Here I asked about implementation or a document/paper with enough details to implement an algorithm. Is it off-topic?


2 Answers 2


Let's break it down:

  • Asking for an open-source implementation is off-topic. Asking for an open-source implementation to do X is equivalent to asking for a software package to do X. They're both off-topic here.

    See https://cs.meta.stackexchange.com/a/31/755 for a relevant policy statement: "[asking] for a complete software package [...] is definitely out of our mandate as the underlying problem/task is not related to computer science".

  • Asking for an algorithm for some well-defined task certainly is on-topic here. That's our bread and butter.

  • Asking for papers that describe an algorithm for some well-defined task is on-topic.

    A note of advice: in my opinion, it's probably more useful to ask for an algorithm to do X, rather than asking for a paper that describes an algorithm to do X. If you ask for an algorithm to do X, you might get back a paper reference, or you might get back an algorithm even if there is no paper. If you ask for a paper, you've ruled out an entire class of useful answers. But either one is in scope and on-topic.

  • Asking for papers that describe how to implement an algorithm for this task feels borderline to me. It could be OK -- or it might be problematic. The risk is that there may not be any paper that describes how to do an entire implementation, including all relevant implementation issues. And many implementation considerations tend to be a matter for Stack Overflow; we focus on the science (e.g., algorithms, concepts, fundamental techniques, rather than programming language specific concerns).

    My advice would be to start by asking for an algorithm for the task; then, if you run into any specific issue with implementing the algorithm, ask another question about that specific issue, either on Stack Overflow or here, depending on whether the issue has to do with fundamental computer science concepts or has more to do with programming/coding. Asking for implementation advice before you even have an approach or algorithm seems to me like it's getting ahead of things. But this is just friendly advice.

  • Asking for "X or Y" where X is off-topic and Y is on-topic is not good practice and risks having your question closed as off-topic. Don't ask for something you know is off-topic here; remove all the off-topic requests from your question, and focus only on the part that is on-topic.

In summary, to make your question on-topic, you'll need to remove the part of your question that asks for an open-source implementation.

If you want to rehabilitate the question to be on-topic here, for your sake, I'd recommend that you focus on asking for an algorithm for power graph analysis/visualization, rather than on asking for a document or paper -- but that is your choice.

Note: In this answer I'm focusing only on what's on-topic. Even if the question is edited to be on-topic, it will still need to meet our other requirements, including being focused enough that it can be answered in a few paragraphs; containing enough information that the task/question is well-defined; and describing what research you've done and what approaches you've considered. As it stands, the question is open-ended and what you mean by "power graph analysis/visualization" is not explained, so you would do better to explain what you mean by that: what tasks do you want to be able to do? what visualizations do you want to produce? what requirements do you have for any such system?

  • $\begingroup$ "As it stands, the question is open-ended and what you mean by "power graph analysis/visualization" is not explained", it was not open-ended, just you are not familiar with the topic so it looked open-ended to you. I found the paper with pseudocode to implement it. It is ignorance and arrogance to describe someone's question as vague instead of realizing you are not able to understand it. $\endgroup$ Commented Jul 30, 2016 at 3:28
  • $\begingroup$ See "Computational Network Theory - Theoretical Foundations and Applications", page 226. $\endgroup$ Commented Jul 30, 2016 at 3:28
  • $\begingroup$ @AliShakiba, I'm afraid I don't own the book, so I can't easily take a look at page 226 right now. Sorry about that. Would it be easy to summarize the ideas there that you think I should know? Incidentally, you might find it interesting to take a look at the guidelines from the CSTheory.SE faq: "Questions should be based on knowledge sharing [..] Try to make your question interesting for others by providing some background knowledge". That's a different site, but their advice may be helpful for improving the question. $\endgroup$
    – D.W. Mod
    Commented Jul 30, 2016 at 3:56
  • $\begingroup$ In any case, you're more than welcome to edit your question to address the feedback you've been given. $\endgroup$
    – D.W. Mod
    Commented Jul 30, 2016 at 3:57
  • $\begingroup$ Already deleted the question. Here is the link to the algorithm (open first result). $\endgroup$ Commented Jul 30, 2016 at 4:06
  • $\begingroup$ "Power Graph" is a specific simplification/visualization of large graphs/networks, there was a link to Wikipedia article about it in my question. $\endgroup$ Commented Jul 30, 2016 at 4:08
  • 1
    $\begingroup$ @AliShakiba, OK. Thanks for elaborating on your concerns -- that helps! I confess I'm still trying to work out what exactly your question is. Are you asking for an algorithm for some task (some task related to power graphs)? Or, do you already have an algorithm in mind and you are asking for details regarding how to implement it? If the latter, is there some specific aspect of the algorithm you are unclear on? $\endgroup$
    – D.W. Mod
    Commented Jul 30, 2016 at 4:23
  • $\begingroup$ If I can understand what question you want to ask, and if you are still interested in seeing the question appear here, I'd be happy to offer advice about how you can edit the question to improve it and make it suitable here. (cc: @AliShakiba) Putting a question on hold is not intended to necessarily be a permanent state; if it can be edited to make it suitable, then it can be re-opened, and that's a great outcome. I am glad to try to help you work out how to edit it to make it suitable here, if I can. $\endgroup$
    – D.W. Mod
    Commented Jul 30, 2016 at 4:24
  • $\begingroup$ @D.W. My task was to find an open-source implementation or implement power graph analysis algorithm for large graph/network simplification/visualization. So I was looking for either an open-source implementation or a detailed explanation so that I can implement it myself. $\endgroup$ Commented Jul 30, 2016 at 4:57
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ Commented Jul 30, 2016 at 9:49

Borderline. While it is a computer science on first sight, it may not be answerable as such.

In my experience, there will often not be such papers, so the question is probably "please tide me over from abstract description to code"; that's a programmer's task and clearly offtopic here.

  • $\begingroup$ This is definitely not a programming request, as I'm fine with a description which includes enough details to implement it myself. Many of papers about algorithms include pseudocode as part of algorithm description. $\endgroup$ Commented Jul 25, 2016 at 17:27
  • $\begingroup$ @AliShakiba Some do, many don't. Is my experience, fwiw. $\endgroup$
    – Raphael
    Commented Jul 26, 2016 at 22:06
  • $\begingroup$ Found my answer: "Computational Network Theory - Theoretical Foundations and Applications", page 226 $\endgroup$ Commented Jul 30, 2016 at 3:30

You must log in to answer this question.

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