I'm confused how to prove this inequality, I tried by multiplying both sides by abc to get ride of fraction which got me

$abc² + bc+ac+ab -(a+c+b)abc-1>0<=> abc²+bc+ac+ab-ab²-bc²-ca²-1 > 0$

I still can't see any way to to transform this inequality.

When i looked at the book solution i found that they come up with formula without explaining how they got there. here's book solution of the inequality

$ (a-\frac 1 {b})(b-\frac 1 {c})(c-\frac 1 {a}) > 0$

What do i need to know so i can make this kind of transformation to the inequality.


which got me $\;\ldots \;\; abc²+bc+ac+ab-ab²-bc²-ca²-1 > 0$

You made a couple of mistakes in the intermediate calculations, which is fairly obvious since the expression you got in the end is not symmetric in $\,a,b,c\,$, as it should be.

What you would actually get, instead, is:

$$ a^2b^2c^2 + ab+bc+ca - a^2bc-ab^2c-abc^2-1 \gt 0 \;\;\iff\;\; (a b - 1) (a c - 1) (b c - 1) \gt 0 $$

[ EDIT ]   The factorization is easier to "see" by defining $\,u=bc, v=ac, w=ab\,$, then:

\begin{align} a^2b^2c^2 + ab+bc+ca - a^2bc-ab^2c-abc^2-1 &= uvw - uv-vw-wu+u+v+w -1 \\ &= (u-1)(v-1)(w-1) \end{align}

  • 1
    $\begingroup$ ops didn't see that thanks! $\endgroup$ – Dota Apr 4 '18 at 2:56
  • $\begingroup$ I still coudn't come up with your formula by myself ? can you go step by step how you got it if thats possible? $\endgroup$ – Dota Apr 4 '18 at 3:05
  • $\begingroup$ @Dota Edited into the answer. Other than that, the expression can also be factored the hard way, by considering it as a quadratic in $a$ (for example) and finding its roots using the quadratic formula. $\endgroup$ – dxiv Apr 4 '18 at 3:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.