documentation

[codecov[bot]]

The author of this PR, rimhajal, is not an activated member of this organization on Codecov. Please activate this user on Codecov to display this PR comment. Coverage data is still being uploaded to Codecov.io for purposes of overall coverage calculations. Please don't hesitate to email us at support@codecov.io with any questions.

[lrnv]

@rimhajal I took a quick look and this looks great for the moment. If you make new files, dont forget to add them to make.jl.

Btw, remember that you can visualize the docs locally: 1) activate the docs folder via ] activate docs 2) using LiveServer 3) servedocs()

[rimhajal]

Thank you for the note on make.jl, it slipped my mind! And yes I'm using the live documentation it's been extremely helpful

[lrnv]

@rimhajal note that you can also include references in the autodocs, see e.g. there https://github.com/lrnv/Copulas.jl/blob/8f47a0589efc49210f3d2177f306eb3bef5e90a6/src/Generator/WilliamsonGenerator.jl#L1-L39

The syntax is abit trickier than in the docs, since when rendered from the REPL these documentations do not pass through bibtex, so you need both the name and the textual name of the reference. Would be great to at least give the PP ref in the PP autodocs, and Grafféo's ref in Grafféo's autodocs

[lrnv]

failluire is because "Error: Key "PoharPerme2014" not found in entries from /home/runner/work/NetSurvival.jl/NetSurvival.jl/docs/src/assets/references.bib" -> maybe you wanted PoharPerme2012 ?

[lrnv]

@rimhajal I partially rewrote the home page and the theory page, did not touch the getting started page yet. Its taking shape !

By the way, the abstract / details of the juliacon submission could be used to enhance the home page (if you think it's relevant). They did go like this:

Abstract:

The NetSurvival.jl package brings standard relative survival analysis modeling routines in Julia. Relative survival analysis is a branch of survival analysis where individuals are subject to two competing risks, but the cause of death is unknown, often for data quality reasons. In these circumstances, standard competing risks approaches are unusable and specific estimators and methods are used (e.g. Pohar-Perme net survival estimator, Graffeo's log-rank-type test among others).

Details:

Data arising from large cancer studies often lack reliable information about the cause of death (supposed binary: death from the studied cancer or from other causes) for each individual. Relative survival analysis is a subfield of survival analysis specifically targeted at this particular type of datasets. The field's goal is to extract survival curves that only takes into account the deaths from cancer, the so-called net survival curves, for comparison purposes between different groups or directly for diagnostic purposes. For that, a few standard estimators were established in the last 50 years, backed by a wide literature.

Standard tools nowadays are composed of R packages, with underlying C and C++ routines, that are hard to read, maintain, and globally use. This package is an attempt to provide a fast, reliable, but most importantly easily maintainable package to implement standard estimators and routines from the field onto the StatsModels.jl API. Our hope is that the junction with classical modeling API in Julia will allow later extensions of the existing modeling methods, with a simple interface for the practitioners.

In this talk, we present the current state of the implementation: the few tools and methods that were implemented, their integration into the Julian ecosystem, and we showcase the functionalities and performance of the implemented methods. We particularly note that the use of native Julia allows for concision and readability of the code.

[rimhajal]

I am unsure what to add for the examples because there are too many graphs that would be useful and we haven't done that yet ... I think I will try to move on with the code before finishing the doc on the examples.

[lrnv]

You have two time "This package is an attempt to..." maybe the two sentences should be merged. Then I agree with you that more implementation of graphs and other methods should be done before we can sketch the examples, so maybe you can simply merge this PR and work on the graphs and/or something else.