sound as if my language matters. \$\endgroup\$
    – Adám
    Jun 15 at 15:35
  • \$\begingroup\$ @Adám I'll change that \$\endgroup\$
    – user100690
    Jun 15 at 15:37
  • \$\begingroup\$ Am I allowed to use \n in my code to insert a newline, or is this considered "functions capable of evaluating such"? \$\endgroup\$
    – Adám
    Jun 15 at 15:58
  • \$\begingroup\$ Let us continue this discussion in chat. \$\endgroup\$
    – user100690
    Jun 15 at 15:59
1
\$\begingroup\$

COBOL(lite) interpreter.

Your task is to code an interpreter for COBOL. This interpreter should use the free format, and cut a few corners, as long as a few test scripts will run on it.

Notice; here I say interpreter, as that is the general direction of my inclination, however if it may save bytes, a compiler would also count.

Try to:

  • Golf your code as much as possible
  • Remove some parts of COBOL (But ask here first to see if they make sense to remove.
  • Accept both CAPITAL and lowercase

Avoid:

  • Standard Loopholes
  • Don't remove too many parts of COBOL. I think that IDENTIFICATION DIVISION. could be ignored, but DATA DIVISION. is probably necessary.

If there is a thing that you would like to clarify on, ask in the comments. I think that this question still could stand with improvement, before it can be added to the main cg se.

\$\endgroup\$
3
  • 3
    \$\begingroup\$ You need to be a lot more specific with what you want implemented. This can range from syntax requirements to required keywords and function. You need to specify anything and everything yo want in a possible golfed interpreter. \$\endgroup\$
    – Razetime
    Jun 15 at 10:39
  • \$\begingroup\$ I recommend checking questions tagged under interpreter on the main site for guidance. \$\endgroup\$
    – Razetime
    Jun 15 at 10:40
  • \$\begingroup\$ @Razetime I didn't have a lot of time to make this post, but that is why it is in the sandbox. I know that I will need to improve it a fair bit, before putting it on the main code golf se \$\endgroup\$
    – smarnav
    Jun 15 at 23:17
1
\$\begingroup\$

Implement a feature-rich calculator


We have a few many calculator challenges on this site, but none seem to require a complex one with many different features.

This challenge is about emulating my old Casio calculator, except with no digit limit. Features are:

  • basic (start with a single operation, then input the next operation)
  • all clear (AC) which starts a new "calculation". No need for C.
  • errors for dividing by zero (otherwise there should be no error)
  • memory: M to set the memory to current number, MC as an inline expression representing the memory, and M- to get rid of the memory.

This should operate either on STDIN (prompt for commands each time) or in a function (taking in either a list of commands or a newline-separated string with commands).

Spec

Your solution should be able to handle floating-point numbers, hence float division should be used regardless of context.

When first run, the calculator will be "ON". It is guaranteed that input syntax is valid.

The chain of commands begins with a number operation number-like structure, where number can be replaced with MC to represent a reference to the memory. For instance 5 * 5 or 6/MC. Whitespace can be ignored or forbidden, that can be your call, but whatever it is you may assume syntax is valid.

Then, if your submission is a full program, it should output the result of that operation, or if a function, append that to the end of a list.

Subsequent operations can either be:

  • M this sets the memory to current number. Memory can only ever hold a single number at a time.
  • AC. Clears the current calculation and resets the number. Memory, however, will still be accessible.
  • M-. (M minus) This clears the memory, and calls to MC will never be given when there is nothing in the memory.
  • operation number where operation is + - * or /, and number is a number which can be replaced with MC.
  • Off. The command Off stops the program, and it is reasonable to assume that no commands will be passed afterwards.
  • - (minus sign) negates current number. Outputs the new number.

Division by zero should trigger an error and restart the calculation. AC, M, MC and M- should output nothing. Memory can be overwritten.

Note that operation number operation number will never be given, and that all operations will be performed on non-negative numbers. (we have the negate operator for negative numbers)

Also note that this challenge is less about parsing / processing a string input and more about responding to different kinds of input.

