Timeline for Estimate peak width from a vector that is a superposition of unknown number of identical Gaussian peaks with different heights?
Current License: CC BY-SA 4.0
32 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 20, 2019 at 11:18 | comment | added | Tom Wenseleers | Well the function I am using, hexp(((x-u)^2)/(-2*(w^2))), has maximum h and the w is the SD of a gaussian density distribution. But it is not strictly speaking a gaussian density distribution, as that would have 1/sqrt(2*piw^2), instead of h... | |
| Aug 20, 2019 at 5:39 | comment | added | Royi | I have zero knowledge in R. What I don't understand is how Gaussian Curve can have width which is independent of height. If you scale $ {e}^{-\frac{ {x}^{2} }{2}} $ its width changes. I assume width has something to do with the STD. | |
| Jan 25, 2019 at 0:00 | history | tweeted | twitter.com/StackSignals/status/1088587394654314502 | ||
| Jan 17, 2019 at 22:54 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
deleted 39 characters in body
|
| Jan 16, 2019 at 19:01 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
deleted 2 characters in body
|
| Jan 12, 2019 at 6:42 | answer | added | Cedron Dawg | timeline score: 2 | |
| Jan 11, 2019 at 19:27 | history | edited | Peter K.♦ | CC BY-SA 4.0 |
Inlined links
|
| Jan 11, 2019 at 16:48 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
edited title
|
| Jan 11, 2019 at 16:41 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
deleted 33 characters in body
|
| Jan 11, 2019 at 9:31 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
edited title
|
| Jan 11, 2019 at 9:18 | history | edited | Tom Wenseleers |
edited tags
|
|
| Jan 10, 2019 at 22:40 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 11 characters in body
|
| Jan 10, 2019 at 19:30 | answer | added | Tom Wenseleers | timeline score: 2 | |
| Jan 10, 2019 at 18:48 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 33 characters in body
|
| Jan 10, 2019 at 17:58 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 47 characters in body
|
| Jan 10, 2019 at 17:46 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 297 characters in body
|
| Jan 10, 2019 at 17:29 | answer | added | Peter K.♦ | timeline score: 0 | |
| Jan 10, 2019 at 17:18 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 9750 characters in body
|
| Jan 10, 2019 at 15:15 | answer | added | A_A | timeline score: 0 | |
| Jan 10, 2019 at 14:42 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 31 characters in body
|
| Jan 10, 2019 at 14:38 | comment | added | Peter K.♦ | Cute question! My go-to for this sort of thing is usually a LMS algorithm... but it wouldn't be fast. And there's a problem of deciding how many gaussians there really are. :-) | |
| Jan 10, 2019 at 14:31 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 57 characters in body
|
| Jan 10, 2019 at 14:29 | comment | added | Tom Wenseleers | Just taking the minimum observed peak width if one would focus on the smallest peak resolved to baseline btw wouldn't work since in my real signals I also have occasional noisy spikes... And there could also be cases where the signal is nowhere resolved to baseline because of the high nr of overlapping peaks... | |
| Jan 10, 2019 at 14:25 | comment | added | Tom Wenseleers | Ideally yes - though the recovery of individual peaks is not that important, it's more that I need a decent estimate of the average peak width over this whole window, taking into account that peaks can of course be superimposed and can overlap. The fact that I assume that all peaks are identically shaped should help with the identifiability of the problem though... | |
| Jan 10, 2019 at 14:22 | comment | added | A_A | What about those peaks that are very very close to each other? Do they need to be resolved individually? | |
| Jan 10, 2019 at 14:21 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 1181 characters in body; edited tags
|
| Jan 10, 2019 at 14:16 | comment | added | Tom Wenseleers | It would be OK to assume the peaks were Gaussian though, even though their width and amplitudes are unknown... The peak locations and amplitudes are not important for my purposes - it's just the average width I need... | |
| Jan 10, 2019 at 14:15 | history | edited | Tom Wenseleers | CC BY-SA 4.0 |
added 1181 characters in body; edited tags
|
| Jan 10, 2019 at 14:12 | comment | added | Tom Wenseleers | Well the problem here would be that the Gaussian is not known, so I would like to estimate the most likely peak shape given that the signal that I measuring is a superposition of many such peaks with different amplitude - I've added some R code to make my question clearer... | |
| Jan 10, 2019 at 13:57 | comment | added | A_A | It really depends on what you are after, you could do Kernel Density Estimation or even Deconvolution with the known gaussian. Do you think you could share a bit more about the problem? | |
| Jan 10, 2019 at 13:50 | review | First posts | |||
| Jan 12, 2019 at 5:01 | |||||
| Jan 10, 2019 at 13:45 | history | asked | Tom Wenseleers | CC BY-SA 4.0 |