Seven colors in a lattice

You have an infinite lattice of equilateral triangles and you would like to fill each node with one of seven colors.

Is there any way to do this so that no equilateral triangle of any size formed by the lattice lines has its three vertices sharing the same color?

For example, the coloring shown would be illegal since the three vertices of the grey triangle share the same color, red.


Inspiration

Yes No

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

1 solution

Geoff Pilling
Oct 26, 2018

Here is a solution provided by @Jon Haussmann:

For the sake of induction, we're going to strengthen the problem. We claim that for n n colors, there is a triangular grid of some size, say f ( n ) f(n) , such that if the grid is colored with n n colors, then there is a monochromatic triangle.

Clearly, f ( 1 ) = 1 f(1) = 1 . Assume that f ( n 1 ) f(n - 1) exists for some n > 1 n > 1 .

Suppose we have an infinite triangular grid with n n colors. Then by Van der Waerden's Theorem , there exists some number W W , so that for any W W vertices in a row, there exist f ( n 1 ) + 2 f(n - 1) + 2 vertices, equally spaced, that have the same color. If we draw the lines through these f ( n 1 ) + 2 f(n - 1) + 2 points, then we effectively have a triangular grid of size f ( n 1 ) f(n - 1) , and at most n 1 n - 1 colors, so by the induction hypothesis, there exists a monochromatic triangle. Thus, we can take f ( n ) = W 1 f(n) = W - 1 . By induction, we are done.

So Van der Waerden's Theorem , on arithmetic sequences (which is a somewhat advanced, but established result) is exactly what we need to generate the smaller grid.

By inspection, many of the smallest triangles are not

Dave Neely - 2 years, 7 months ago

What if we assigned the colors a number n where the lattice consists of the sequence in increasing order from n, n+1, n+1,..., n+k, in which n+k is the maximum number of colors and then in the next line under the lattice restart the sequence. Ex.: 1 2 3 4 5 6 7 1 2 3 4 5 6 7 7 1 2 3 4 5 6 7 7 1 2 3 4 5 6 7...

Would that be an example of a lattice that works? (Just trying to get a better understanding of the concept.)

Samir Fridhi - 2 years, 7 months ago

Log in to reply

Not sure. I understand... Where do the numbers/colors fit into the lattice?

Geoff Pilling - 2 years, 7 months ago

Even without van der Waarden's theorem, I think compounding an intuitive use of pigeon hole principle gives what you need.

Chris Maitland - 2 years, 7 months ago

You should also make explicit the conclusion that it is not possible to avoid some triangle having two vertices the same color.

Kermit Rose - 2 years, 5 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...