#santa-2025
1 messages ยท Page 1 of 1 (latest)
yes i am
i am attempting to understand what's being asked by the competition. is not at all clear to me. i guess enough folks have heard this problem before that it is well-known?
I am
Nice too meet you
It was also no really clear to me. In the evaluation code for the competition https://www.kaggle.com/code/metric/santa-2025-metric they define this class to make a Christmas tree polygon. Looks like this.
๐ . cant put images here . we get n xmas tree shaped polygons defined in the evaluation code and we are tasked to fit them in the tightest square posssible. Shape of trees stays the same and we aim at finding smallest square containing n trees for each n from 1 to 200.
I recommend looking through some code examples. It is explained pretty well there
Thnx
I'm getting ID column id not found in submission even though id column is there
Hi guys, in this competition we package Xmas trees in 2-D I think as there are only x and y coordinates and angle degree in the submit file.
So basically, we should calculate the location (this is the middle point of the polygon I guess) and the angle of each Xmas tree for 200 packages each containing 1 to 200 Xmas trees. It should be optimized to have minimum score.
THis is what I understood so far ๐
Hello everyone
What I did not understand is the id which is of the form N_i
What is this? There are 200 different values of N and 200 of i
that is, the solution row ids for the problem of let's say 5 trees should be 005_0, 005_1 ... 005_4
you have the submit the n coordinates for each of the N problems, toguether
So we are supposed to submit the the coordinates of every tree in each of 200 packages which contain the smallest one 1 tree and the largest one 200 trees
Yes position and direction
20100 trees in total
So a quick doubt, is there any float we are using or float64 or some decimal precision?
https://media.discordapp.net/attachments/1444971360047726605/1445085758598938824/image1.gif?ex=692f107d&is=692dbefd&hm=94f18cd6e7350e7cc612826beb5d11a9fd125485a58ee1e39a16a03b6f9e2426&=&width=237&height=315
https://media.discordapp.net/attachments/1444971360047726605/1445085766937088000/image2.gif?ex=692f107f&is=692dbeff&hm=51e8429e6818b166e21485a613e8f0c706d64c765aefc93f65a7bcefa10907c2&=&width=864&height=1152
https://media.discordapp.net/attachments/1444971360047726605/1445085774562197535/image3.gif?ex=692f1081&is=692dbf01&hm=e520e8e4edd4eea02e82168a7059a868ea59c19d9b90c7c34402f7bb3616c76f&=&width=864&height=1152
https://media.discordapp.net/attachments/1444971360047726605/1445085781801566319/image4.gif?ex=692f1082&is=692dbf02&hm=bdc0715977fdcda4b7804916e5bfb36af1d3132f535d1b4327894a067fbfc769&=&width=725&height=907
so its bin packing right
My team is. This is our first time joining.
Tiling problem is more fitting the description.
Hey I have a question regarding the competition. I started with an easy simple algorythm, but get a score of 116000. this cant be right i think. my code is bad but not bad. any ideas on how i could fix this? help would be greatly appreciated, it's my first competition ๐
I think it might be a scaling issue, but i am not sure.
Because my square is consitantly smaller than the one of the getting started example and i get huge scores
are you appling scale correctly? Try /= 1e15 x and y before generating your csv
Why should I? I never scaled up or anything. Is there a penalty for overlaps maybe? Because I just discovered that my angles in the csv all default to 0 thus creating overlaps
I fixed it still same result.
It must be scaling but i dont understand what you meant. I had no scalefactor.
I have absolute values from 0 to N and normalize them to be between -100 and 100
And why am i getting overlaps even though i just safe stable solutions๐ญ
what?
nvm everything is working now. misread sth
๐
Hi guys, i'm curious, what methods have you guys tried for this competition
seconded
simulated annealing
trying to find other methods currently. Any teams still going at this problem?
I'm still working on this competition... the submission says that there are trees that overlap, but can't figure out how to fix that part yet