Timeline for How to measure the shift between two cumulative distribution functions (CDFs)?
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 15, 2022 at 13:20 | comment | added | mhdadk | For future readers who are interested in a proof of why the signed area is the difference in expectations, a partial proof is given here under "Proof of expected survival time formula". | |
| Jan 24, 2018 at 0:52 | history | edited | kjetil b halvorsen♦ | CC BY-SA 3.0 |
deleted 26 characters in body; edited tags
|
| Aug 25, 2017 at 4:17 | vote | accept | thanks_in_advance | ||
| Aug 25, 2017 at 3:59 | comment | added | thanks_in_advance | @Nick: You're correct, I'll use red/blue in the end. | |
| Aug 23, 2017 at 14:02 | answer | added | Danica | timeline score: 23 | |
| Aug 23, 2017 at 13:17 | comment | added | Nick Cox | Orthogonal comments, picky if you like, but in graphics details can matter a lot. 1. Difficulty distinguishing red and green is a common problem many people have; here it doesn't matter because the two curves would be distinct whatever the colours, but even for this kind of problem, red and green curves that crossed could be a puzzle. Red and blue or orange and blue are much less problematic. 2. Diagonal shading is meaningless here. Appropriate shading should just be monochrome. | |
| Aug 23, 2017 at 12:43 | answer | added | joceratops | timeline score: 6 | |
| Aug 23, 2017 at 12:42 | comment | added | whuber♦ | The shaded (signed) area is the difference in expectations of the two distributions. There are no "weights" in any parts of the area--it's just an area. "Signed" means that areas where the red curve is to the left of the green curve are subtracted from the total. (No such areas appear in this image). | |
| Aug 23, 2017 at 11:24 | review | First posts | |||
| Aug 23, 2017 at 11:45 | |||||
| Aug 23, 2017 at 11:20 | history | asked | thanks_in_advance | CC BY-SA 3.0 |