Tutorials¶
These notebooks are maintained as deterministic, self-contained tutorials. They use only public Python APIs, small generated arrays, and no network access or credentials.
Before opening a notebook, install the published package and notebook tools in its kernel environment:
- First reconstruction compiles a model, simulates an object, runs alternating projection, and plots the result.
- Synthetic objects previews the in-memory object constructors and reconstructs one test target.
- Reconstruction diagnostics records convergence and Fourier coverage and renders overview plots.
Notebook execution is deliberately disabled during mkdocs serve and
docs-build, keeping documentation builds fast and deterministic. The source
notebooks have no stored large outputs. Their code cells are executed as tests
by python/tests/test_diagnostic_notebooks.py in the Python test suite. Those
offline checks build the in-tree package instead of downloading from PyPI.
Two more focused examples, python/examples/diagnostics_debug.ipynb and
python/examples/diagnostics_plots.ipynb, remain in the repository but are not
published in the primary tutorial path. They are maintained test fixtures for
advanced per-frame inspection and direct plot customization.