Notebook guide

ForestFlow notebooks are paired Jupytext Python files and Jupyter notebooks. Paths are case-sensitive: the supported directory is notebooks/Tutorials.

Start here

notebooks/Tutorials/Tutorial_archive.ipynb

Load GadgetArchive3D, inspect training and testing snapshots, and compare measured P3D and P1D with an Arinyo fit.

notebooks/Tutorials/Tutorial_emulator.ipynb

Load the pretrained cINN, predict Arinyo coefficients, and compute spectra.

notebooks/Tutorials/Tutorial_Arinyo.ipynb

Evaluate the analytic Arinyo model and vary its coefficients.

notebooks/Tutorials/Tutorial_Pcross.ipynb

Compute transverse cross-power. Install the px optional dependencies.

Focused tutorials

notebooks/Tutorials/covariance

Finite-volume and leave-one-out covariance exercises.

notebooks/Tutorials/training

Emulator input preparation and leave-one-out validation. These workflows require training data and are not needed for ordinary prediction.

Other directories

notebooks/Figures

Publication-figure reproduction using paper-specific data products.

notebooks/priors

Arinyo, cosmological, and IGM prior studies.

notebooks/emulator

Emulator covariance and derivative diagnostics.

notebooks/developers

Exploratory, historical, or maintenance workflows; APIs and external paths here may not be supported.

Synchronize one pair after editing its Python source with:

jupytext --sync notebooks/Tutorials/Tutorial_emulator.py

Run make notebooks from the repository root to synchronize every non-checkpoint notebook.