Jump to content

Talk:Wave front set

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

If s.o. wants to delete "the notion of" from the intro, maybe some more care should be taken to precise that ξ must be nonzero in order to belong to Σ (i.e. add "ξ=0 or" after \iff)... MFH: Talk 23:44, 25 May 2005 (UTC)[reply]

(DONE) MFH: Talk

unprecise precisions

[edit]

Dear C.M.,

thanks for putting the [ [ ... ] ] at the right place (cotangent bundle vs WF set). However, I'm not so happy about your other edits, and rather prefer the old version:

  • 'direction' is a bit misleading, because the Fourier transform is about frequency (i.e. roughness) and not spatial direction. And, I think at this quite ealy point of the article, such a precision is not yet needed and rather tears apart the 'preface'. (Which, I admit, is not obvious as I did not "start" the article with an ==introduction== or ==history== section.) However, your point about dimension > 1 is worth mentioning. Maybe we should duplicate this, and put your developed version in a separate paragraph, leaving my short version in the 'preface'.
  • I can't see why you put the miniscule T*(X) into a 'displayed' equation. Well, I can imagine: you choose to display all math tags as PNG. But this is your choice, not the standard setting, and I feel it's not justified.

On both points, I'm quite tempted to put back the original version. Would you mind very much? MFH: Talk 12:44, 26 May 2005 (UTC)[reply]

Direction: it is how I think of microlocal, somehow. That is, local analysis which is like localisation in algebra, or inverting operators, has to be supplemented by something talking about directions at a point. The cotangent 'directions' are available to talk about, and the Fourier interpretation is just one way of looking at that. Well, I suppose I am just discussing my own intuitions. General ideas of this kind really belong at the level of introduction to the microlocal analysis article; which is yet to be written.

Format: at present the consensus seems to be MathHTML for inline formulae. I don't feel it is harmful to display more formulae: it makes a page less dense to read.

Charles Matthews 14:47, 26 May 2005 (UTC)[reply]

Geometric Asymptotics by Gullemin and Sternberg, Ch. VI Geometric Aspects of Distributions, introductory paragraph, contains this:

... the notion of the wave front set, a subset of the cotangent bundle, which measures the singular codirections for distributions.

Charles Matthews 15:46, 27 May 2005 (UTC)[reply]

Thanks for recalling the notion of codirection which I did not think of and thus did not mention in my article, which definitely is far from being complete. So I reiterate what I already suggested:

However, your point about dimension > 1 is worth mentioning. Maybe we should duplicate this, and put your developed version in a separate paragraph, leaving my short version in the 'preface'.

MFH: Talk 18:34, 27 May 2005 (UTC)[reply]

Typo

[edit]

Just to point that there is 2 different choice of functions "phi" in the formal definition ( in tex this would be \phi and \varphi but I don't do math in html) so please fix: First reference to test function is used with \phi second one is typed outside math env. and doesn't not look at all the same (phi(x_0) not 0)