OJ Board

I changed every float to double and widen the alfa angle precision, still WA...
I also tried another implementation of graham scan, it sorts the vertices by coordinates, not by polar angle range, and it computes 2 loops, first the upper convex hull, then the bottom part. The code goes like this ...

upgrade for one point input, now the output consiists of route from the point to itself... still WA, someone posted some auto generatoed points with correct output, and I got it all right with my code
http://online-judge.uva.es/board/viewto ... =218#p5388
still WA

okay, now I`m pretty sure I`ve covered it all. I ran like 20 tests on collinear points covering every possible situations with collinear points, but still getting WA from the judge... here`s the modified code, and I`m pretty clueless right now...


#include <cstdio>
#include <math.h>
#include ...

okay, I know what's wrong.. It was alrdy 4a.m when I was dealing with those collinear points, and I screwed pretty much... there are 2 issues I noticed so far, first if there`s only one line of points in total, and they`re collinear, then the convex hull must contain all of the inner vertices twice ...

Hi there, I`ve been struggling for some hours to get my project through this uva judge, but I`m still getting WA although I`ve checked everything like ten times. I`ve also tried all the possible inputs I`ve found on this forum so far and I`m getting the correct answers all the time... Got no idea ...