[Planning] Initial Todolist

[lrnv]

This is a first general todolist for the package. @rimhajal feel free to modify it as much as you want (it is yours). Do not hesitate to check the boxes that you think are ok.

• Datasets specific to the field
  • Mortality tables: slopop, frpop, survexp.us, survexp.fr (moved to RateTables.jl)
  • Datasets: colrec, others ?
• Net survival log-rank-tests
  • fit() interface (dont duplicate code if possible! )
  • Grafféo's log-rank-type test (R's rs.diff function)
  • Stratification
• Benchmarking
  • Check that our benchmark coverage of our code is good enough
  • Link to benchmark outputs in the README (including badge)
  • Show benchmarks in docs
  • Also benchmark Hakulinene / EdererI and EdererII ? Crude mortality ? Others ?
• Non-parametric net survival
  • fit() interface.
  • Pohar Perme
  • Ederer 1, Ederer 2 and Hakulinen variants.
  • Stratification correct
  • [Interface] Ensure that no nans are outputed (and remove the ones that are already there) #43
  • Crude Mortality
• Nessie function #49
• [FR] Net survival Ad-RMNST test  #53
• Documentation
  • Show out the fact that we have the same results as R (so compute the R versions too).
  • Describe the different methods mathematically in the docs, take a look at the docs of Copulas.jl for the formats, and at Nathalie's lecture for the mathematical content.
  • One complete case-study example: re-write Maja's practice fully in Julia, with the same data and the same graphs.
  • [Docs] Show how to use NetSurvival.jl from R through JuliaConectoR or JuliaCall  #55
  • [Docs] Clarify readme and add benchmarks #41
  • [Docs] Add timings w.r.t. R #24
• Testing
  • Write tests that compare our results to R on several datasets and several mortality tables.
  • [Tests] check coherence of the fit() interface. #48
• [FR] Parametric Net Survival  #54
• Registration of the package :)