MHB Can You Prove this Specific Inequality Involving Positive Integers?

  • Thread starter Thread starter anemone
  • Start date Start date
AI Thread Summary
The discussion focuses on proving the inequality involving positive integers \( l, k, m, n \) under the conditions \( l+m \le 1982 \) and \( \frac{l}{k} + \frac{m}{n} < 1 \). The goal is to demonstrate that \( 1 - \frac{l}{k} - \frac{m}{n} > \frac{1}{1983^3} \). Participants suggest exploring alternative methods to approach the problem, indicating a desire for varied solutions. The hint encourages considering different problem variants to enhance understanding. The conversation emphasizes the complexity of the inequality and the need for rigorous proof techniques.
anemone
Gold Member
MHB
POTW Director
Messages
3,851
Reaction score
115
Positive integers $l,\,k,\,m,\,n$ satisfying $l+m \le 1982$ and $\dfrac{l}{k}+\dfrac{m}{n}<1$. Prove that $1-\dfrac{l}{k}-\dfrac{m}{n}>\dfrac{1}{1983^3}$.
 
Last edited:
Mathematics news on Phys.org
Hint:

Let $1-\dfrac{l}{k}-\dfrac{m}{n}=\dfrac{a}{kn}$ for some positive integer $a$. And see what happens if we let $a>n$ and $a\le n$.

The skill of using the concept of contradicting of minimality with respect to certain variable helps too.
 
Solution of other:

Let us consider the general case with 1982 replaced by $P$. We have $l
\dfrac{1}{P}$. So there is no need in making $n$ too big.

Certainly we must have $n<2Pm<2P^2$. Similarly for $k$. So there are only finitely many candidates for $l,\,m,\,n,\,k$. Hence there is a set of values which minimizes $1-\dfrac{l}{k}-\dfrac{m}{n}$. Let us adopt these values.

Clearly $1-\dfrac{l}{k}-\dfrac{m}{n}=\dfrac{a}{kn}$ for some positive integer $a$. We may assume that $k>n$. The fact that $l,\,m,\,n,\,k$ is an optimal set means that $a$ cannot be too large. Multiplying across, $a=kn-ln-km$. So if $a>n$, we could increase $l$ by 1. That would reduce $1-\dfrac{l}{k}-\dfrac{m}{n}$ to $\dfrac{a-n}{kn}$, contradicting minimality. So, $a\le n$.

Now, $kn=a+ln+km\le a+lk+km=a+k(l+m)\le a+kP$. Hence, $n\le P+\dfrac{a}{k}\le P+1$.

But $\dfrac{a}{kn}+\dfrac{l}{k}=1-\dfrac{m}{n}\ge \dfrac{1}{n}\ge \dfrac{1}{P+1}$.

We have $\dfrac{a}{kn}+\dfrac{l}{k}=\dfrac{1}{k(l+1)}\le \dfrac{1}{k(l+1)}\le \dfrac{1}{kP}$. So, $\dfrac{1}{k}\ge \dfrac{1}{P(P+1)}>\dfrac{1}{(P+1)^2}$.

Hence $1-\dfrac{l}{k}-\dfrac{m}{n}=\dfrac{a}{kn}>\dfrac{a}{(P+1)^3}\ge \dfrac{1}{(P+1)^3}=\dfrac{1}{1983^3}$.
 
I am thinking it would be nice to show another variant of the same problem that uses other method to tackle it...here goes the problem:

Let $l,\,k,\,m,\,n\in\Bbb{Z_+}$ and $a=1-\dfrac{l}{k}-\dfrac{m}{n}$. If $a>0$ and $l+m\le 1982$, then prove that $a>\dfrac{1}{1983^3}$.

We have 3 cases to analyze:

Case I: If $k,\,n\ge 1983$, then $a\ge 1-\dfrac{l+m}{1983}\ge 1-\dfrac{1982}{1983}>\dfrac{1}{1983^3}$.

Case II: if $k,\,n\le 1983$, then $a=\dfrac{kn-ln-km}{kn}>0$ so that $kn-ln-km \ge 1$. Thus, $a\ge \dfrac{1}{kn}>\dfrac{1}{1983^3}$.

Case III:

Suppose now that $k<1983<n$, if $n>1983^2$ and $a<\dfrac{1}{1983^3}$, then $\dfrac{m}{n}<\dfrac{1982}{1983}$, thus, $1-\dfrac{l}{k}<\dfrac{1}{1983}$. This implies that $k>1983(k-l)>1983$ because $l\le k+1$, which is absurd. If $n\le 1983^2$, then as in the case II above, we have $a\ge \dfrac{1}{kn}>\dfrac{1}{1983^3}$.
 
Seemingly by some mathematical coincidence, a hexagon of sides 2,2,7,7, 11, and 11 can be inscribed in a circle of radius 7. The other day I saw a math problem on line, which they said came from a Polish Olympiad, where you compute the length x of the 3rd side which is the same as the radius, so that the sides of length 2,x, and 11 are inscribed on the arc of a semi-circle. The law of cosines applied twice gives the answer for x of exactly 7, but the arithmetic is so complex that the...
Back
Top