https://github.com/beekill95/numpyro-doing-bayesian
My implementation of John K. Kruschke's Doing Bayesian Data Analysis 2nd edition using Python and Numpyro.
https://github.com/beekill95/numpyro-doing-bayesian
bayesian-data-analysis bayesian-inference dbda2e hierarchical-models kruschke mcmc numpyro probabilistic-programming
Last synced: about 2 months ago
JSON representation
My implementation of John K. Kruschke's Doing Bayesian Data Analysis 2nd edition using Python and Numpyro.
- Host: GitHub
- URL: https://github.com/beekill95/numpyro-doing-bayesian
- Owner: beekill95
- License: mit
- Created: 2022-05-14T14:23:25.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2024-04-23T04:19:53.000Z (about 1 year ago)
- Last Synced: 2025-03-24T09:46:41.294Z (2 months ago)
- Topics: bayesian-data-analysis, bayesian-inference, dbda2e, hierarchical-models, kruschke, mcmc, numpyro, probabilistic-programming
- Language: Python
- Homepage:
- Size: 190 MB
- Stars: 7
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# numpyro-doing-bayesian
My implementation of John K. Kruschke's
[Doing Bayesian Data Analysis 2nd edition](https://sites.google.com/site/doingbayesiandataanalysis/what-s-new-in-2nd-ed)
using Python and Numpyro.
This implementation is not comprehensive,
I'll just focus on the generalized linear model only,
which is from chapter 16 onward.
Suggestions for improvement are welcome!
## Chapters
* [Chapter 9: Hierarchical Models](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_09)
* [Chapter 10: Model Comparison and Hierarchical Modeling](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_10)
* [Chapter 12: Bayesian Approaches to Testing a Point ("Null") Hypothesis](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_12)
* [Chapter 16: Metric-Predicted Variable on One or Two Groups](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_16)
* [Chapter 17: Metric Predicted Variable with one Metric Predictor](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_17)
* [Chapter 18: Metric Predicted Variable with Multiple Metric Predictors](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_18)
* [Chapter 19: Metric Predicted Variable with One Nominal Predictor](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_19)
* (WIP) [Chapter 20: Metric Predicted Variable with Multiple Nominal Predictors](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_20)
* [Chapter 21: Dichotomous Predicted Variable](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_21)
* [Exercise 21.3: Heterogeneous Concentration Parameters](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_21_exercise_21_3)
* [Chapter 22: Nominal Predicted Variable](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_22)
* [Chapter 23: Ordinal Predicted Variable](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_23)
* (WIP) [Exercise 23.2: Handle Outliers](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_23_exercise_23_2)
* [Chapter 24: Count Predicted Variable](https://www.nguyenmbquan.space/numpyro-doing-bayesian/chapter_24)
## Dependencies
This project uses a combination of [Conda](https://docs.conda.io/en/latest/)
and [Poetry](https://python-poetry.org/) for dependencies management.
To install the dependencies for this project, make sure that you have `conda` installed on your system.
First, create a virtual environment managed by `conda`:
```
conda env create -f environment.yml
```
The above command will create a virtual environment named `doing_bayes`
and install `poetry` package manager into that environment.
After that, activate the environment `conda activate doing_bayes`
and use `poetry` to install the remaining dependencies:
```
poetry install
```
## Jupyter Notebook
Activate the `doing_bayes` environment,
and then start the `jupyter-lab` server:
```
jupyter-lab --no-browser
```
Then, you can click on the link to open notebooks on your browsers.
Each chapter's notebook are a normal python script thanks to [Jupytext](https://jupytext.readthedocs.io/en/latest/).
To generate a notebook for a chapter from the python script, you can follow this [instruction](https://jupytext.readthedocs.io/en/latest/paired-notebooks.html#how-to-open-scripts-with-either-the-text-or-notebook-view-in-jupyter).
## Credits
My implementation refers to [JWarmenhoven's implementation](https://github.com/JWarmenhoven/DBDA-python) a lot,
especially those figures with data and posterior predictive distributions.