Timeline for Is there an intuitive interpretation of a negative variogram "nugget" value?
Current License: CC BY-SA 3.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 13, 2012 at 22:03 | review | First posts | |||
| Sep 17, 2012 at 17:55 | |||||
| Sep 6, 2012 at 17:11 | comment | added | whuber♦ | The nugget actually is not the extrapolation to zero: it is the (necessarily nonnegative) jump at zero. When extrapolation of a fitted ("model") variogram based on the data is negative, this suggests that the spatial process is smooth, which for certain kinds of images might be no surprise, and you should be using a smooth model. This behavior also arises from data with finite (rather than point) support, which is usually the case with digital images, whose values are pixel averages. The true underlying variance is "smeared" out, frequently resulting in what you describe. | |
| Sep 6, 2012 at 16:34 | answer | added | Sam Livingstone | timeline score: 2 | |
| Aug 18, 2012 at 20:57 | history | tweeted | twitter.com/#!/StackStats/status/236929724369231872 | ||
| Aug 18, 2012 at 19:46 | history | edited | chl | CC BY-SA 3.0 |
edited title
|
| Aug 18, 2012 at 18:11 | history | edited | timday | CC BY-SA 3.0 |
stress
|
| Aug 18, 2012 at 17:53 | history | asked | timday | CC BY-SA 3.0 |