-
Option to activate or deactivate non-robust statistics (max, min, range, etc.)
-
The statistics we keep track of are currently generated on-demand and are then cached for a period of time. Ultimately, we should keep track of our statistics as they change, making it so that we'd (a…
-
Depth Statistics computes a suite of summary statistics. Our initial version works on a fixed window size, but user feedback has indicated a more general version which supported multiple window sizes…
-
Each constraint in a constrained scan is costed using the "random IO" cost, which is a 4x multiplier. When the estimated row count of the scan is very low, this can cause the optimizer to choose a sca…
-
If we will be sticking to paired-samples/one-samples tests then robust statistics could help us when there are small samples. We could implement robust group-level tests such as Yuen's t or a percenti…
Feugs updated
7 years ago
-
![image](https://github.com/user-attachments/assets/13e1d068-f558-4e5b-9fff-dc02f1d40ffc)
![image](https://github.com/user-attachments/assets/ac39bcf2-d63e-4eb1-a540-526cbbc4b73d)
java.lang.Illega…
-
Since sorting and I assume ranking will be part of stdlib, here is some related functionality to consider. In another issue I mentioned the median. The median is a special case of a trimmed mean, wher…
-
Author Name: **alobo -** (alobo -)
Original Redmine Issue: [5369](https://issues.qgis.org/issues/5369)
Redmine category:rasters
---
Optionally, median and median absolute deviance (mad) should be i…
-
For implementation it looks like we could reuse sandwich covariance.
For RLM: H1 is the analog of OLS "nonrobust" covariance
I haven't figured out if H3 is HC, it has a summation term that looks simi…
-
See `tmle` function in https://www.notion.so/basisresearch/Design-document-robust-statistics-with-influence-functions-0a3ecaab5b944e2fade867a6958039b7