PySide6 desktop demo: load NanoBerry edge CSVs from MASTER_SD/, fit a wafer circle, show X / Y / theta offset and sigma vs a calibration baseline. No live camera required.

Stack: Python 3.11+, PySide6, NumPy.

Quick start

python3 -m venv .venv && source .venv/bin/activate
pip install -e .
wafer-user          # or: python src/wafer_edge/App.py

Click Read Wafer to load the demo CSVs. Tests: python -m pytest tests/ (UI smoke needs a display).

Demo data

CSV files in [MASTER_SD/](MASTER_SD/). Names: {S|R}{1|2}{seq}P0.CSV — local/remote board, SPI camera, 4-digit sequence (e.g. S10540P0.CSV).

Default cameras: CsvSource.DEMO_CAMERA_PATHS. Change that dict or use parse_csv_filename() for other captures.

Calibration

First run creates baseline.json (gitignored). Key fields: center_mm, radius_mm (75), center_px, radius_px (pixel scale).

Layout

Path	Role
`src/wafer_edge/App.py`	Main window
`VisionEngine.py`	Circle fit + offsets (no Qt)
`CsvSource.py`	CSV loader
`ui/widgets.py`	Plot + metric cards
`ui/Theme.qml`	Styles (loaded via `ui/theme.py`)

Pipeline: DEMO_CAMERA_PATHS → load_camera_records_from_csv() → snapshot_from_records() → UI.