pySEMTM0009

Small, transparent Python utilities for SEMTM0009 phase-plane analysis, simulation, equilibrium stability, and numerical continuation. The installed distribution is named pysemtm0009; the import package remains semtm0009.

Google Colab

For quick testing against the moving default branch:

!pip install -q "git+https://github.com/DRAGGON-Lab/pySEMTM0009.git"

P1 temporarily follows main while the new figure-output API is under pre-release testing. Pin the next reviewed full commit as soon as it is pushed; do not use the earlier 3faa173... revision because it predates that API.

Once a reviewed v0.1.0 tag and PyPI release actually exist, the preferred commands will be:

!pip install -q "git+https://github.com/DRAGGON-Lab/pySEMTM0009.git@v0.1.0"
!pip install -q "pysemtm0009==0.1.0"

Then import the stable teaching API:

from semtm0009 import (
    IZHIKEVICH_PRACTICAL_PARAMETERS,
    continue_equilibrium,
    continue_periodic_orbit,
    reduced_neuron,
)

Install before importing semtm0009. A tag or full commit hash prevents equations or numerical behavior changing silently between teaching sessions.

Local development and build

The supported interpreters are Python 3.12 and 3.13. In the standalone package repository, create an environment and run the same checks as CI:

python3.12 -m venv .venv
.venv/bin/python -m pip install -e ".[dev]" build twine
.venv/bin/python -m pytest
.venv/bin/python -m mypy
.venv/bin/python -m build
.venv/bin/python -m twine check dist/*

Runtime dependencies are NumPy, SciPy, and Matplotlib. The current continuation backend is included and does not require pyoomph.

Public continuation API

branch = continue_equilibrium(
    reduced_neuron,
    initial_state=[-65.953, 2.77e-4],
    parameter_name="I",
    params={**IZHIKEVICH_PRACTICAL_PARAMETERS, "I": 0.0},
    initial_parameter_step=0.05,
    arclength_step=0.15,
    max_points=100,
    parameter_bounds=(-1.0, 6.0),
)

Returned branches contain ordinary NumPy arrays for parameters, states or orbit envelopes, stability data, convergence flags, and bifurcation labels.

Notebook figure output

Use the public renderer factory to keep notebook setup cells short while still supporting inline teaching views and slide-ready files:

from pathlib import Path
from semtm0009 import configure_figure_output

render_figure = configure_figure_output(
    "both",  # "inline", "files", or "both"
    Path.cwd() / "figures" / "generated",
)

# After constructing a Matplotlib figure:
render_figure(fig, "phase-portrait")

The callable validates its configuration, creates the output directory when needed, writes PDF and PNG files at 180 dpi, displays an in-memory PNG in IPython for inline/both, and closes the figure. Named DRAGGON palette values are available as d_purple, d_orange, d_green, and d_red.

Synchronizing from the workshop

The workshop python/ directory is the upstream mirror. In the standalone repository, preview and then apply a synchronization with:

./scripts/sync_from_workshop.sh /path/to/semtm0009-python-workshops/python
./scripts/sync_from_workshop.sh --apply /path/to/semtm0009-python-workshops/python

Repository-only .github/, scripts/sync_from_workshop.sh, and graft/ are protected. Git metadata, environments, caches, builds, and bytecode are excluded. Review every dry run because synchronization uses deletion for unprotected mirrored paths.

Releasing with Trusted Publishing

The repository owner must configure a protected GitHub environment named pypi and a PyPI Trusted Publisher for project pysemtm0009, owner DRAGGON-Lab, repository pySEMTM0009, workflow publish.yml, and environment pypi. Create v0.1.0 only after CI is green on the reviewed release commit. The workflow uses OpenID Connect and should not contain a long-lived PyPI token.