Example

As a full program, but note that a function will have to take in an array of commands. There is a > before each command, but in your version, it is not required.

> 5*5
25
> *4
100
> /200
0.5
> /0
ERROR (or some form of ERROR)
> 15*6
90
> M
> +MC
180
> AC
> 7-MC
-83
> -
83
> M-
> +5
88
> Off

Meta: is it a duplicate? any clarifications needed?

\$\endgroup\$
3
  • \$\begingroup\$ Maybe use the [interactive] tag, and make it clearer that this is less parsing/processing a given string input and more responding to inputs as given \$\endgroup\$ Jun 18 at 14:16
  • \$\begingroup\$ The "error" should not terminate the program, right? Would it be OK to, on error, print anything that is distinguishable from numeric output (say "X", "Inf", or "NaN"), or print something different across errors (e.g. "Inf" for 1/0, "NaN" for 0/0)? \$\endgroup\$
    – Bubbler
    Jun 25 at 1:40
  • \$\begingroup\$ @Bubbler Yes, that's fine. You can also output Inf for 1/0 and NaN for 0/0 like you mentioned. \$\endgroup\$
    – user100690
    Jun 25 at 5:10
1
\$\begingroup\$

KotH - Floating Point Prisoners Dilemma

\$\endgroup\$
8
  • 1
    \$\begingroup\$ Looks intriguing! How does each round work though? So one is coop, zero is defect, but what do the floats in between do? Look forward to hearing! \$\endgroup\$
    – AviFS
    Jun 11 at 2:33
  • \$\begingroup\$ It uses an equation from the wonderful fellows at math.stackexchange.com to determine what the in-between states should do. I'm bad at explaining, so the equations in my (probably horrible) code are in the play function of the controller class in controller.py \$\endgroup\$
    – 4D4850
    Jun 11 at 2:46
  • \$\begingroup\$ @4D4850 Would you link to the question if you have it still? \$\endgroup\$
    – math
    Jun 11 at 17:46
  • 1
    \$\begingroup\$ It was actually closed, but here it is: math.stackexchange.com/questions/4152360/… \$\endgroup\$
    – 4D4850
    Jun 11 at 18:27
  • \$\begingroup\$ Understood it now, good challenge! \$\endgroup\$
    – math
    Jun 12 at 8:39
  • \$\begingroup\$ Should I post the question? \$\endgroup\$
    – 4D4850
    Jun 17 at 19:08
  • \$\begingroup\$ Posted on Main: codegolf.stackexchange.com/questions/229926/… \$\endgroup\$
    – 4D4850
    Jun 17 at 22:14
  • \$\begingroup\$ I've edited this down to a stub now that it's been posted to save space \$\endgroup\$ Jun 18 at 14:33
1
\$\begingroup\$

Encode a Lenguage

\$\endgroup\$
1
\$\begingroup\$

Outputting Blum Integers

\$\endgroup\$
2
  • \$\begingroup\$ I'd suggest using standard sequence rules (answers can choose \$n\$th, first \$n\$ or infinite output). I've also edited the challenge a bit, feel free to revert any changes you dislike \$\endgroup\$ Jun 18 at 14:29
  • \$\begingroup\$ @cairdcoinheringaahing thanks for the suggestion, i'll apply it \$\endgroup\$
    – user100752
    Jun 19 at 9:14
1
\$\begingroup\$

What's missing

\$\endgroup\$
5
  • \$\begingroup\$ This is probably too easy \$\endgroup\$
    – rak1507
    Mar 22 at 12:14
  • \$\begingroup\$ Some languages may have built-in Bag / MultiSet / Counter collection type. And a simple - operate or something similar may get correct result. Will you allow answers use such type of values as input / output? \$\endgroup\$
    – tsh
    Mar 23 at 1:41
  • \$\begingroup\$ @tsh I guess? It doesn't really make a difference for the languages that have creative solutions. It's code-golf, which means answers generally don't get accepted. \$\endgroup\$
    – emanresu A
    Mar 23 at 8:22
  • \$\begingroup\$ ruby does have array-array, but it wont work for arrays with similar items missing. I feel this is easy, but it can work as a codegolf challenge \$\endgroup\$
    – user100752
    Jun 22 at 9:24
  • \$\begingroup\$ @EliteDaMyth Honestly, I'm really surprised this doesn't already exist. \$\endgroup\$
    – emanresu A
    Jun 22 at 9:33
1
\$\begingroup\$

Do these points approximately make up a regular n-sided polygon?

