Talk:Mollifier

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

A comment to note 14 =[edit]

It is not true that Hörmander did not define f(t) on the whole real line, since otherwise it would be a mistake to consider f(1 - |x|^2) because the argument is not always positive for x in R^n. Also it is not clear who \varphi_2 is under "Smooth cutoff functions". Maybe an edit of the wiki page can improve the article. — Preceding unsigned comment added by 89.162.87.196 (talk) 07:28, 21 May 2013 (UTC)[reply]

Oleg, thanks for correcting my silly mistake. I must have been really tired. - Gauge 04:57, 16 January 2006 (UTC)[reply]

You're welcome! :) It does look look confusing, if you expect something like
. :) Oleg Alexandrov (talk) 07:09, 16 January 2006 (UTC)[reply]
should we change it to exp( -1 / (1 - x² )) ? — MFH:Talk 23:46, 30 March 2006 (UTC)[reply]

In the end, it says you can create a smooth cutoff funtion from any set, but this is of course, not true. The set must be compact, otherwise what is this "distance from the set"...

Applications[edit]

The intro touches on applications of mollifiers: to create a sequence of smooth functions approximating a nonsmooth one. It doesn't go into much detail, though. Could someone add some example applications? Also, this seems closely related to scale space, although that uses a Gaussian kernel which doesn't have compact support. —Ben FrantzDale (talk) 01:53, 29 May 2008 (UTC)[reply]

What's the source of the "Puritan" remark?[edit]

The article says that Donald Alexander Flanders was a "puritan." Now my understanding is that he was not a Calvinist living in the 16th or 17th century. Are you saying he was sexually repressed? Or attempted to sexually repress others? Or what, exactly? What is the meaning and source of this info? It's quite an odd claim to make about anyone without citing proof. In Googling around the only other references are pages that are now citing this anecdote on Wikipedia! So someone has invented some reality here, and I'm hoping that this remark will either be supported with hard data or else retracted.

189.222.221.221 (talk) 02:06, 20 October 2014 (UTC) Steve L[reply]