Road plan for the implementation
Opened this issue · 0 comments
tomaskrehlik commented
The package will implement the standard StatsBase
- Make use of the weighing functions available in the package
- Implement the abstraction layer
The following methods will be supported
- Weighted least squares using the Hampel, Huber, and BiSquare weighing functions and the S-estimation (basically just using the GLM machinery)
- MM-estimation
- Least trimmed squares
- Least weighted squares
Include standard datasets
- Phonecalls
- Engine data
- Heterogeneity simulation
Rich documentation
- Examples on the standard datasets
- IPython notebooks explaining each method