be integers, let 
be a set with 
be pairwise distinct non-empty, not necessary disjoint subset of 
is called nice if there exists an index 
such that ![\[\sum_{x \in X_k} f(x)>\sum_{x \in X_i} f(x) \quad \text{for all } i \ne k.\]](http://latex.artofproblemsolving.com/3/d/a/3da5ad1bcade98ee08caa35763de226027bcb2c6.png)
Prove that the number of nice functions is at least 
.
.![\[\sqrt{a+\frac{9}{b+2c}}+\sqrt{b+\frac{9}{c+2a}}+\sqrt{c+\frac{9}{a+2b}}+\frac{2(ab+bc+ca)}{9}\ge\frac{20}{3}\]](http://latex.artofproblemsolving.com/b/3/2/b32e0bcba967ca847e878ac186c3168cd6c5fc66.png)
be positive real numbers such that 
Find the minimum value of 
Prove that
, s.t. ecuation: 
have 2 real solutions .
,
.
be a ring with unity such that for every 
there exist 
such that 
.
then 
then the result from a) may no longer hold.
and 
be fixed positive integers. Show that the set of primes that divide at least one of the terms of the sequence 
is infinite.
![\[y''+by'+cy=g(t)\]](http://latex.artofproblemsolving.com/7/8/8/7882dc954ab6dba2445b68d370d630ce1c66a6db.png)
with 
,
are constants has the form ![\[y(t)=\int^{t}_{t_0}K(t-s)g(s)ds\,\]](http://latex.artofproblemsolving.com/6/3/c/63c34df1c0013bf659bc76506a93873a5f879070.png)
and use variation of parameters in each case to find the solution. Is there any short way to calculate solution?
Y by mmkkll, feranjos, CrazyFok, Littlelame, emmelin, KhaiMathAddict, cubres, ItsBesi
Determine all composite integers 
that satisfy the following property: if 
, 
, 
, 
are all the positive divisors of 
with 
, then 
divides 
for every 
.
that satisfy the following property: if 
, 
, 
, 
are all the positive divisors of 
with 
, then 
divides 
for every 
.This post has been edited 7 times. Last edited by v_Enhance, Sep 18, 2023, 12:33 AM
Reason: minor latex pet peeve
Reason: minor latex pet peeve
be the smallest two prime divisors of 
.
for some 
and 
for some nonnegative integer 
.
,
and 
which is a contradiction.
and a positive integer 
.
,
and hence 
and hence 
and hence 
and hence 
and hence 
and hence 
etc.
and hence 
is a prime power which indeed works.
one can get that 
so 
,
in the same way, which leads to 
and 
for prime 
so 
and this is contradiction
for distinct prime divisors 
of 
for some prime divisor ![\[\frac{\frac{n}{p} + n}{\frac{n}{q}} = \frac{q(p+1)}{p},\]](http://latex.artofproblemsolving.com/0/5/c/05c967bf45ae5d33ffba891e1d9c652bc72acc03.png)
obviously not an integer.
and 
or 
;![\[ \frac{\frac{n}{p} + \frac{n}{p^2}}{\frac{n}{q}} = \frac{q(p+1)}{p^2}, \]](http://latex.artofproblemsolving.com/5/0/5/5059120c1c8d485fe7447ba05c92c0a5593df531.png)
obviously not an integer either.
.
Then, for some 
)![\[\frac{n}{q}\large \mid\frac{n}{p^k}+\frac{n}{p^{k-1}}=\frac{n}{p^k}(k+1).\]](http://latex.artofproblemsolving.com/a/d/5/ad52ef76dad656f90976900413d51f04a0a18c4a.png)
This means that for some integer 
,![\[\frac{n\cdot l}{q}=\frac{n}{p^k}(p+1) \iff l\cdot p^k=pq+q\]](http://latex.artofproblemsolving.com/6/8/1/68157585856dc13142624a2d5fc58ee9edff7020.png)
so 
.
distinct prime factors as desired.
for all primes 
,
.![\[
\frac{n}{q} \mid \frac{n}{p^{k-i}} + \frac{n}{p^{k-i-1}},
\]](http://latex.artofproblemsolving.com/4/0/a/40aaa0ef28c0aca6ab6f5bb96699eb71b6d2046a.png)
but this is obviously false. We are done.
for distinct prime divisors 
for some prime divisor 
or 
;
.
is divisible by 
,
divides 
.
.
.
,
for some 
,
.
so 
are both divisible by 
is not.
for primes 
.
which is not an integer.
which are not integers. Thus only 
for some primes 
and some 
.
and hence
Which is impossible unless 
,
to be an integer, which is impossible.
,
and 
be the first two primes of 
be the largest number, such that 
![\[\frac{n}{p_2},\frac{n}{p_1^s},\frac{n}{p_1^{s-1}}\]](http://latex.artofproblemsolving.com/7/a/2/7a2e24917485ff73d99d64ee636ca0a168b0b005.png)
are consecutive divisors of 
Which is a contradiction, hence 
,
,
on both sides, the left side has 
,
,
This does not hold true
for some prime 
.
, 
.![\[d_i=p_{i-1}\mid p_i + p_{i+1}=d_{i+1}+d_{i+2}\]](http://latex.artofproblemsolving.com/d/7/e/d7ec55f74aa6c1f73d60b87f3a29dbf4713c30d3.png)
quite clearly.
for all 
for some 
and 
as the divisors are arranged in increasing order. Now, if 
which is a very clear contradiction since 
are both primes. Thus, there are no such 
where 
work evidently. We now show these are the only ones.
.
![$$\implies \boxed{d_{i}d_{i+1} \mid d_{i+2}(d_{i}+d_{i+1})} [\because i \mapsto m-i-1]$$](http://latex.artofproblemsolving.com/c/1/6/c16495b4b20ccabcb4b9e38e9198166de8adc88a.png)
Now 
gives, 
since 
.
is a prime number (say 
has any prime factor 
which is a blatant contradiction. Thus 
as it can not be 
otherwise 
again, a contradiction. Now by an easy strong induction and similar argument we can conclude that 
.
where 
)
.
-
:
,
.
,
and 
and 
,
.
and 
,
.
,
.
so we must have 
.
.
or 
.
,
,
for a prime 
,
,
is just 
,
always divides 
.
.
,
.
to be an integer, but 
is clearly not an integer. There are no cases here.
for 
.
must be an integer, but this is clearly impossible.
be the smallest two prime divisors of 
Let 
be the two consecutive divisors. Hence, we get 
so 
or 
All perfect powers work.
so 
.
,
.
(easy to see why)
,
with 
are the only solutions (easy to check why these work).
be the largest integer such that 
and 
(assume 
as the case 
is trivial). One may notice that 
,
and 
.
From the above we say 
,
Hence applying this argument in general yields 
So 
So if some prime 
where 
So there is no other prime 
dividing 
It is easy to see that this works.
