adding RSS and RMS #176
adding RSS and RMS #176
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #176 +/- ##
==========================================
+ Coverage 90.74% 90.77% +0.02%
==========================================
Files 23 23
Lines 2475 2482 +7
==========================================
+ Hits 2246 2253 +7
Misses 229 229 ☔ View full report in Codecov by Sentry. |
| float | ||
| Root Mean Square of the residuals. | ||
| """ | ||
| return np.sqrt(np.nanmean(np.square(self.model_residuals))) |
There was a problem hiding this comment.
Sorry if I'm blind right now, but is this not the same as Variogram.rmse?
We could update the the existing rmse with this implementation, which is way cleaner; or let Variogram.rmse just call your function.
There was a problem hiding this comment.
It seems like mean square of the residuals (RMS) could be the Root Mean Square Error (RMSE), however in the book 1 its not explicitaly menctioned ~ which is why I added it separately; similarly here is used with the same nomenclature 2.
It could be a typo in the book (?) and its just the RMSE, for me leaving RMSE is fine..
There was a problem hiding this comment.
I pointed my old Variogram.rmse to your Variogram.root_mean_square and all the tests still work, so in fact, it is the same thing.
Then, we keep both with Variogram.rmse simply calling Variogram.root_mean_square, as it has the much cleaner implementation.
| def rss(self): | ||
| return self.residual_sum_of_squares |
There was a problem hiding this comment.
I added this, cause I like to have both, the full name and the short name.
|
Thanks again. Just published as |
Contributions:
included and tested the metrics RSS and RSM, it is suggested in the reference below
Reference:
Oliver, M. A., & Webster, R. (2015). Basic steps in geostatistics: the variogram and kriging (No. 11599). Cham, Switzerland: Springer International Publishing.
Link: https://link.springer.com/book/10.1007/978-3-319-15865-5