A positive integer sequence a 1 , a 2 , … a n , a n + 1 , … has a property that for all integers i , j with i > j , the sequence satisfies i a j > j a i
If a 1 0 0 0 = 2 0 1 4 , then what is the the minimum possible value a 5 0 0 can take?
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.
Wow, this problem is rated 320. I realized that level 5 is 200 points and above, not 300 points and above.
I'm really surprised that everything cancelled out so neatly in the induction part. This makes me think there is an easier way...
I was also really surprised that I found the answer to be 1 0 1 4 , not 1 0 0 8 .
Log in to reply
The way I thought about it, is to rephrase your condition as
i a i is a decreasing sequence.
Now, suppose we have such a sequence, with a 1 0 0 0 = 2 0 1 4 . Consider the points ( i , a i ) . Connect them to the origin. We get a "fan" shape, which looks like this:
Claim: If n a n > k for some integer k , then a n − a n − 1 < k .
Proof: This follows immediately from the geometric picture of the fan. Going from the point ( n , a n ) to the point ( n − 1 , a n − 1 ) , we cannot decrease at a rate (strictly) steeper than k , hence moving by unit 1 in the x direction cannot decrease the y direction by (strictly) more than k . □
Corollary: Since 1 0 0 0 a 1 0 0 0 > 2 , hence a k ≥ a 1 0 0 0 − 2 × ( 1 0 0 0 − k ) = 1 4 + 2 k
Corollary. This also tells that the largest n for which we could possibly have a n − a n − 1 ≥ 3 , must also satisfy n a n > 3 . In particular, k < 1 4 .
Corollary. The minimum value of a 1 is less than 1 4 + 2 × 1 = 1 6 . Note that Daniel's proof only applied to k > 1 4 .
Followup: Can you find the minimum value of a 1 ?
Log in to reply
I was a bit confused by what you were asking in the followup, because you stated that a 1 > 1 6 so the minimum should be 1 7 . Oops.
I'm not quite getting the fan shape you are talking about. Can you draw an example?
Also, at the second to last corollary, is it supposed to be a k − a k − 1 < 3 not a k − a k − 1 > 3 ?
I'm pretty sure I'm not getting this. Your claim states that if n a n > k then a n − a n − 1 < k . However, plugging k = 2 in, we have that n 1 4 + 2 n > 2 ⟹ 1 4 > 0 which is already true, so a n − a n − 1 < 2 is always true. But that means that the difference between consecutive terms is never more than 1 which doesn't make sense! I don't know what I'm missing here.
Log in to reply
@Daniel Liu – Ah sorry, I had a bunch of typos. (The last two corollaries were wrongly phrased.) Let me go back and clean things up.
Lol how you can admit that a999>=2012 and not for 500 there's no logic at all
Log in to reply
I have no idea what you are trying to convey to me...
Problem Loading...
Note Loading...
Set Loading...
At first glance, plugging in j = 5 0 0 and i = 1 0 0 0 gives a 5 0 0 > 1 0 0 7 , so the answer should be 1 0 0 8 . But there is a tighter bound; this is because the sequence is of integers , not real numbers.
First, we compare a 1 0 0 0 to a 9 9 9 . We see that using the condition, a 9 9 9 > 1 0 0 0 9 9 9 ⋅ 2 0 1 4 ≈ 2 0 1 1 . 9 so a 9 9 9 ≥ 2 0 1 2 Comparing a 9 9 8 with a 9 9 9 , we obtain (skipping calculations) a 9 9 8 ≥ 2 0 1 0
Comparing a 9 9 8 with a 9 9 7 , we obtain a 9 9 7 ≥ 2 0 0 8
It appears that this pattern continues, so we conjecture that for all k , the sequence satisfies a k ≥ 1 4 + 2 k
Now let's try to prove it using induction. We first check the base case: does a 1 0 0 0 ≥ 1 4 + 2 ( 1 0 0 0 ) ? Yes it does.
Now we assume a k ≥ 1 4 + 2 k and try to prove that a k − 1 ≥ 1 2 + 2 k .
Plugging in i = k and j = k − 1 , we see that k a k − 1 > ( k − 1 ) a k
Rearranging: a k − 1 > k k − 1 ⋅ a k
Using the fact that a k ≥ 1 4 + 2 k :
a k − 1 > k ( k − 1 ) ( 1 4 + 2 k )
Now to prove that a k − 1 ≥ 1 2 + 2 k , we must prove that 1 2 + 2 k > k ( k − 1 ) ( 1 4 + 2 k ) > 1 1 + 2 k
Multiply k on all sides of the inequalities:
k ( 1 2 + 2 k ) > ( k − 1 ) ( 1 4 + 2 k ) > k ( 1 1 + 2 k )
Expanding: 2 k 2 + 1 2 k > 2 k 2 + 1 2 k − 1 4 > 2 k 2 + 1 1 k
Clearly, 2 k 2 + 1 2 k > 2 k 2 + 1 2 k − 1 4 for al k . However, when does 2 k 2 + 1 2 k − 1 4 > 2 k 2 + 1 1 k ?
Simplifying, we see that we must have k > 1 4 . However, this is not a problem since we just want to go down to k = 5 0 0 .
Since we have proved that a k ≥ 1 4 + 2 k for all k > 1 4 , we can just plug in k = 5 0 0 to obtain a 5 0 0 ≥ 1 0 1 4 and we are done.