Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Is it true for $1 \leq p<\infty$ that for $f \in L^p(\mathbb R)$ and $q$ is such that $1/p+1/q=1$, $$\|f\|_p= \sup \left\{\int_\mathbb R fg : g\in L^q(\mathbb R), \|g\|_q \leq 1 \right\} ?$$

The integral is with respect to the Lebesgue measure.

share|cite|improve this question

By Hölder's inequality, $LHS \geqslant RHS$, and for the choice $g:=f^{\frac qp}$, assuming without loss of generality that $f\geqslant 0$, we can see that $LHS\leqslant RHS$.

share|cite|improve this answer
Thank you for the answer. If I take the sup over functions with some smoothness, say $g \in C^k$ do I still get the same $p$-norm or something strictly smaller? – Cantor Mar 28 '13 at 14:49
Smooth functions with compact support are dense in $L^q$ for $1\leqslant q<\infty$, so for $1<p<\infty$, it still work. – Davide Giraudo Mar 28 '13 at 14:51
Do I need absolute values inside the integral (i.e. $\sup\int|fg|$)? Also, could you write out the details of $LHS \leq RHS$? It seems to me that $g$ should be taken to be $\frac{f^{p/q}}{\|f\|_p^{p/q}}$ – Cantor Mar 28 '13 at 19:37

Let $f = \chi_{[0,2]}$. Then $||f||_2 = \sqrt 2$. On the other hand, if $g = 100 \chi_{[0,2]}$, then $\int fg = 200$, a contradiction.

share|cite|improve this answer
I think I need to take the sup over $\|g\|_q \leq 1$. – Cantor Mar 28 '13 at 14:31
That sounds right. – user66345 Mar 28 '13 at 14:31
@Cantor, now Davide's answer is right. – user66345 Mar 28 '13 at 14:41
As it stands it's not correct or at least missing some crucial details. Do you care to fill in the details? – Cantor Mar 29 '13 at 11:19

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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