The input is a set of integer coordinates. Take them in whatever form you want, e.g. a list of tuples or just a plain list. The question is: can you draw an n-sided regular polygon with vertices at the specified coordinates? n being the amount of coordinates provided. Answer truthy if yes, falsy if no.

The level of precision is in integers. That means you will have to round both the x and y values of the ideal polygons to the nearest whole values. Here's an example diagram:

example

The triangle has points at approximately (5,5), (4,9), (1,6). You can draw a regular triangle to fit these coordinates, even if the actual points of the triangle are not at quite the same coordinates.

The same goes for the pentagon. The coordinate set (3,1), (7,1), (8,5), (5,8), (2,5) makes up a regular pentagon-ish, accurate to whole values.

Any set of coordinates can make up an infinite amount of approximate n-gons. You answer truthy if there's at least one such n-gon. One way of thinking about it is to consider a unit square around each of the points, then determine if you can stretch a regular polygon such that every point falls within the squares.

Further ground rules:

  • The coordinates will always be positive.
  • They do not have to be in any particular order.
  • Coordinate values will never exceed 1000.
  • There will not be more than 10 coordinates.

Examples:

(5,5), (4,9), (1,6) -> true
(3,1), (7,1), (8,5), (5,8), (2,5) -> true
(1,1), (1,1), (2,1) -> true
(1,1), (2,1), (130,1) -> false

(42,42) -> true

(according to Wikipedia, a monogon is still a polygon, meaning any input is valid.)

(0,1), (5,10) -> true 

(any two points make up a regular bigon so this should work for any coordinates)

(1,3), (1,3), (1,3), (1,3) -> true

(any set identical coordinates will always match an n-gon that's tiny enough to have all of its coordinates rounded to the same integer).

\$\endgroup\$
4
  • \$\begingroup\$ I just found this question. It seems to be the same question but it is closed with no answers, and I think my requirements are clearer. \$\endgroup\$
    – KeizerHarm
    Jun 16 at 12:12
  • 1
    \$\begingroup\$ I don't think the current explanation addresses the question "How are we supposed to tell if the polygon is regular without knowing its exact coordinates?" I feel like there's some way to gradually constrain polylines, but not quite sure. Also, I think "One way of thinking about it is to consider a circle of half a unit around one of the points" should be "... consider a unit square around one of the points" instead. What is the expected output of (0, 0), (0, 2), (1, 1), (2, 0) (which is borderline square with (-0.5, -0.5), (-0.5, 1.5), (1.5, 1.5), (1.5, -0.5))? \$\endgroup\$
    – Bubbler
    Jun 23 at 0:15
  • \$\begingroup\$ @Bubbler Thanks for the comment. You are correct about the circle being a square! My mistake. And yes, that would be truthy. \$\endgroup\$
    – KeizerHarm
    Jun 23 at 8:59
  • \$\begingroup\$ @Bubbler regarding "how are we supposed to tell if polygon is regular" that's asking for an example answer strategy, not part of the question, is it? The question is whether there is any regular polygon where each point falls in those unit squares. That's almost an optimisation problem. It's not impossible, all the required information is there. But I could try to come up with one approach - but answers should not feel obligated to copy that approach because for this problem there would be many solution strategies. \$\endgroup\$
    – KeizerHarm
    Jun 23 at 9:03
1
\$\begingroup\$

Alteration of a challenge that was considered too close to an older quasi-duplicate.

ISO Computus

? ?

