lfrerot
/
good_simulation_practices
1
1
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update Research simulation

Lucas FRÉROT 2024-12-15 14:32:36 +00:00
parent 6e8ef99c27
commit d8056fe707
1 changed files with 45 additions and 38 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
Research-simulation.md

@ -1,38 +1,45 @@
# Principles of code and data management # Principles of code and data management
This page lists **mandatory** rules to be followed when working with code and This page lists **mandatory** rules to be followed when working with code and
data. data.
## Working with code ## Working with code
The objectives of these rules is to ensure: The objectives of these rules is to ensure:
1. Code developped in the lab is preserved. 1. Code developed in the lab is preserved.
2. Code can be easily shared in the lab and outside. 2. Code can be easily shared in the lab and outside.
3. Code can be easily reused by other people *(this includes future you!)*. 3. Code can be easily reused by other people *(this includes future you!)*.
4. Simulations are reproducible. 4. Simulations are reproducible.
### Tenets ### Tenets
1. Every simulation and post-processing code must be versionned in a 1. Every simulation and post-processing code must be versioned in a
[Git](https://git-scm.com/book/en/v2) (or similar) repository on [Git](https://git-scm.com/book/en/v2) (or similar) repository on
[git.dalembert.umpc.fr](https://git.dalembert.upmc.fr) (or similar). [git.dalembert.umpc.fr](https://git.dalembert.upmc.fr) (or similar).
[Commit](https://git-scm.com/book/en/v2/Git-Basics-Recording-Changes-to-the-Repository) [Commit](https://git-scm.com/book/en/v2/Git-Basics-Recording-Changes-to-the-Repository)
messages must be meaningful. messages must be meaningful.
2. Every repository must have a `README` file explaining what is the purpose of 2. Every repository must have a `README` file explaining what is the purpose of
the code, what are the dependencies, and how to run the code, what the code the code, what are the dependencies, and how to run the code, what the code
produces, how to run tests. produces, how to run tests.
3. Every library repository must contain tests. Running the tests should be done 3. Every library repository must contain tests. Running the tests should be done
with a single command. with a single command.
4. Every library must have a documented API, i.e. each function of the API has a 4. Every library must have a documented API, i.e. each function of the API has a
basic description of what it does, along a description of inputs and outputs. basic description of what it does, along a description of inputs and outputs.
5. Every library must have usage examples. 5. Every library must have usage examples.
6. Code versions used in a publication must be saved on 6. Code versions used in a publication must be saved on
[SoftwareHeritage](https://archive.softwareheritage.org/save/) and the [SoftwareHeritage](https://archive.softwareheritage.org/save/) and the
resulting SWHID cited in the publication. resulting SWHID cited in the publication.
## Working with data ## Working with data
The objective of these rules is to enure: The objective of these rules is to ensure:
1. Data produced in the lab is preserved. 1. Data produced in the lab is preserved.
2. Data can be easily shared in the lab and outside. 2. Data can be easily shared in the lab and outside.
3. Data can be easily resued by other people *(this includes future you!)*. 3. Data can be easily reused by other people *(this includes future you!)*.
4. Data origin can be traced. 4. Data origin can be traced.
### Tenets
1. Your `$HOME` must have scheduled daily backups on an external drive / remote server. Periodically make sure backups are working and can be recovered.
2. Simulation data for a workflow / pipeline / paper is grouped in a dataset. Datasets must be documented with a `README` file explaining what is the data, how it was generated and how it can be used.
3. Datasets must be (as much as possible) published to Zenodo at the time of submission, and the dataset DOI cited in the article.
4. Open-source file formats must be used to store data and metadata.
5. All datasets must be uploaded to ... when leaving the lab.