Shikhar Mishra

WIP: Add general `CI` and integration with `mean`

3

EDIT: 26.02.23 Remaining to do on this PR. Integration with other summary functions (`total` etc) can be done in separate PR's. See their issues - [ ] Add doctests -...

WIP: Hartley Rao and other improvements

4

- Update docs with `add Survey` - Add back Horvitz thompson total with Hartley Rao variance estimation - Integrate HR with `total` with `SurveyDesign` - Add some analytical formulae for...

Draft old progress on raking. Function is not complete yet. Algorithm is about 50% there, plenty of work still to go. @sayantikaSSG can you link all of the references for...

Add Taylor Series linearisation

3

Add Taylor series linearisation for variance estimation for arbitrary `SurveyDesign` schemes - [ ] Add taylor series linearisation function(s) - [ ] Integrate with `SurveyDesign`

Variance Estimation for Multistage and other Complex designs

9

Calculating survey means and totals is relatively straighforward as compared to their variances in certain designs. For example in the Horvitz Thompson estimator, their is a double inclusion probability term...

`bydomain` is hardcoded for `bootstrap` variance

5

See line 22-27 in [by.jl](https://github.com/xKDR/Survey.jl/blob/main/src/by.jl) ```julia for i = 1:nd filtered_dx = filter(!isnan, Xt_mat[i, :] .- X.statistic[i]) push!(ses, sqrt(sum(filtered_dx .^ 2) / length(filtered_dx))) end replace!(ses, NaN => 0) X.SE =...

- [ ] push R code as comments in testing suite, if used R for checking "correctness" of Julia code

Thoroughly think about type stability and improved abstractions

1

To utilise Julian optimisations, we should rethink some aspects of the current codebase for improving type stability. I think a discussion around type hierarchy to improve modularity in code would...

Running the following in R: ```R > dclus2 dclus2_jk

Allow `fpc` argument in the constructor and implement finite population correction

2

TODO