harp-ww-feb2023

Example code and projects for the harp working week at DMI in February 2023.

Rendered versions of the examples can be found on the harp Training Course website.

Before beginning, it is highly recommended that you have a Github Public Access Token (PAT). This will prevent Github from throttling downloads based on too many files and / or lack of authentication. You will need a Github account and you can follow the instructions in Happy Git with R for how to set it the PAT from within R.

This repository is for reproducing spatial verication examples that were shown during the working week. The repository is set up as follows:

├── data
│   ├── meps - deterministic forecast from the MEPS model - file only contains precipitation
│   ├── met_analysis - Nordic 1km precipitation analysis
├── harphub - project directory using the harpSpatial version at harphub
│   ├── renv.lock - lockfile to restore project: use renv::restore() from this directory
│   ├── harphub_test.Rmd - notebook markdown
├── harpspatialteam - project directory using the development harpSpatial version at harpSpatialTeam
│   ├── renv.lock - lockfile to restore project: use renv::restore() from this directory
│   ├── harphub_test.Rmd - notebook markdown
├── ├── conf
├── ├── ├── my_conf.R -configuration file for running spatial verification
├── ├── ├── my_domain.rds - verification domain

In order to get up and running, you first need to restore the projects from the lockfiles. You should do this in separate R sessions and run the projects in separate R sessions as well.

Brief Instructions

To restore a project, first make sure that you have the renv package installed. Navigate to one of the project directories (harphub or harpspatialteam) and start R. Then at the command line type renv::restore(). The correct versions of all of the packages will be installed to the project. You may get a warning about the R version - this only matters if the R version is older than 3.5. Do the same for the other project. Note that the installation could take a while for the first project, but should be quick for the second as it will use a linked cache for the packages.

The notebooks can be rendered with rmarkdown::render(), or by clicking on Preview in RStudio. Note that you will need to have the pandoc system library available to render markdown documents. Pandoc is shipped with RStudio, so outside of RStudio you will need to get the library installed.

As an exercise, you may want to take the code in the notebooks and rewrite as a standalone R script.

Detailed Instructions for the harphub project (do exactly the same for harpspatialteam)

After cloning the repository

RStudio

• Open RStudio
• File > Open Project > Directory where you installed the repository > harp-ww-feb2023 > harphub > harphub.Rproj
• In the console pane: renv::restore()
• Say yes to activating project
• Say yes to proceeding with installation of packages
• Wait (in general warnings can be ignored)
• File > Open File > harphub_test.Rmd
• In Source pane (normally top left) > Run > Run All
• When finshed.. In Source Pane > Preview

R from Terminal

• cd to path/to/repository/harp-ww-feb2023/harphub
• Start R
• renv::restore()
• Say yes to activating project
• Say yes to proceeding with installation of packages
• Wait (In general warnings can be ignored)
• rmarkdown::render("harphub_test.Rmd")
• When finished open harphub_test.nb.html in a web browser