Getting Started

Prerequisites

You need Python 3.11 and either conda/mamba or uv.

Installation

With conda + uv (recommended)

The FEniCSx chapters (14 and 16) require packages that are only available through conda-forge. Everything else is pure Python and handled by uv.

# 1. Create the conda environment (Python 3.11 + FEniCSx + pyvista)
conda env create -f environment.yml
conda activate fem-env

# 2. Install the workbench and remaining Python deps
uv pip install -e ".[lectures,dev]"

Without FEniCSx (core only)

If you only need the symbolic workbench and the companion notebooks (no FEniCSx chapters):

pip install -e ".[lectures]"

Verify the installation

import symbolic_fem_workbench as sfem

# Quick smoke-test: build the 1D bar element problem
result = sfem.workflow.build_bar_1d_local_problem()
print(result["Ke"])   # 2×2 symbolic stiffness matrix

Running the companion notebooks

# From the project root
jupyter lab notebooks/

Or render the full HTML site with output:

make notebooks-html

Build everything

make all          # notebooks (HTML + PDF) + book (HTML + PDF) + slides
make book         # lecture-note book only
make notebooks    # companion notebooks only
make slides       # revealjs lecture slides only

Package layout

src/
└── symbolic_fem_workbench/
    ├── __init__.py       # public API surface
    ├── symbols.py        # domains and symbolic fields
    ├── forms.py          # form containers
    ├── fe_spaces.py      # shape functions
    ├── reference.py      # reference elements + affine maps
    ├── quadrature.py     # integration rules
    ├── transforms.py     # IBP, Neumann, Galerkin substitution
    ├── extract.py        # element matrix / vector extraction
    ├── assembly.py       # global assembly helpers
    ├── elasticity.py     # constitutive matrices, B-matrix
    ├── workflow.py       # high-level end-to-end helpers
    ├── viz.py            # matplotlib plotting
    ├── codegen/          # I❤LA code generation
    └── printers/         # custom SymPy printers