Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

documentation improvements #362

Open
7 of 59 tasks
nhejazi opened this issue Sep 29, 2021 · 0 comments
Open
7 of 59 tasks

documentation improvements #362

nhejazi opened this issue Sep 29, 2021 · 0 comments

Comments

@nhejazi
Copy link
Member

nhejazi commented Sep 29, 2021

More work on documentation improvements. 😭

Learners to document:

  • CV_lrnr_sl: only requires an example (possibly wrapped in \dontrun{...} to avoid timing issues from running its examples). Maybe we can rename this learner (the capital "CV" is clunky and breaks the current naming pattern), perhaps Lrnr_sl_cv?
  • Lrnr_base: the variety of methods it exposes should be documented, both for developers and for users. This might also be a good place to give basic explanations of the x$train() and x$predict() methods inherited by all learners as well as any internal methods that might impact users.
  • Lrnr_bilstm: uses a documentation style that differs from that used by other learners and lacks an example. TODO: @imalenica
  • Lrnr_caret: missing documentation updates and examples.
  • Lrnr_cv: missing documentation updates and examples, as well as documentation for its helper functions.
  • Lrnr_dbarts: missing documentation updates and examples.
  • Lrnr_define_interactions: missing documentation updates and examples.
  • Lrnr_density_discretize: missing documentation updates and examples. TODO: @nhejazi
  • Lrnr_density_hse: missing documentation updates and examples. TODO: @nhejazi
  • Lrnr_density_semiparametric: missing documentation updates and examples. TODO: @nhejazi
  • Lrnr_grf: missing documentation updates and examples.
  • Lrnr_gts: missing documentation updates and examples.
  • Lrnr_h2o_glm: missing documentation updates and examples, as well as the same for the define_h2o_X model definition utility function.
  • Lrnr_h2o_grid and helper functions in the same file: missing documentation updates and examples.
  • Lrnr_hts: missing documentation updates and examples.
  • Lrnr_independent_binomial: missing documentation updates and examples.
  • Lrnr_multiple_ts: missing documentation updates and examples.
  • Lrnr_multivariate: missing documentation updates and examples.
  • Lrnr_nnet: missing documentation updates and examples.
  • Lrnr_optim: missing documentation updates and examples.
  • Lrnr_pca: missing documentation updates and examples. Maybe we just drop this?
  • SuperLearner interface, including Lrnr_pkg_SuperLearner, Lrnr_pkg_SuperLearner_method, and Lrnr_pkg_SuperLearner_screener all require documentation updates and examples.
  • Lrnr_polspline: missing documentation updates and examples.
  • Lrnr_pooled_hazards: missing documentation updates and examples. TODO: @nhejazi
  • Lrnr_randomForest: missing documentation updates and examples.
  • Lrnr_revere_task: requires some documentation updates. Maybe this doesn't need to be exported?
  • Lrnr_rpart: missing documentation updates and examples.
  • Lrnr_screener_augment: missing documentation updates and examples.
  • Lrnr_screener_coefs: missing documentation updates and examples.
  • Lrnr_screener_correlation: missing documentation updates and examples.
  • Lrnr_sl: missing documentation updates and examples. Could also point to the handbook chapter in the docs.
  • Lrnr_solnp: missing documentation updates and examples.
  • Lrnr_solnp_density: missing documentation updates and examples.
  • Lrnr_stratified: missing documentation updates and examples.
  • Lrnr_subset_covariates: requires some documentation updates. Maybe this doesn't need to be exported?
  • Lrnr_tsDyn: missing documentation updates and examples. TODO: @imalenica
  • Lrnr_ts_weights: missing documentation updates and examples.
  • Pipeline: missing documentation updates and examples. Could also point to the handbook chapter in the docs.
  • Shared_Data: requires some documentation updates. Maybe this doesn't need to be exported?
  • Stack: missing documentation updates and examples. Could also point to the handbook chapter in the docs.
  • Utility functions in debug_helpers.R. Examples wrapped in \dontrun{...} and some long-form description of how to use these tools would make them a lot more developer-friendly.
  • Utility functions in density_utils.R, look mostly like those appearing in haldensify's data-reformatting function(s). TODO: @nhejazi
  • importance and importance_plot: documentation looks good, need examples.
  • Could we convert importance_plot to use ggplot2? TODO: @rachaelvp
  • Utility functions in learner_helpers.R: documentation looks good but need examples if exported. Maybe these don't need to be exported? Some pointers to how/when to use these would still be useful.
  • Loss function definitions in loss_functions.R: documentation looks good but need examples. Very easy to do.
  • Metalearner definitions in metalearners.R: documentation looks ok, but would be helpful to have some description contrasting these for users to refer to. Also require examples if will remain exported.
  • Utility functions in multinomial_helpers.R: requires some documentation updates. Maybe this doesn't need to be exported?
  • prediction_plot: requires some minor documentation updates and examples.
  • process_data: would be great to have some documentation describing what this is doing, especially for developer reference purposes.
  • process_missing: I guess we can delete this file based on the comment at the top ("This file is deprecated")?
  • Utility functions in reindex_folds.R: some documentation would be very helpful to make sure that we can all easily make sense of what these are doing. Example required for the exported subset_folds, but this probably doesn't need to be exported?
  • sl3_revere_Task: requires some documentation updates. Maybe this doesn't need to be exported?
  • inverse_sample (sampling.R): requires some minor documentation touch-ups; probably doesn't need an example. Maybe this doesn't need to be exported?
  • sl3_Task: requires thorough documentation, including description of commonly used optional arguments (e.g., id, weights, folds); examples illustrating creation of different tasks, e.g., with folds respecting IDs, would be very helpful to users. Some description of less commonly used arguments (e.g., flag, row_index) would also be welcome.
  • pooled_hazard_task (survival_utils.R): requires some minor documentation updates. Maybe this doesn't need to be exported? Needs an example if it is to remain exported.
  • Utility functions in ts_utils.R: need the bare minimum of standard documentation.
  • Utility functions in utils.R: documentation is mostly good, some functions need minor touch-ups. Exported functions need examples, but probably don't need to be exported.
  • Variable_Type: explicit documentation would be very helpful in making clear how automatic variable type definitions are currently performed. Examples would help make this clear, both to users and current/potential developers. Questions routinely come up from new users about the option "sl3.pcontinuous".
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants