blackjax
PyMC
blackjax | PyMC | |
---|---|---|
1 | 3 | |
727 | 8,202 | |
3.3% | 1.2% | |
8.2 | 9.5 | |
3 days ago | 5 days ago | |
Python | Python | |
Apache License 2.0 | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
blackjax
-
AutoBNN: Probabilistic Time Series Forecasting
What are the other four frameworks?
> For one, who wants to do stuff in tensorflow anymore let alone tensorflow-probability.
AutoBNN is a JAX library and has nothing to do technically with TF Probability. It was developed by the TF Probability team.
> DL community prefers pytorch and stats community prefers Stan.
It looks like the JAX ecosystem for stats is growing: NumPyro is based on JAX, PyMC has a JAX backend, https://github.com/blackjax-devs/blackjax has effective samplers, there is https://github.com/jax-ml/bayeux, and now AutoBNN.
> This one seems theoretically more interesting than some others but practically less useful.
Are there other factors why you think AutoBNN is not practically useful, apart from being based on the wrong foundation (which was a mistaken belief of yours)?
PyMC
- PYMC Release: v5.0.0
-
An Astronomer's Introduction to NumPyro
I believe the pymc versions were resolved into developing version 4 of pymc. Development at https://github.com/pymc-devs/pymc
It still depends on theano now evolved and renamed
-
What is Probabilistic Programming?
This tutorial explains what is probabilistic programming & provides a review of 5 frameworks (PPLs) using an example taken from Chapter 4 of Statistical Rethinking by Dr. Richard McElreath. Frameworks (PPLs) reviewed are - Stan (https://mc-stan.org/) PyMC3 (https://docs.pymc.io/) Tensorflow Probability (https://www.tensorflow.org/probability) Pyro/NumPyro (https://pyro.ai/) Turing.jl (https://turing.ml/stable/) I also provide the basic review of a great library called arviz (https://arviz-devs.github.io/arviz/), which can be used for all the above-mentioned PPLs to do Exploratory Data Analysis of Bayesian Models. Here is the link to the notebook in which I have implemented the example model using the above Frameworks/PPLs https://colab.research.google.com/drive/1zgR2b0j2waGi1ppnIe1rw7emkbBXtMqF?usp=sharing
What are some alternatives?
statsmodels - Statsmodels: statistical modeling and econometrics in Python
Dask - Parallel computing with task scheduling
stan - Stan development repository. The master branch contains the current release. The develop branch contains the latest stable development. See the Developer Process Wiki for details.
Numba - NumPy aware dynamic Python compiler using LLVM
SymPy - A computer algebra system written in pure Python
pyro - Deep universal probabilistic programming with Python and PyTorch
SciPy - SciPy library main repository
zipline - Zipline, a Pythonic Algorithmic Trading Library
Pandas - Flexible and powerful data analysis / manipulation library for Python, providing labeled data structures similar to R data.frame objects, statistical functions, and much more
harold - An open-source systems and controls toolbox for Python3
NumPy - The fundamental package for scientific computing with Python.
RDKit - The official sources for the RDKit library