Ayush Patnaik
Ayush Patnaik
I have dumped most of the stuff from the slideshow. @smishr could you take it from here? This is the author's guide: https://juliacon.github.io/proceedings-guide/author/ The limit is 2 pages.
These is no test for: ```julia ratio(x::Vector{Symbol}, domain, design::AbstractSurveyDesign) ```
For every dataset included with the package, there should be a link to its data dictionary. The place to put this is the documentation.
```julia julia> dclus1 = SurveyDesign(apiclus1; weights=:pw, clusters=:dnum, popsize=:fpc) SurveyDesign: data: 183×45 DataFrame strata: none cluster: dnum [637, 637, 637 … 448] popsize: [757, 757, 757 … 757] sampsize: [15, 15,...
@arinbasu my answer in JuliaCon was wrong, factor analysis is a part of the core survey package in R. https://r-survey.r-forge.r-project.org/survey/html/svyfactanal.html Please let us know if you need us to prioritise...
In the show method, instead of showing all the strata, I could be better to just show the unique values of the vector. ```julia julia> dstrat = SurveyDesign(apistrat; strata=:stype, weights=:pw)...
If there are no clusters, a column called `false_clusters` is created. If there are no strata, a column called `false_strata` is created. Perhaps it's better to leave them as non...
Make a documentation page for the datasets included in the package and with their corresponding Julia code for creating a `SurveyDesign`.
In R `apiclus2.fpc2` isn't a simple vector. @iuliadmtru can you please take a look at this and update the file in assets accordingly?
Structures in the package have giant if and else ladders. This is because a user should be allowed to pass probs, weights or popsize, and they all can be vectors...