@JohanWiltink Really sorry for my incorrect editing and hope you can revert them. I searched a lot but didn't find clear documents about how to update a kata. And even after @Blind4Basics replied to me in another kata, I still didn't get the right steps.
It is strange that allowing anyone to click the edit button. Why does codewars design like this, instead of like Github? And where are buttons similar to "fork" and "send pull request"? Tutorials with screenshots will be great.
Once again, my last time to make kata broken. Apologize for the inconvenience.
I'll translate it to more languages later if no issues arise in the near future.
Yep, I can make sense of it now.
I have a solution, but I don't know enough Python. Can you do a JS or Haskell translation? :P
Too late. It's already retired.
Yes, that could have been an Issue. But why would I take more points from you than you already did to yourself? ( Unresolved Issues cost the author points. Unsatisfied votes also cost you points. )
besides i dont get the option of unpublishing this.
at least I included random tests
umm... doesnt that belong in an ISSUE comment not a SUGGESTION comment?
This lacks all originality and novelty.
Please unpublish this kata.
I think I've mixed the rows 4 and 5 at some point, now it should be correct.
No, the existence of pairs (x, y) and (y, z) in a binary relation doesn't imply that x > z, the pair (x, z) must be present too for that.
x > z
I wrote something, but on second thought it didn't make sense.
The example matrix and graph do not seem to be equivalent. The matrix has 5 -> 7 and 6 -> 7; the graph has 4 -> 7 and 6 -> 7. ( I stopped checking after stumbling over this; not saying there isn't more. )
Binary relations are not transitive?
fix scanr to pass bonus tests, by adding a check about fn.length
That change was actually incorrect. Tests possibly did not catch that ( another things to look into when I'm doing the update ), but you put a scanr with wrong logic in the Example solution.
A fork with updated tests ( like @natan's, above ) would have been very useful and helpful. That could have been tested, and iterated on, until it was ready for a big bang replace of the tests.
This, unfortunately, just creates difficult work to undo the changes, after which all the effort ( mine and yours ) will have resulted in, in the best case, no effect at all ( and possibly some lingering problems ).