Your task is to calculate
from a signed integer year number provided as sole input
the integer number of the calendar week of Easter Sunday as sole output
with as few variables and as few arithmetic operations as possible.

Easter Sunday is determined according to the Gregorian Computus (as used in the all Western Christian churches, including the Catholic Church, and some others), for which there are several equivalent algorithms available, see Computus, Gent, Stockton and ESTRALGS.TXT. There are 35 possible month-day dates, from 22 March to 25 April.

The week of the year is specified by international standard ISO 8601-1: All weeks start on Monday; there are no partial weeks; the first week of the year has 4 January (or the first Thursday) in it. For more information see the Mathematics of the ISO 8601 Calendar. There are just six possible results, weeks 12 through 17. W17 will hardly ever occur – it has never occurred since 1583 at least.

Restrictions

  • You may use all basic arithmetic operators, including addition, substraction, multiplication, division, integer division, modulo, rounding (with integer flooring and ceiling), exponentiation, logarithms.
  • You may not use library functions that implement an algorithm for the Computus, like PHP's easter_date(), EasterSunday[year] from the Mathematica Calendar package or Easter from the R package timeDate.
  • You may not use other predefined date functions, e.g. to find the day of the week or the week of the year.

Ranking score

Your solution must be an algorithm implemented as an executable script in a programming language of your choice.

  • Each numeric variable and constant, whether used by name or literal value, counts as 1 point.
  • Each basic operation, as defined above, counts as 1 point.
  • Each definition of a custom function or method counts as 5 points.
  • Each call of a custom function or method counts as 1 point.
  • Imports and similar initializations required by the programming language are ignored.
  • Neither reading the input value and returning or printing the output value nor eventual variables for storing their value are counted.
  • The answer with the fewest points wins.
  • TBD…

Example:

function CFAQEaster(year) { // Calendar FAQ
                       // 5 points for Div() function
  var G, C, H, I, J, L // 6 points for variables
  G = year % 19        // 2: 1 point for modulo, 1 point for 19
  C = Div(year,100)    // 2: 1 point for function call, 1 point for 100
  H = (C - Div(C,4) - Div(8*C+13,25) + 19*G + 15) % 30 
    // 16: 8 points for operations, 2 points for function calls, 6 points for additional numeric values
  I = H - Div(H,28)*(1 - Div(29,H+1)*Div(21-G,11)) 
    // 13: 5 points for operations, 3 points for function calls, 5 points for additional numeric values
  J = (year + Div(year,4) + I + 2 - C + Div(C,4)) % 7
    // 6 points for operations, 2 points for function calls, 2 points for additional numeric values
  L = I - J            // 1 point for substraction
  EasterMonth = 3 + Div(L+40,44) 
    // 6: 2 points for addition, 1 point for function call, 3 points for additional numeric values
  EasterDay = L + 28 - 31*Div(EasterMonth,4)
    // 6: 3 points for operations, 1 point for function call, 2 points for additional numeric values
  return {Y:year, M:EasterMonth, D:EasterDay} 
    // 63 points in total
}

Contemporary example input: output

  • 2001: 15
  • 2002: 13
  • 2003: 16
  • 2004: 15
  • 2005: 12
  • 2006: 15
  • 2007: 14
  • 2008: 12
  • 2009: 15
  • 2010: 13
  • 2011: 16
  • 2012: 14
  • 2013: 13
  • 2014: 16
  • 2015: 14
  • 2016: 12
  • 2017: 15
  • 2018: 13
  • 2019: 16
  • 2020: 15
  • 2021: 13

The return value must be correct for a complete cycle of input years, i.e. from 1 through 5700000, even though the original event that is celebrated did not occur before 30 (if ever at all) and the Gregorian Calendar was not used before late 1582, i.e. the proleptic Gregorian Calendar applies.

