A template for conducting scientific exploration using the research lifecycle and version control.
flowchart LR
subgraph Research Lifecycle
direction LR
Share:::nohighlight -.-> Discover
Discover:::nohighlight --> Plan
Plan:::nohighlight --> Acquire
Acquire:::nohighlight --> Process
Process:::nohighlight --> Analyze
Analyze:::nohighlight --> Preserve
Preserve:::nohighlight --> Share
end
classDef nohighlight fill:#fff,stroke:#333
The research lifecycle may be defined as follows (referencing the diagram above):
The Research Lifecycle is a heuristic model for understanding the steps of the process of scientific discovery (research), often represented in a wheel to emphasize the continuous nature of research with one project leading into the next. (https://nnlm.gov/guides/data-glossary/research-lifecycle)
Version control technologies (such as git) provide a consistent way to track, deploy, and collaborate on code associated with research, enabling and enhancing reproducible science. Providing a template repository for reference or direct use helps reduce costs associated with individual reinvention of common steps for any research process.
- Treat code as data necessary to the reproducibility of science for research project. (code becomes reproducibility outcome artifact)
- Treat documentation as code for research project. (documentation is treated with just as much importance as code and data)
- Reduce time to delivery of scientific results.
- Increase documentation of all steps, not just outcomes of scientific research project.
- Use software design principles to enhance scientific development agility.
- Incorporate automated linting for documentation, code, and data for early warning signs.
This repository makes various presumptions about code and related technologies used within. The following are used for example purposes but may be changed as necessary for more specific needs.
- Python >= v3.9
- Install Python
- Install Poetry
- Install Dagger
- Initialize Poetry dependencies:
poetry install - Initialize Dagger:
dagger project update
- Inspiration for this repository comes from work performed on the following Data Endeavor Template, which is itself a continuation of ideas from Microsoft's Team Data Science Process Project Template