A Nice Mandelbrot Problem
by djmathman, Jan 13, 2018, 4:23 AM
oops I head off to school tomorrow for second semester junior year whee
Solution
Mandelbrot 2017-2018 wrote:
Let
be the set containing all positive integers less than
that are not multiples of
. Suppose
is partitioned into forty sets, each containing three of the numbers, along with one set containing the single leftover number. What is the largest possible value for the single leftover number, if each set of three numbers form an arithmetic sequence with difference
or
?






Solution
I probably would not have gotten this live (this took me thirty minutes to solve), but I like it quite a bit.
The crucial claim (at least for me) is that the set of almost-coverings of
with arithmetic progressions bijects to the set of tilings of an
board with
and
tiles and one square uncovered. This is easily seen by taking the square in the
row and
column (where here the lower left corner is row
and column
) and associating it with the number
. Then horizontal tiles correspond to arithmetic progressions of difference
while vertical tiles correspond to arithmetic progressions of difference
.
Now we perform a standard roots of unity argument. For all
and
, write the number
in the cell in the
row and
column, where
is a primitive third root of unity. Then the sum of the numbers in the cells covered by any tile is zero. But the sum of the numbers in all
cells is
Hence the lone cell must have its row and column number sum to something which is
. This reduces the possible locations of this lone cell to those dotted in the diagram.
![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
for(int j=0;j<=10;j=j+1)
{
if ((i+j) % 3 == 1 && (i < 11))
dot((i+0.5,j+0.5));
}
}
[/asy]](//latex.artofproblemsolving.com/4/8/1/481e5ba9e0eddac3caadda8ba5b6ed8037fd3e37.png)
But note that since we are restricted to just
and
tiles, we may perform this exact same analysis on a board with the numbering system of the columns reversed (i.e. with the board reflected about its vertical line of symmetry)! This actually leaves only nine possible slots for the lone cell left, as shown.
![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
for(int j=0;j<=10;j=j+1)
{
if ((i+j) % 3 == 1 && (i < 11) && ((i-j) % 3 == 0))
dot((i+0.5,j+0.5));
}
}
[/asy]](//latex.artofproblemsolving.com/4/e/d/4ed0422dbfd60f9be0423aee8d8556d9f739cb11.png)
The one requested in the problem statement is the one in the upper right hand corner, which corresponds to the number
. A construction is shown below.
![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
if(i < 11)
{
draw((i+0.5,0.5)--(i+0.5,2.5));
draw((i+0.5,3.5)--(i+0.5,5.5));
}
}
draw((0.5,6.5)--(0.5,8.5)^^(1.5,6.5)--(1.5,8.5)^^(9.5,8.5)--(9.5,10.5)^^(10.5,8.5)--(10.5,10.5));
for(int i=0;i<=2;i=i+1)
{
for(int j=0;j<=1;j=j+1)
{
draw((3*i+2.5,j+6.5)--(3*i+4.5,j+6.5));
draw((3*i+0.5,j+9.5)--(3*i+2.5,j+9.5));
}
}
draw((2.5,8.5)--(4.5,8.5)^^(5.5,8.5)--(7.5,8.5));
dot((8.5,8.5));
[/asy]](//latex.artofproblemsolving.com/9/c/e/9ce3e39b7ae92caac8ce820e0608b8a8eb35badb.png)
The crucial claim (at least for me) is that the set of almost-coverings of











Now we perform a standard roots of unity argument. For all







![\[(1+\omega + \cdots + \omega^{10})^2 = (1+\omega)^2 = \omega.\]](http://latex.artofproblemsolving.com/4/1/c/41cf17047d7e1b2572b924ed2121657fff95e526.png)

![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
for(int j=0;j<=10;j=j+1)
{
if ((i+j) % 3 == 1 && (i < 11))
dot((i+0.5,j+0.5));
}
}
[/asy]](http://latex.artofproblemsolving.com/4/8/1/481e5ba9e0eddac3caadda8ba5b6ed8037fd3e37.png)
But note that since we are restricted to just


![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
for(int j=0;j<=10;j=j+1)
{
if ((i+j) % 3 == 1 && (i < 11) && ((i-j) % 3 == 0))
dot((i+0.5,j+0.5));
}
}
[/asy]](http://latex.artofproblemsolving.com/4/e/d/4ed0422dbfd60f9be0423aee8d8556d9f739cb11.png)
The one requested in the problem statement is the one in the upper right hand corner, which corresponds to the number

![[asy]
size(180);
defaultpen(linewidth(0.75));
for(int i=0;i<=11;i=i+1)
{
draw((i,0)--(i,11)^^(0,i)--(11,i));
if(i < 11)
{
draw((i+0.5,0.5)--(i+0.5,2.5));
draw((i+0.5,3.5)--(i+0.5,5.5));
}
}
draw((0.5,6.5)--(0.5,8.5)^^(1.5,6.5)--(1.5,8.5)^^(9.5,8.5)--(9.5,10.5)^^(10.5,8.5)--(10.5,10.5));
for(int i=0;i<=2;i=i+1)
{
for(int j=0;j<=1;j=j+1)
{
draw((3*i+2.5,j+6.5)--(3*i+4.5,j+6.5));
draw((3*i+0.5,j+9.5)--(3*i+2.5,j+9.5));
}
}
draw((2.5,8.5)--(4.5,8.5)^^(5.5,8.5)--(7.5,8.5));
dot((8.5,8.5));
[/asy]](http://latex.artofproblemsolving.com/9/c/e/9ce3e39b7ae92caac8ce820e0608b8a8eb35badb.png)
This post has been edited 2 times. Last edited by djmathman, Jan 13, 2018, 4:42 AM
Reason: small mistake in sol
Reason: small mistake in sol