https://github.com/viktor-shcherb/reproducible-research-project-template
A Docker-based template for reproducible Python research, integrating pip‑tools dependency pinning, pytest, and Codecov reporting
https://github.com/viktor-shcherb/reproducible-research-project-template
cicd codecov data-science docker docker-compose pip-tools pytest python reproducible-research template
Last synced: about 2 months ago
JSON representation
A Docker-based template for reproducible Python research, integrating pip‑tools dependency pinning, pytest, and Codecov reporting
- Host: GitHub
- URL: https://github.com/viktor-shcherb/reproducible-research-project-template
- Owner: viktor-shcherb
- License: cc0-1.0
- Created: 2025-07-23T15:01:52.000Z (11 months ago)
- Default Branch: master
- Last Pushed: 2025-07-27T12:47:52.000Z (11 months ago)
- Last Synced: 2025-08-28T04:20:52.233Z (10 months ago)
- Topics: cicd, codecov, data-science, docker, docker-compose, pip-tools, pytest, python, reproducible-research, template
- Language: Dockerfile
- Homepage:
- Size: 44.9 KB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Research Project Title
[](https://codecov.io/gh/OWNER/reproducible-research-project-template)
## Overview
## Reproduction Steps
---
## Environment Setup
This project is fully containerized with Docker to guarantee bit‑for‑bit reproducibility. You can either drive everything from the CLI or hook into your IDE (PyCharm) for an integrated experience. Follow the instructions that match your workflow.
### Prerequisites
* **Docker** (Engine & CLI) installed and running
* **Docker Compose** (if not bundled with Docker)
* A local checkout of this repository
---
### 1. CLI‑Based Workflow
> **Assumes** you have `docker` and `docker-compose` in your `$PATH` and that you’re in the project root.
1. **Build the dev image**
```bash
docker-compose build
```
2. **Launch the container**
```bash
docker-compose up -d
```
* Your project directory is bind‑mounted into `/app` inside the container.
* A persistent pip cache volume speeds up installs.
3. **Run commands inside the container**
```bash
# Open a shell
docker-compose exec app bash
# From within: install extras, run tests, start analysis, etc.
pytest --cov=common --cov-report=term-missing
# Coverage results will be uploaded to Codecov by the CI pipeline
python scripts/your_analysis.py --input data/data.parquet --output results/
# Or exit when you’re done:
exit
```
4. **Stop & clean up**
```bash
docker-compose down
```
---
### 2. PyCharm‑Based Workflow
> **Assumes** you’re using PyCharm Professional (with Docker support).
**Add a Docker‑Compose interpreter**
* **Preferences → Project → Python Interpreter → ⚙️ → Add… → Docker Compose**
* Select:
* **Compose file:** `docker-compose.yml`
* **Service:** `app`
* **Python interpreter path:** `/usr/local/bin/python3`
* Click **OK** and let PyCharm probe the container.
---
### 3. Syncing Dependencies (`sync_dependencies.sh`)
Whenever you update your `requirements/*.in` files, you need to recompile them into `.txt`, rebuild your Docker images, and restart the services. Instead of doing these steps by hand, use the provided `sync_dependencies.sh` script:
```bash
# From the project root:
./sync_dependencies.sh
```
This script will:
1. Run `pip-compile` inside the Docker environment to update `requirements/*.txt` (ensuring the correct Python/OS context).
2. Rebuild your Docker Compose images.
3. Redeploy all services (with `--remove-orphans` to clean up any old containers).
After it finishes, your dependencies and containers will be fully up-to-date without manual intervention.
---
This repository uses the following [template](https://github.com/viktor-shcherb/reproducible-research-project-template). See [TEMPLATE.md](TEMPLATE.md) for instructions on customizing the template.