Resources

\$\endgroup\$
2
  • \$\begingroup\$ I believe this challenge is better done as alorithm-golfing rather than counting characters or bytes. However, I'm not sure at all how to award a score. The rules above are very much up to bikeshedding. \$\endgroup\$
    – Crissov
    Jul 1 at 10:46
  • \$\begingroup\$ The main problem with any kind of scoring like this is that many programming languages have very different set of built-in operations and features. How many points for using a mapping or reduction on an array? How many points for swapping two numbers on the stack? How many points for using the 1 command in Hexagony, which computes x>=0?x*10+1:x*10-1? A better way would be to fix the language and a very strict set of features available, like "use Python 3, only +-*/ operators on numbers are allowed". \$\endgroup\$
    – Bubbler
    Jul 14 at 7:28
1
\$\begingroup\$

Days of the Week

\$\endgroup\$
1
  • 1
    \$\begingroup\$ A few suggestions: perhaps add the date tag, and say something like Output "WEEKDAY" if there are 5 dates, all from Monday-Friday, "WEEKEND" if there are two dates: Saturday and Sunday, and "ALLDAYS" if the input contains seven days all from Monday-Sunday. \$\endgroup\$
    – user100690
    Jun 28 at 12:46
1
\$\begingroup\$

Generate the shortest regex to match these but not those

\$\endgroup\$
3
  • \$\begingroup\$ A possible short program just enumerates all the regexes. This might not be feasible, and you may not be able to score submissions at all... \$\endgroup\$
    – Trebor
    Jun 28 at 17:34
  • \$\begingroup\$ @Trebor I don't quite understand what you mean by this. \$\endgroup\$
    – user197974
    Jun 28 at 18:13
  • \$\begingroup\$ If you just write a program and tests, one by one, all the regexes, and outputs the first one that works, this would be a valid program. But to give it a score, you would probably need days, if not years. \$\endgroup\$
    – Trebor
    Jun 29 at 8:06
1
\$\begingroup\$

Solve Fermat's Last Theorem with matrices

Fermat's Last Theorem states that there is no such triple of positive integers \$(x, y, z)\$ such that

$$x^n + y^n = z^n$$

for all integers \$n > 2\$. However, this conjecture does not hold for integer matrices:

$$\begin{align} & \left( \begin{matrix} 1 & 3 \\ 0 & 1 \end{matrix} \right)^3 + \left( \begin{matrix} -1 & 0 \\ 1 & -1 \end{matrix} \right)^3 \\ = & \left( \begin{matrix} 1 & 9 \\ 0 & 1 \end{matrix} \right) + \left( \begin{matrix} -1 & 0 \\ 3 & -1 \end{matrix} \right) \\ = & \left( \begin{matrix} 0 & 9 \\ 3 & 0 \end{matrix} \right) \\ = & \left( \begin{matrix} 0 & 3 \\ 1 & 0 \end{matrix} \right)^3 \\ \end{align}$$

You are to take two \$k \times k\$ integer matrices \$A\$ and \$B\$, and a positive integer \$n \ge 2\$, and output a \$k \times k\$ integer matrix \$C\$ such that

$$A^n + B^n = C^n$$

You may output any such \$C\$. You may also take \$k\$ as an input, and you may assume that \$k \ge 2\$

This is , so the shortest code in bytes wins

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Is it guaranteed that A, B, n are given so that such a C exists? Why not just make a challenge about taking the nth root of a given matrix (which is arguably the hardest part of the task)? \$\endgroup\$
    – Bubbler
    Jul 14 at 6:54
1
\$\begingroup\$

Chess Squad March

Chess Squad March

\$\endgroup\$
1
\$\begingroup\$

I'm a lizard, cut here!

\$\endgroup\$
2
  • \$\begingroup\$ I think you mean n≥2 \$\endgroup\$
    – xnor
    Jul 21 at 5:05
  • \$\begingroup\$ @xnor sure thanks \$\endgroup\$
    – Domenico
    Jul 21 at 5:11
1
\$\begingroup\$

How effective is compression?

Vyxal has a very simple format for compressed integers: a base-255-encoded string of characters wrapped in «.

