Difference between revisions of "2008 IMO Problems/Problem 5"
Line 21: | Line 21: | ||
a_i \textrm{ if } a_i \in A \\ | a_i \textrm{ if } a_i \in A \\ | ||
a_i-n \textrm{ if } a_i \in B | a_i-n \textrm{ if } a_i \in B | ||
− | \end{array} \right for i=1,2,\ldots,k</cmath> | + | \end{array} \right \textrm{for }i=1,2,\ldots,k</cmath> |
What we want to show now is that each element of <math>\cal{M}</math> is an image of exactly <math>2^{k-n}</math> elements from <math>\cal{N}</math>, which would imply <math>N = 2^{k-n}M</math> and solve the problem. | What we want to show now is that each element of <math>\cal{M}</math> is an image of exactly <math>2^{k-n}</math> elements from <math>\cal{N}</math>, which would imply <math>N = 2^{k-n}M</math> and solve the problem. | ||
− | Consider an element <math>y</math> of <math>\cal{M}</math> let <math>l_i</math> be the number of | + | Consider an arbitrary element <math>y</math> of <math>\cal{M}</math> and let <math>l_i</math> be the number of appearances of the number <math>i</math> in <math>y</math> for <math>i=1,2,\ldots n</math>. Now consider the set of pre-images of <math>y</math>, that is <math>X_y = \{ x | f(x) = y \}</math>. |
It is easy to see that each element <math>x\in X_y</math> is derived from <math>y</math> by ''flipping'' an ''even'' number of its <math>1</math>-s, <math>2</math>-s, and so on, where flipping means changing the number <math>j\in A</math> to <math>j+n\in B</math>. Since each such set of flippings results in a unique <math>x</math>, all we want to count is the number of flippings. We can flip exactly <math>0, 2, 4,\ldots </math> of the <math>1</math>-s, so that results in | It is easy to see that each element <math>x\in X_y</math> is derived from <math>y</math> by ''flipping'' an ''even'' number of its <math>1</math>-s, <math>2</math>-s, and so on, where flipping means changing the number <math>j\in A</math> to <math>j+n\in B</math>. Since each such set of flippings results in a unique <math>x</math>, all we want to count is the number of flippings. We can flip exactly <math>0, 2, 4,\ldots </math> of the <math>1</math>-s, so that results in |
Revision as of 05:22, 4 September 2008
Problem 5
Let and
be positive integers with
and
an even number. Let
lamps labelled
,
, ...,
be given, each of which can be either on or off. Initially all the lamps are off. We consider sequences of steps: at each step one of the lamps is switched (from on to off or from off to on).
Let be the number of such sequences consisting of
steps and resulting in the state where lamps
through
are all on, and lamps
through
are all off.
Let be number of such sequences consisting of
steps, resulting in the state where lamps
through
are all on, and lamps
through
are all off, but where none of the lamps
through
is ever switched on.
Determine .
Solution
For convenience, let denote the set
and
the set
.
We can describe each sequences of switching the lamps as a -dimensional vector
, where
signifies which lamp was switched on the
-th move for
.
Let consist of those sequences that contain each of the numbers in
an odd number of times and each of the numbers in
an even number of times. Similarly, let
denote the set of those sequences that contain no numbers from
and each of the numbers in
an odd number of times. By definition,
and
.
Define the mapping as
\[f(a_1, a_2, \ldots, a_k) = (b_1,b_2,\ldots b_k) : b_i = \left\{ \begin{array}{l} a_i \textrm{ if } a_i \in A \\ a_i-n \textrm{ if } a_i \in B \end{array} \right \textrm{for }i=1,2,\ldots,k\] (Error compiling LaTeX. Unknown error_msg)
What we want to show now is that each element of is an image of exactly
elements from
, which would imply
and solve the problem.
Consider an arbitrary element of
and let
be the number of appearances of the number
in
for
. Now consider the set of pre-images of
, that is
.
It is easy to see that each element is derived from
by flipping an even number of its
-s,
-s, and so on, where flipping means changing the number
to
. Since each such set of flippings results in a unique
, all we want to count is the number of flippings. We can flip exactly
of the
-s, so that results in
flippings. Combine each of them with the
,
, etc. ways of flipping the
-s,
-s etc. respectively to get the total number of flippings:
This shows that
and the proof is complete.