blab / ncov-forecasting-fit

Assessing accuracy of fitness model forecasts
1 stars 2 forks source link

Make `figures/` directory with individual .Rmd scripts #16

Open trvrb opened 11 months ago

trvrb commented 11 months ago

I’m using Rstudio rmd to generate the figures and have them organized with an outline where u can click to whichever figure to see how it is generated. The file is script_result_vis.rmd

It's fine to use R Studio instead of Jupyter to make your figures, but I'd still recommend making this clearer from a reproducible research standpoint. I'd suggest:

  1. Make a new top-level directory called figures/ in the repo
  2. Split out your current 3800 line (!) file script_result_vis.rmd into five files titled figure_1.Rmd, figure_2.Rmd, etc... that each produce just one figure.
  3. The PNG output from each of these files should be exactly what gets versioned in manuscript/figures/.
  4. Add a README.md under top-level figures/ directory that describes how to install necessary R packages (maybe with something like install.packages(c("dplyr", "ggplot2", "tidyverse")), but I don't know R very well at all) and how to load and run a file in R Studio.