Timeline for Contours containing a given fraction of $(x,y)$ points
Current License: CC BY-SA 4.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| S Nov 4, 2022 at 13:32 | history | suggested | J-J-J | CC BY-SA 4.0 |
the link to the image has changed, and is broken. This is hopefully the correct link (if we refer to the image title "FINAL_FIGURES_b11_spt_lcdm", which corresponds to the former URL).
|
| Nov 4, 2022 at 9:21 | review | Suggested edits | |||
| S Nov 4, 2022 at 13:32 | |||||
| Mar 9, 2017 at 17:30 | history | edited | CommunityBot |
replaced http://inspirehep.net/ with https://inspirehep.net/
|
|
| Aug 23, 2013 at 13:14 | comment | added | Andy W | I just found a copy of the bagplot paper, and each are implemented in open-source R libraries. See aplpack and hdrcde. The approach whuber is describing is synonymous with the highest density region by Hyndman. The bagplot is a bit different though, and follows from defining the halfspace depth of points. | |
| Aug 23, 2013 at 9:06 | history | tweeted | twitter.com/#!/StackStats/status/370834460549591041 | ||
| Aug 22, 2013 at 21:18 | comment | added | whuber♦ | I hear you. My concern is that this way of framing the question is likely to lead to huge computational difficulties. Controlling a density bandwidth (or something in that spirit) is easy to do, computationally reasonable, and should be sufficiently flexible to meet most needs. | |
| Aug 22, 2013 at 21:09 | comment | added | Kyle | @whuber Hm that's a good point. Might be that the best answer is just to smooth, but I'm still curious along these lines. Maybe the constraint should be to find a non-self-intersecting contour which links between points with straight line segments and minimises... not area because that would give some sort of very spiky shape I suspect, but... something? | |
| Aug 22, 2013 at 20:54 | comment | added | whuber♦ | You might want to view your choice of kernel as tantamount to satisfying an additional constraint. BTW, the constraints you suggest won't work: the infimum of the areas is always zero and the supremum of the densities within those areas is, of course, infinite. Look instead to constraining the tortuosities of the contour regions (which is more or less what is varied when you change kernel widths). | |
| Aug 22, 2013 at 20:51 | comment | added | Kyle | @AndyW looks interesting, do you happen to know if the paper is available anywhere without a paywall? Otherwise I can probably access it next time I'm in my university library I guess. | |
| Aug 22, 2013 at 20:49 | comment | added | Kyle | @whuber I've played around with density methods (similar to the one you linked) and they work alright, but I wanted to try something that is independent of any smoothing kernel choice or similar. | |
| Aug 22, 2013 at 19:56 | comment | added | whuber♦ | A closely related question is asked and answered at stats.stackexchange.com/questions/63447/…. In this case, if you are not firmly wedded to the idea of imposing an extra constraint explicitly, an answer is easy to come by: just compute a kernel density estimate on a fine grid and select the largest 68%, 95%, or 99% from the cumulative sum of its sorted values: they form a region within the grid. A smoothed version of the boundary of that region will do. | |
| Aug 22, 2013 at 18:44 | comment | added | Andy W | Also see highest density region (if it is ok that the contour is not entirely enclosed in one polygon). | |
| Aug 22, 2013 at 18:22 | review | First posts | |||
| Aug 22, 2013 at 18:24 | |||||
| Aug 22, 2013 at 18:13 | comment | added | Andy W | You may be interested in bag plots. | |
| Aug 22, 2013 at 18:03 | history | asked | Kyle | CC BY-SA 3.0 |