On p.179, ex.4.5.7, there should be instead of , because inequality is obvious.

*[Correction added, thanks – T.]*

*[Correction added, thanks – T.]*

Sorry, I made a typo too, that is Ex: 8.3.4, not Ex: 3.3.4.

Here are another possible typo. On page 318, Ex: 8.3.5, is that \Omega(|A|^{5/2}) instead of \Theta(|A|^{5/2}). I’ve checked the reference [79] and found out that the result is not that strong.

*[Corrected, thanks – T.]*

Here are a possible typo. On page 318, Ex: 3.3.4, is that |A|^{1/2} |B|^{3/2} |C|^{1/2} rather than |A|^{1/2} |B|^{1/2} |C|^{3/2}.

*[Corrected, thanks – T.]*

*[Corrected, thanks – T.]*

On p.163, in Lemma 4.6:

In max(…) there should be instead of . We should apply Chernoff’s bound not for , but for .

Otherwise the bound is too weak for , .

*[Correction added, thanks – T.]*

For Lemma 4.36, one applies Lemma 4.35 with d taken to be slightly larger than the upper bound in the second display, then no dissociated sets will occur.

]]>Hello. I’m reading your book and find chapter 4 an extremely fascinating introduction to Fourier analytic methods. Yet there are two points on which I wonder if you could kindly elaborate a little.

The first is exercise 4.2.3 (p. 159), about “dyadic decomposition” just below the display. I suppose, following the previous hints, one should split the function into a sum, with each summand ranging within a dilation of a dyadic interval. In this regard, the strategy is unable to cover all scenarios, for example when, arranging values in ascending order, we have the subsequent one is always more than twice (or N times) the preceding. In this case, each dyadic (or N-adic) interval can only contain one value, and the logarithm estimate degrades.

The second has to do with the “Fourier concentration lemma”, in p. 182. In the proof, the sentence before the second display confuses me. To be precise, if one directly uses lemma 4.35, then he merely can find a union of dissociated sets plus a cube, instead of the promised cube alone. If one incorporated the dissociated sets into the cube, then I truly wonder how he could deal with the dimension. In addition, I have taken a glance at [48], but only found the proof of the latter part.

Sincerely I look forward to your reply.

]]>