https://github.com/damian-oswald/process-rate-estimator
A GitHub repository for the process rate estimator developed for the Sustainable Agroecosystems Group at ETH.
https://github.com/damian-oswald/process-rate-estimator
agroecology biochemistry ethz nitrogen nitrous-oxide
Last synced: 7 months ago
JSON representation
A GitHub repository for the process rate estimator developed for the Sustainable Agroecosystems Group at ETH.
- Host: GitHub
- URL: https://github.com/damian-oswald/process-rate-estimator
- Owner: Damian-Oswald
- Created: 2023-09-28T15:39:34.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-08-12T15:49:02.000Z (about 1 year ago)
- Last Synced: 2024-08-13T16:05:44.070Z (about 1 year ago)
- Topics: agroecology, biochemistry, ethz, nitrogen, nitrous-oxide
- Language: R
- Homepage: https://damian-oswald.github.io/process-rate-estimator/
- Size: 224 MB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# :test_tube: Process Rate Estimator
This repository contains the source code and documentation for the process rate estimator. For a detailed (scientific) description of the process rate estimator, see Decock et al. ([2022](https://doi.org/10.5194/bg-2022-221)).
The project was done for the [sustainable agro-ecosystems group at ETH Zurich](https://sae.ethz.ch/). While the project documentation is presented on [this website](https://damian-oswald.github.io/process-rate-estimator/), the source code for the process rate estimator in contained in an R package called `PRE`. You can explore said package in [this GitHub repository](https://github.com/damian-oswald/PRE/). Meanwhile, this repository contains the scripts for running the process rate estimator on some collected data as well as the source code for reproducing the documentation page.
# :file_folder: The repository structure
This repository contains:
1. **The source code for the documentation page:** This repository contains the quarto documents necessary to render the documentation page (i.e., a bunch of `.qmd` and `.yml` files). Images, GIFs, bibliography etc. is in the `/resources` folder.
2. **The R scripts for running the process rate estimator:** The R scripts in the folder `/scripts` run some part of the process rate estimator. Specifically, they (1) prepare the collected data appropriately (`scripts/data-preparation`), (2) find the best hyper-parameters (`scripts/hypertuning`), (3) run the process rate estimator itself (`scripts/run-process-rate-estimator`), (4) run the sensitivity and uncertainty analyses (`scripts/sensitivity-analysis`).
Note that the latter two are subdivided into scripts that perform the calculation and scripts that generate visualizations based on the output.
# :question: How can I run the process rate estimator?
To run the process rate estimator, first install the R package `PRE`.[^readme-1]
[^readme-1]: This R package contains all functions that are repeatedly used throughout the workflow of this project. It's stored [on its own GitHub repository](https://github.com/Damian-Oswald/PRE/).
``` r
remotes::install_github("https://github.com/Damian-Oswald/PRE")
```
After successful installation, you can clone this GitHub repository.
``` bash
git clone https://github.com/Damian-Oswald/process-rate-estimator
cd process-rate-estimator
```
## :globe_with_meridians: Reproducing the documentation site
The [documentation site](https://damian-oswald.github.io/process-rate-estimator/) was built using the [scientific publishing system Quarto](https://quarto.org). To reproduce the documentation, first install Quarto, then run `quarto render` to render the entire site.
## :microbe: Running the main script
After successful installation of the R package `PRE` and cloning of this repository, you can run the main script, which will reproduce the results for the process rate estimations.
``` bash
Rscript scripts/run-process-rate-estimator/run-PRE.R
```
These results are also documented [here](https://damian-oswald.github.io/process-rate-estimator/results.html).
## :abacus: Running the sensitivity analysis
To reproduce the sensitivity analysis (which is documented [here](https://damian-oswald.github.io/process-rate-estimator/sensitivity-analysis.html)), enter the following command:
``` bash
Rscript scripts/sensitivity-analysis/sensitivitiy-analysis.R
```
> [!IMPORTANT]
> Running the sensitivity analysis R script is very compute intensive and may take a few days on an ordinary computer.
You can open the corresponding R script to change parameters such as `SAMPLESIZE` and `SAMPLEREPEAT`; however, doing this will change the exact numerical results.[^readme-2]
[^readme-2]: Although the *expected* results will not change.
The results written by the sensitivity analysis are also used by the uncertainty analysis.