Date: May 24th, 2015 12:16 PM Author: Claret pea-brained church
Consider the usual x-y coordinate plane. Assign every point (ordered pair of real numbers) the letter A,B, or C. Prove there exist two points with the same letter that are 1 unit apart.
This problem has the advantage of being both easier and more interesting than the dumb combinatorics problems that one dude was posting.
1. What does "every point" mean? Do you mean "every coordinate pair with integral x and y", or do you mean "every coordinate pair with real x and y"? If the latter, LJL at this joke of a question. If the former, read on for a disproof.
2. Assuming you mean that, then the statement is false. If you define the set of A points as {(x,y)|(x+y)mod2=0} and B points as {(x,y)|(x+y)mod2=1} where x and y are integers, then there is no A point within 1 unit of another A, and there is no B unit within another B, and there is no C unit within another C since there are no C units.
Proof that there is no A point within 1 unit of another A:
Choose two points in A, n and m. n = (x1,y1), m=(x2,y2).
The distance between m and n = sqrt((x1-x2)^2+(y1-y2)^2)).
Assume these two points are within 1 unit of each other. Because by postulation x1,y1,x2,y2 are all integers, in order to be within 1 unit of each other, either x1=x2 and |y1-y2|=1 or y1=y2 and |x1-x2|=1. The proof of this is left as an exercise for the reader, but it's pretty fucking obvious, merely tedious.
Now we defined A such that (x+y)mod2=0, so let's look at the first point, W/OLOG:
(x1+y1)mod2=0, and now do some substitution
(x2+y1)mod2=0, and by a theorem of absolute value addition I don't wnat to prove here,
(x2+(y2+1))mod2=0
x2mod2 + y2mod2 + 1mod2 = 0
By the same definition of A, we know
(x2+y2)mod2 = 0
x2mod2 + y2mod2 = 0
So combine those,
x2mod2 + y2mod2 + 1mod2 = x2mod2 + y2mod2
simplify
1mod2 = 0
1 = 0
This is a contradiction. The proof regarding two points in B is almost exactly the same and is left as an exercise for the reader.
Accordingly, the theorem has been disproved.
In fact, you can generalize the opposite statement as a theorem in N dimensions:
For any set of N-dimensional points with integral components, you can assign these points with N different labels such that no two points with the same label have a distance <= 1
Date: May 24th, 2015 9:17 PM Author: tripping address really tough guy
i'm dumb, bc i thought the question was to prove that there are two points *less* than one unit apart with teh same label, which actually you can do with only 2 circles, like this:
make a circle around "A". all points in this circle must be either B or C points. wlog, choose a "B" point on the boundary of this circle. draw a circle around it. all points within this circle must be A or C points. but there is a region where the A circle and the B circle intersect. all points inside this region must be C points. but since they are all C points and they are in a continuous region where the max distance is less than one, there are two C points within distance 1.
Date: May 24th, 2015 11:52 PM Author: Mahogany hairraiser associate
Mathbro here. gonna hit this question. So this means every point is randomly assigned a A,b or c? (Every unit could be A?) also, do you count diagnolly as one unit or what