Connect the Colours

Connect the Colours - Part 2

Rate this contribution

+13

Make the contribution public (lvl 20+)

Comments

74 Comments

Harry.B. Unnamed CodinGamer

7 days ago

Slightly altered error messages and also changed the black colour to a gray for easier viewing.

M4RGUS Unnamed CodinGamer

19 days ago

Ok for me! Ready to go!

Cg_TheLegend Unnamed CodinGamer

24 days ago

Ok for me! Ready to go!

CodinBot Unnamed CodinGamer

about 2 months ago

I hope you received enough feedback from the community to improve your contribution. Once you have made the necessary updates, publish your contribution!

Harry.B. Unnamed CodinGamer

about 2 months ago

Updated title to "Part 2" as an easier version is being created for "Part 1". Updated the images from imgur due to inaccessibility in the UK.

CodinBot Unnamed CodinGamer

about 3 months ago

I hope you received enough feedback from the community to improve your contribution. Once you have made the necessary updates, publish your contribution!

Timinator Unnamed CodinGamer

about 3 months ago

I am stuck on the 4 Large Sparse test cases. Test Case 18 seems quite different to me from the other 3 in that it is not easily divided into rectangular sections that can be solved independently of the bigger puzzle. For the other 3, it seem I can break the grid into rectangles that can be solved independently. Test case #18 looks like it needs a completely unique strategy. If there is some specific logic strategy that solves the sparse test cases, I wonder if more sparse test cases are needed to make sure the user doesn't splice something together like I'm hoping to do.

Timinator Unnamed CodinGamer

about 3 months ago

I can now solve everything except test cases 18 and 35. Getting to this point has not been a "medium" difficulty experience for me. Maybe I'm doing something wrong, but really minimizing the number of legitimate paths has been brutal (in a fun sort of way 🙂). I can add a 'janky' assumption like you mentioned in the comments below and solve all test cases. Are you able to solve all test cases in Python without any jankiness now?

darkhorse64 Unnamed CodinGamer

about 4 months ago

Hmm ... Isn't that another mission for AlgorithmX ?

Timinator Unnamed CodinGamer

about 3 months ago

Did I see "logical puzzle" in the first sentence??? Outstanding job on the puzzle @Harry.B. Great test case design! I cannot solve the big grids with Algorithm X alone. I will need some logical problem-space reduction and just thinking about how to do that has been super fun this morning!

5DN1L Unnamed CodinGamer

about 4 months ago

Congratulations on drafting your first puzzle contribution using the CodinGame SDK! 🎉 Your question 2 * 'Unsure if "incorrect paths" should be displayed or not': are you referring to the console or the viewer? In either case, could you clarify the scenarios you're considering? Rules * Please specify the location of the origin (0, 0), and indicate the directions which the x- and y-coordinates increase. Loss Conditions * Consider reordering the points to match the order in which the loss conditions are checked: "Response…", "The…", "Attempted…" and "Number…". Output * Please change the heading to "Output per turn". * Please change "first coordinate" to "first tile". * Please change "second coordinate" to "second tile". * Please state that each section of the path must be strictly horizontal or vertical. (i.e. either [[x1]] = [[x2]] or [[y1]] = [[y2]]) Your question 1 * Unsure on the limits for turns / time limits: Have you tested your solution to verify that it's possible to solve all cases within the current time limits? This should serve as the initial basis. We'll also need feedback from other CodinGamers to determine whether the limits feel too loose, just right, or too tight. Example * Please add "3 3" as the first line of the input. * (optional) You could embed an image (e.g. of a grid showing the coordinates and the solution) here if it helps illustrate the example. Default stub * The order of the variables in the default output line differs from the one stated in Output section. Error messages * "Invalid colour identifier number": Consider deleting the word "number" for consistency with Output section. * "Same input or not straight line": Would it be more precise to say "point" instead of "input"? Viewer * One path segment was missing in the viewer after I passed the case "Around the World": https://imgur.com/... . Here's the solution I used: h, w = [int(i) for i in input().split()] for i in range(h): row = input() solution = """1 0 0 0 1 1 2 2 2 2 0 0 0 3 1 3 0 3 3 1 0 3 3 3 1 1 0 1 1 1 2 0 2 1 1 3 0 2 0 1""".splitlines() for line in solution: print(line) Cases * Look good, except that the "Large" cases aren't really that large? I'll probably examine the cases more closely later. Thanks. P.S. Yay, I found the Easter egg too! 🤣

Harry.B. Unnamed CodinGamer

about 4 months ago

Thanks 😇 Many points to brush up on! Key responses: * Question 1 time - I've done in java and python and both work fine with current time restraints - I've not tested to find the actual execution time of them but was more a general question due to other languages - My understanding is there is a variable time limit for different languages in regular puzzles - so wondering how I work out a good rule for setting a single time limit for all. * Question 1 turn limit - Technically with the tight loss conditions not allowing "invalid paths" the turn limit is h*w - (no. of start_tiles / 2) - The only thing that will change that if allowing for creation of the same path/link multiple times or allow for changing the colour of links/paths. I guess the question refers to Q2 also about allowing invalid moves or removing mention of turn limit as it can't be reached / setting it to a variable based on the equation above. edit: Alternatively I could set the limit to be h*w - (no. of start_tiles / 2) - 1 which would force the user to have at least 1 path connecting more than 2 tiles. * Question 2 - I mean in terms of visuals - adding a connector even if the output is invalid (e.g. connecting wrong colour to a 'start node' or crossing over) Suppose it would help with debugging but is also a bit ugly - https://imgur.com/... * Viewer - That wasn't actually visual, that was a bug in the end game function thinking those were connected... Should be fixed now. * Cases - Originally were 7x7, which isn't large but computationally is a lot larger but I decided that 6x6 is probably a good limit. I could/will test dense 7x7s to see how bad they actually are though. * Error messages and case names were just testing/low effort - Just wanted to get the puzzle out there for others to test / comment on 😃 Hope you enjoyed finding the egg 😂 Thanks for your comments, as always!

Konstant Unnamed CodinGamer

about 4 months ago

Found an Easter egg

Harry.B. Unnamed CodinGamer

about 4 months ago

Thought it might go a bit longer without being found but apparently not... Hope you enjoyed 🙂

ABOUT US CAREERS PRIVACY FAQ BLOG

The #1 tech hiring platform

SCREENING INTERVIEWING