Skip to content

Astroinformatics/BayesianComputing

Repository files navigation

Bayesian Computing & Hierarchical Modeling Labs

Astroinformatics Summer School 2022


This repository contains several computational notebooks:

  • monte_carlo.jl (Pluto notebook): Compares methods for performing numerical integration
  • app_hbm_galaxy_evolution.ipynb (Jupyter notebook): Applies hierarchical Bayesian modeling to make inferences about galaxy evolution.
  • ppl_intro.jl (Pluto notebook): Introduces the use of probabilistic programming languages (PPLs) using Turing.jl
  • ppl_hbm.jl (Pluto notebook): Demonstrates performing inference with hierarchical Bayesian models using the Turing PPL

monte_carlo.jl is mostly independent of the other notebooks in this lesson and could reasonably be done first or last. We recommend that students complete ppl_intro.jl before ppl_hbm.jl.
app_hbm_galaxy_evolution.ipynb and ppl_hbm.jl could be worked in either order.

Files ending in .jl are Pluto notebooks written in Julia and files ending in .ipynb are Jupyter notebooks written in Python. Labs do not assume familiarity with either language. While it can be useful to "read" selected portions of the code, the lab tutorials aim to emphasize understanding how algorithms work, while minimizing need to pay attention to a language's syntax.


Running Labs

Instructions will be provided for students to run labs on AWS severs during the summer school. Below are instruction for running them outside of the summer school.

Running Pluto notebooks on your local computer

Summer School participants will be provided instructions for accessing a Pluto server. Others may install Julia and Pluto on their local computer with the following steps:

  1. Download and install current version of Julia from julialang.org.
  2. Run julia
  3. From the Julia REPL (command line), type
julia> using Pkg
julia> Pkg.add("Pluto")

(Steps 1 & 3 only need to be done once per computer.)

  1. Start Pluto
julia> using Pluto
julia> Pluto.run()
  1. Open the Pluto notebook for your lab

Running Jupter/Python notebooks

Summer School participants will be provided instructions for accessing JupyterLab server.
Others may install Python 3 and Jupyter (or JupyterLab) on their local computer or use Google Colab to open the Jupyter notebooks.


Additional Links

Contributing

We welcome people filing issues and/or pull requests to improve these labs for future summer schools.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •