Jupyter notebooks

These python notebooks can be found in the main code repository, under notebooks, and are fully interactable. We also provide Google collab extensions, which can be found in the introduction to each tutorial. They provide a high-level introduction to interfacing with harmonic, ranging from how to easily pick up harmonic and apply it to a project to more involved checkpointing and cross-validation for large scale, complex inferences.

One should note that harmonic acts on chains of posterior samples to infer the evidence, and is thus agnostic to the select sampler. In these notebooks we adopt the popular and highly accessible emcee package for sampling, though alternate (perhaps custom) samplers are equally applicable.

Tutorials:

• Basic Usage
• Checkpointing