Difference between revisions of "1977 USAMO Problems/Problem 5"
m (→Problem) |
(→Solution) |
||
(6 intermediate revisions by 4 users not shown) | |||
Line 1: | Line 1: | ||
== Problem == | == Problem == | ||
If <math> a,b,c,d,e</math> are positive numbers bounded by <math> p</math> and <math> q</math>, i.e, if they lie in <math> [p,q], 0 < p</math>, prove that | If <math> a,b,c,d,e</math> are positive numbers bounded by <math> p</math> and <math> q</math>, i.e, if they lie in <math> [p,q], 0 < p</math>, prove that | ||
− | <cmath> (a | + | <cmath> (a+b +c +d +e)\left(\frac{1}{a} +\frac {1}{b} +\frac{1}{c} + \frac{1}{d} +\frac{1}{e}\right) \le 25 + 6\left(\sqrt{\frac {p}{q}} - \sqrt {\frac{q}{p}}\right)^2</cmath> |
and determine when there is equality. | and determine when there is equality. | ||
== Solution == | == Solution == | ||
− | {{ | + | Fix four of the variables and allow the other to vary. Suppose, for example, we fix all but <math>x</math>. Then the expression on the LHS has the form <math>(r + x)(s + \frac{1}{x}) = (rs + 1) + sx + \frac{r}{x}</math>, where <math>r</math> and <math>s</math> are fixed. But this is convex. That is to say, as <math>x</math> increases if first decreases, then increases. So its maximum must occur at <math>x = p</math> or <math>x = q</math>. This is true for each variable. |
+ | |||
+ | Suppose all five are <math>p</math> or all five are <math>q</math>, then the LHS is 25, so the inequality is true and strict unless <math>p = q</math>. If four are <math>p</math> and one is <math>q</math>, then the LHS is <math>17 + 4\left(\frac{p}{q} + \frac{q}{p}\right)</math>. Similarly if four are <math>q</math> and one is <math>p</math>. If three are <math>p</math> and two are <math>q</math>, then the LHS is <math>13 + 6\left(\frac{p}{q} + \frac{q}{p}\right)</math>. Similarly if three are <math>q</math> and two are <math>p</math>. | ||
+ | |||
+ | <math>\frac{p}{q} + \frac{q}{p} \geq 2</math> with equality iff <math>p = q</math>, so if <math>p < q</math>, then three of one and two of the other gives a larger LHS than four of one and one of the other. Finally, we note that the RHS is in fact <math>13 + 6\left(\frac{p}{q} + \frac{q}{p}\right)</math>, so the inequality is true with equality iff either (1) <math>p = q</math> or (2) three of <math>v, w, x, y, z</math> are <math>p</math> and two are <math>q</math> or vice versa. | ||
== See Also == | == See Also == | ||
{{USAMO box|year=1977|num-b=4|after=Last Question}} | {{USAMO box|year=1977|num-b=4|after=Last Question}} | ||
+ | {{MAA Notice}} | ||
[[Category:Olympiad Algebra Problems]] | [[Category:Olympiad Algebra Problems]] | ||
[[Category:Olympiad Inequality Problems]] | [[Category:Olympiad Inequality Problems]] |
Latest revision as of 22:08, 27 June 2018
Problem
If are positive numbers bounded by and , i.e, if they lie in , prove that and determine when there is equality.
Solution
Fix four of the variables and allow the other to vary. Suppose, for example, we fix all but . Then the expression on the LHS has the form , where and are fixed. But this is convex. That is to say, as increases if first decreases, then increases. So its maximum must occur at or . This is true for each variable.
Suppose all five are or all five are , then the LHS is 25, so the inequality is true and strict unless . If four are and one is , then the LHS is . Similarly if four are and one is . If three are and two are , then the LHS is . Similarly if three are and two are .
with equality iff , so if , then three of one and two of the other gives a larger LHS than four of one and one of the other. Finally, we note that the RHS is in fact , so the inequality is true with equality iff either (1) or (2) three of are and two are or vice versa.
See Also
1977 USAMO (Problems • Resources) | ||
Preceded by Problem 4 |
Followed by Last Question | |
1 • 2 • 3 • 4 • 5 | ||
All USAMO Problems and Solutions |
The problems on this page are copyrighted by the Mathematical Association of America's American Mathematics Competitions.