Timeline for Problem in accepting R-squared (coefficient of determination) as a perfect measure of goodness of fit
Current License: CC BY-SA 3.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 31, 2023 at 14:41 | answer | added | Dave | timeline score: 0 | |
| Oct 31, 2023 at 14:10 | history | edited | Christoph Hanck |
edited tags
|
|
| Jun 2, 2017 at 13:34 | comment | added | Tim |
No, the reason why your examples differ is that they are not equivalent (!) and lead to different models as noted by @dbwilson. Notice that if you took y2 = x1 * 2 than you'd get the same $R^2$.
|
|
| Jun 2, 2017 at 13:32 | vote | accept | Hossein | ||
| Jun 2, 2017 at 13:19 | comment | added | Hossein | @Tim Thanks. So, one of the reasons of being far from perfect is that I mentioned in the problem? | |
| Jun 2, 2017 at 13:18 | comment | added | Hossein | @lucasfariaslf Yes I know the formula! I just don't understand why this is a perfect measure for goodness of fit. | |
| Jun 2, 2017 at 13:17 | comment | added | Tim | Nobody said it is perfect, it is actually far from perfect. | |
| Jun 2, 2017 at 13:17 | comment | added | Tim | Possible duplicate of Is $R^2$ useful or dangerous? | |
| Jun 2, 2017 at 13:13 | answer | added | dbwilson | timeline score: 3 | |
| Jun 2, 2017 at 12:26 | comment | added | Lucas Farias | Not an answer but a hint: have you checked the formula for $R^2$? | |
| Jun 2, 2017 at 12:18 | answer | added | Christoph Hanck | timeline score: 3 | |
| Jun 2, 2017 at 12:10 | history | asked | Hossein | CC BY-SA 3.0 |