Difference between revisions of "Talk:2012 USAMO Problems/Problem 4"
Line 17: | Line 17: | ||
7. For <math>n\ge 3</math>, <math>f(n)=C_1(n-1)+1=C_2(n-2)+2</math> implies that <math>f(n)>2</math>, otherwise <math>C_1=C_2=0</math> and that <math>f(n)=1=2</math>, which is a contradiction. Since <math>f(n)>2</math>, now we have <math>C_2\ge 1</math>, so <math>f(n)\ge (n-2)+2=n</math>. Now we have a lower bound. What is more difficult is to find an upper bound of <math>f(n)</math>. | 7. For <math>n\ge 3</math>, <math>f(n)=C_1(n-1)+1=C_2(n-2)+2</math> implies that <math>f(n)>2</math>, otherwise <math>C_1=C_2=0</math> and that <math>f(n)=1=2</math>, which is a contradiction. Since <math>f(n)>2</math>, now we have <math>C_2\ge 1</math>, so <math>f(n)\ge (n-2)+2=n</math>. Now we have a lower bound. What is more difficult is to find an upper bound of <math>f(n)</math>. | ||
− | ''' | + | |
− | (I know from the main Page it is not too hard to get the upper bound, but to be honest it takes me a while to figure it out.) | + | '''(I know from the main Page it is not too hard to get the upper bound, but to be honest it takes me quite a while to figure it out.)''' |
− | ''' | + | |
8. Therefore <math>n|f(n)!</math>. Now <math>f( (n+1)!) = f(n!) + C( n \cdot n!)</math>, therefore if <math>f(n)=n</math>, then suppose <math>f(n+1)=D(n+1)</math>, then we have | 8. Therefore <math>n|f(n)!</math>. Now <math>f( (n+1)!) = f(n!) + C( n \cdot n!)</math>, therefore if <math>f(n)=n</math>, then suppose <math>f(n+1)=D(n+1)</math>, then we have | ||
<cmath> (D(n+1))! = n! + Cn \cdot n!</cmath> | <cmath> (D(n+1))! = n! + Cn \cdot n!</cmath> |
Revision as of 21:27, 3 May 2012
Though not as elegant as the inductive proof, my proof for this problem is quite different from the one posted in the Page, so I would like to paste it here just for reference.
1. First, , equal to or .
2. Second, suppose that for all , . Then for ,
( is an integer.) , therefore and so is the only possibility. Hence . Similar argument yields that , so is a constant function, which can only be or .
3. From now on, suppose that there exists such that .
4. By and that we know that , or .
5. By we know that , and by and we know that , therefore .
6. Then by we know that is odd, so .
7. For , implies that , otherwise and that , which is a contradiction. Since , now we have , so . Now we have a lower bound. What is more difficult is to find an upper bound of .
(I know from the main Page it is not too hard to get the upper bound, but to be honest it takes me quite a while to figure it out.)
8. Therefore . Now , therefore if , then suppose , then we have
, which is not divisible by and so . By induction we have for all .
--Lightest 22:26, 3 May 2012 (EDT)