Lab 1 — Reference simulation and scientific dataset

Goals¶

Deliver an SMIB simulator, event schedule, dataset manifest, and convergence check. Compare at least two numerical settings. Explain scenario-based splits. Retain numerical failures as flagged records.

Materials¶

• Notebook: lab01_reference_sim.ipynb · Running notebooks

• ·

• Teaching helper: src/sciml_ps/smib.py (simulate_smib, SMIBParams, FaultEvent)

Setup (local — canonical)¶

conda activate sciml_ps
cd /path/to/SciML_PS   # repository root
pip install -e .
jupyter notebook book/900_labs/lab_01_reference_sim/lab01_reference_sim.ipynb

Binder and Colab need the public GitHub repo; until then use local install. No ANDES required.

Deliverables¶

1. Plot comparing two solver settings on one fault scenario

2. Manifest JSON with scenario IDs, solver metadata, and train/test scenario split

3. Short note: whether stability labels flipped between solver settings

Linked chapters¶

• Reference simulation

• SMIB dynamics