simChef
simChef copied to clipboard
Yu-Group
An R package to facilitate PCS simulation studies.
Create an issue template for new contributors, and link template to CONTRIBUTING.md
- **Naming**: We have many `create_*` functions that create `R6` objects like the `Experiment`, `DGP`, and `Method` classes. On the other hand, we used `init_docs()` for setting up the Rmd...
Add column `.time_taken` to output of `fit_experiment()`, indicating the time it took to run that method-DGP combination. This is to facilitate timing comparisons.
Motivated by #175, the [documentation](https://yu-group.github.io/simChef/articles/simulation-directory-setup.html) and [vignette](https://yu-group.github.io/simChef/articles/simulation-directory-setup.html) for `create_sim` could be improved. Here are some specific items: - [ ] Change "character" to "string" in the description of the `path`...
In https://github.com/Yu-Group/simChef/issues/175, @rcannood makes a great point: > In the context of reproducibility I would expect the rendered report to contain information on software versions of the used environments.
Create an example using multiple nodes on a Slurm cluster to run a simulation.
If `set.seed()` is called from within DGPs/Methods/Evaluators/Visualizers and `add_vary_across(.dgp = ...)` is used, the seed affects all replicates after the first vary-across parameter value. Consequently, the results are the same...
Caching does not work properly with DGPs (and Methods) that take in function arguments that explicitly contain a nested function. Example: ```r rm(list = ls()) library(simChef) library(magrittr) n
Add option to specify whether `ggplot2` objects should be displayed in an interactive way using `plotly::ggplotly()` in `render_docs()`. Only possible when rendering to html.