With base 255, the number to be compressed needs to be 10000 or greater for the compression to have an advantage against ordinary numbers, as with 1 character, you can represent at most 3-digit numbers (up to 254), which ties with 1 + 2 (for the «) = 3 bytes; with 2 you can represent up to 65024 in 4 bytes, beating 5 for n ≥ 10000.

But what if it was another base?

If it were base 99, it would take 3 bytes, since you can only express up to 9800 with 2, which isn't any better than expressing the number. But with 3, you can reach 970299, which has 6 digits

Your challenge is to take a positive integer \$b\$ 11 or greater, and find the smallest number \$n\$ such that \$log_b(10^n) + 2 < n\$. You may output \$10^n\$ instead.

This is , shortest wins!

Testcases coming soon.

\$\endgroup\$
3
  • \$\begingroup\$ Be sure to include the testcases for n=100 and n=1000. The correct answer is floor(1+2/(1-1/log10(n))), but the wrong answer ceil(2/(1-1/log10(n))) gives identical results for all other n. (The correct answer can be also wrong due to floating-point imprecision. Is this allowed?) \$\endgroup\$
    – Bubbler
    Jul 29 at 7:45
  • \$\begingroup\$ @Bubbler Ok, will do. Shouldn't be wrong as it's taking a single logarithm. \$\endgroup\$
    – emanresu A
    Jul 29 at 7:56
  • \$\begingroup\$ Straightforward Python fails because of 1/3, not single logarithm btw \$\endgroup\$
    – Bubbler
    Jul 29 at 8:13
1
\$\begingroup\$

KOTH: Prisoners Dilemma Single Elimination

I was playing around with this and found some interesting strategies. This is sufficiently different from this post because you are playing different bots each game, instead of repeatedly playing the same ones.

How the game works

The basic game is the prisoners dilemma. A bot and it's opponent may choose to defect (1) or cooperate (0). Defecting gives you 1 point, and if your opponent cooperates you gain 2 points. If you defect and your opponent cooperates this adds up to a total of 3 points.

In every round, you play against every opponent (including yourself) in random order. The bot does not know who it is playing against. This process repeats 100 times to form a single round, after each round the player with the lowest score is eliminated, and the scores of the rest of the players is divided are 2. Round continue until either every remaining player always defects, or always cooperates, at which point the best player is decided by their score. (After this points scores converge quite fast, so to avoid noise deciding the winner I measure at the point with the most pronounced differences)

Input

Your bot does not know who it is playing against, not even if it is playing against itself. However, the bot does know what its opponent decided on its last 3 games, and for each of those 3 games what the opponent's opponent decided. Basically, the input will look like this:

[
    [0, # Three games ago, your opponent decided not to defect
        [1, # Your opponent's opponent from three games ago defected three games before that
         0,
         0],
    [1,  # Your opponent defected in their second to last game
        [1, # Your opponent's opponent from two games ago defected the 3 games before that
         1,
         1]
    [0, [0, 0, 0]
]

(With 1 and 0 replaced with True and False)

Any of these players might be the same. Your opponent's opponent could actually be your current opponent if they played against themselves last, or it could be yourself if you played against eachother twice in a row. (Unlikely, but possible at the end of a round)

Your bot does not receive what it's own last decisions are but is free to keep that info in it's own state.

Output

You can output any value, if bool(output) == True your bot defects, otherwise your bot cooperates.

Example Bots

Nice bot never defects:

class NiceBot:
    def get_choice(self, last_3_games):
        return 0 # Any False-y value will work, for example False or an empty string

Mean bot always defects:

class MeanBot:
    def get_choice(self, last_3_games):
        return 1

Majority bot chooses whatever it's opponent chose the most often:

class MajorityBot:
    def get_choice(self, last_3_games):
        choices = [i[0] for i in last_3_games]
        return max(set(choices), key=choices.count)

Bandwagon bot will avoid decisions by chosing what the bots opponents chose the most.

class BandwagonBot:
    def get_choice(self, last_3_games):
        total_true = sum(
            i[1].count(True)
            for i in last_3_games
        )

        return total_true >= 5

Bandwagon bot will only play if the number of bots is even.

Other specifics

A bot may implement a reset() method that will be called every time an opponent is eliminated. For bots that keep state this might be a good time to clear it.

Other rules:

  • You are not allowed to use any form of randomness.
  • You are not allowed to use IO
  • Standard loopholes apply, in particular no exploiting the controller
  • Standard library imports are allowed for simple functionality that you could implement yourself without breaking any of the other rules eg itertools, functools but not for example sys, time or asyncio. Ask if you are not sure. Definitely no attempts to import parts of the controller or other bots.

Controller

I have a controller written but I would need to simplify it a little to remove some parts that have become redundant. Don't want to put too much effort into making it presentable if it is not a suitable challenge.

\$\endgroup\$
1
\$\begingroup\$

Train A Single Perceptron

You must write a program that will train a perceptron to simulate the things described in the test cases section of this question.

What is a perceptron and how to train it

A perceptron takes inputs and returns 1 if the dot product of the inputs and its list of weights is greater than or equal to 0 and it returns 0 otherwise. To take the dot product, multiply the elements of the list and take the sum. In other words, your program must find the line (w⋅x=1) that linearly separates the two possible classes of outputs. Example: dot product of [0,2] and [8,3] is 0+6, which equals 6. Your job is to return the list of weights that will make the perceptron return the correct outputs for a list of inputs 100% of the time.

See these links for more information:

https://desmos.com/calculator/d2nryjmw2t

https://towardsdatascience.com/perceptron-learning-algorithm-d5db0deab975

Inputs

The program should take as input training inputs and training labels. Each list of inputs corresponds to exactly one output. This data must be used to train the perceptron.

Outputs

The program should output the final list of weights as integers. The dot product of these weights and any valid input to the perceptron should be the correct output of the perceptron 100% of the time.

Training

How you train the perceptron is up to you. To train a perceptron to simulate an AND gate, your program should take training inputs and the correct output for each list of inputs. In the case of the AND gate, your program would take [[0,0],[0,1],[1,0],[1,1]] as the list of training inputs and [0,0,0,1] as the list of what the neuron should output given each input. The program should then output a list of weights, which will output the correct output when used in a perceptron.

Test Cases

  1. AND gate:
    • Input List: [[0,0],[0,1],[1,0],[1,1]]
    • Correct Outputs: [0,0,0,1]
    • Your program should output a list of two weights that will solve the problem of behaving like an AND gate when used in a perceptron.
  2. Only outputs 1 if the third input is 1:
    • Input List: [[0,0,0],[0,0,1],[0,1,0],[0,1,1],[1,0,0],[1,0,1],[1,1,0],[1,1,1]]
    • Correct Outputs: [0,1,0,1,0,1,0,1]
  3. OR gate:
    • Input List: [[0,0],[0,1],[1,0],[1,1]]
    • Correct Inputs: [0,1,1,1]
  4. Make up your own test case, and try to make it interesting!

Scoring

This is code golf, so the shortest program in bytes wins.

\$\endgroup\$
3
  • \$\begingroup\$ Could you maybe add a note that this task is to find a line (\$w \cdot x=1\$) that linearly separates the two classes? That might make the challenge more accessible to people unfamiliar with perceptrons. As a side note, the decision are of linear perceptrons can be visualized quite nicely: desmos.com/calculator/d2nryjmw2t \$\endgroup\$
    – ovs
    Aug 1 at 10:51
  • \$\begingroup\$ I guess you changed the threshold to \$1\$ to avoid the bias? I'm not quite sure this works because now the origin \$(0,0)\$ is always classified as \$w\cdot 0=0\$. I think just going with the usual definition should be fine. \$\endgroup\$
    – ovs
    Aug 1 at 10:55
  • \$\begingroup\$ Ok, I updated it. \$\endgroup\$ Aug 4 at 2:11
1
59 60
61
62 63
119

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .