SpaceY — Reusability Prediction API

FastAPI service that serves a tiny scikit-learn model (saved with joblib) to predict the probability that a booster is reusable. Includes tests and GitHub Actions CI so reviewers can verify it runs.

---

Why this repo matters

• Product thinking: turns a model into an HTTP API that anyone can call.
• Reproducible: artifact built from a script; environment pinned via requirements.txt.
• Quality: unit tests + CI on every push/PR.
• Clarity: minimal endpoints, typed payloads, and a 2-minute quickstart.

---

Quickstart

Windows PowerShell shown; macOS/Linux: replace the activate line with source .venv/bin/activate.

# 1) Setup
python -m venv .venv
. .venv/Scripts/activate
pip install -r requirements.txt

# 2) Build the demo model artifact
python scripts/train_dummy_model.py    # writes models/model.joblib

# 3) Run the API
uvicorn app.main:app --reload
# Open docs: http://127.0.0.1:8000/docs

Smoke test (new shell):

# Single prediction
curl -X POST http://127.0.0.1:8000/predict \
  -H "Content-Type: application/json" \
  -d "{\"features\":[0.1,0.5,1.2]}"

# Batch prediction
curl -X POST http://127.0.0.1:8000/predict-batch \
  -H "Content-Type: application/json" \
  -d "{\"batch\":[[0.1,0.5,1.2],[0.9,1.0,0.8]]}"

Endpoints GET /health → { "status": "ok" }

POST /predict → body:

{ "features": [0.1, 0.5, 1.2] }

response:

{ "reusability_probability": 0.462 }

POST /predict-batch → body:

{ "batch": [[0.1, 0.5, 1.2], [0.9, 1.0, 0.8]] }

response:

{ "reusability_probabilities": [0.462, 0.873] }

Run tests

# from repo root, with venv active
pytest -q

Project structure

The-SpaceY-project/
├─ app/
│  └─ main.py              # FastAPI app (uses lifespan to load the model)
├─ src/
│  ├─ model.py             # joblib loader + single/batch predict
│  └─ schemas.py           # pydantic request models
├─ scripts/
│  └─ train_dummy_model.py # builds models/model.joblib
├─ tests/
│  ├─ test_api.py          # health + single predict
│  └─ test_api_batch.py    # batch predict
├─ models/                 # generated artifacts (ignored in git)
├─ .github/workflows/
│  └─ ci.yml               # GitHub Actions (install → build artifact → tests)
├─ requirements.txt
├─ pytest.ini
└─ README.md

CI (GitHub Actions) On every push/PR to main, CI will:

Set up Python

pip install -r requirements.txt

python scripts/train_dummy_model.py

pytest -q

Badge:

![CI](https://github.com/GitDario79/The-SpaceY-project/actions/workflows/ci.yml/badge.svg)

If your workflow filename isn’t ci.yml, change the badge URL to match (e.g., api-ci.yml).

Notes & decisions Artifacts: models/*.joblib is git-ignored; re-build locally/CI via the script.

Startup: app uses FastAPI lifespan to load the model at startup (modern replacement for @on_event).

Typing: Pydantic models validate request bodies; responses return rounded probabilities for readability.

Troubleshooting Badge looks failing or “no status”:

Confirm the workflow file is at .github/workflows/ci.yml (or update the badge URL to your actual filename).

Confirm default branch is main.

Push any change to trigger a new run.

ModuleNotFoundError: app or src:

Ensure app/init.py, src/init.py exist and pytest.ini contains:

[pytest]
testpaths = tests
pythonpath = .

FileNotFoundError: models/model.joblib:

Run python scripts/train_dummy_model.py before starting the API or running tests.

License MIT

I wrapped a scikit-learn model as a FastAPI service, added schema validation, tests, and CI. The artifact is built in CI and the API exposes both single and batch prediction with docs at /docs.