https://github.com/larsgeb/hmclab
https://github.com/larsgeb/hmclab
bayesian-inference geophysics hamiltonian-monte-carlo inversion markov-chain-monte-carlo tomography
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/larsgeb/hmclab
- Owner: larsgeb
- License: bsd-3-clause
- Created: 2019-04-15T15:43:06.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2023-09-22T09:06:22.000Z (over 1 year ago)
- Last Synced: 2025-03-13T22:46:14.354Z (3 months ago)
- Topics: bayesian-inference, geophysics, hamiltonian-monte-carlo, inversion, markov-chain-monte-carlo, tomography
- Language: Jupyter Notebook
- Size: 70.1 MB
- Stars: 25
- Watchers: 4
- Forks: 1
- Open Issues: 12
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# HMC Lab
## Andrea Zunino, Lars Gebraad, Andreas Fichtner
[](https://codecov.io/gh/larsgeb/hmclab) [](https://opensource.org/licenses/BSD-3-Clause) [](https://github.com/psf/black) [](https://github.com/larsgeb/hmclab/releases/latest)
**HMC Lab** is a numerical laboratory for research in Bayesian seismology, written in Python and Julia. Jump to [Docker one-command setup](#docker-one-command-setup).
- **Website:** https://hmclab.science
- **Python documentation:** https://python.hmclab.science
- **Source code:** https://github.com/larsgeb/hmclab
- **Docker image:** https://hub.docker.com/repository/docker/larsgebraad/hmclab
- **Bug reports:** https://github.com/larsgeb/hmclab/issues
It provides all the ingredients to set up probabilistic (and deterministic) inverse
problems, appraise them, and analyse them. This includes a plethora of prior
distributions, different physical modelling modules and various MCMC (and
other) algorithms.
In particular it provides prior distributions, physics and appraisal algorithms.
**Prior distributions:**
- Normal
- Laplace
- Uniform
- Arbitrary composites of other priors
- Bayes rule
- User supplied distributions
**Physics:**
- Linear equations
- Straight ray tomography
- 3d source location
- 2d elastic full-waveform inversion
- User supplied physics
**Algorithms:**
- Hamiltonian Monte Carlo (and variations)
- Random Walk Metropolis Hastings
- Stein Variational Gradient Descent
- Gradient descent
- Interfaces to non-linear optimization methods from SciPy
- Animated versions of various algorithms
To get staerting with the tutorial and example notebooks, one can use a single command
in Docker. This will pull a Docker image based on the Jupyter Datascience stack. The
final container is approximately 5GB.
```bash
docker run -p 9123:9123 larsgebraad/hmclab \
start-notebook.sh --NotebookApp.token='hmclab' \
--NotebookApp.port='9123' --LabApp.default_url='/lab/tree/Home.ipynb'
```
Then either copy-past the link from your terminal, or navigate manually to [http://127.0.0.1:9123/lab/tree/Home.ipynb?token=hmclab](http://127.0.0.1:9123/lab/tree/Home.ipynb?token=hmclab).
# Online tutorial notebooks
All tutorial notebooks can also be accessed online in a non-interactive fashion. Simply
use https://python.hmclab.science or use the following links:
**Tutorials:**
- [Getting started.ipynb](notebooks/tutorials/0%20-%20Getting%20started.ipynb)
- [Tuning Hamiltonian Monte Carlo.ipynb](notebooks/tutorials/1%20-%20Tuning%20Hamiltonian%20Monte%20Carlo.ipynb)
- [Separate priors per dimension.ipynb](notebooks/tutorials/2%20-%20Separate%20priors%20per%20dimension.ipynb)
- [Creating your own inverse problem.ipynb](notebooks/tutorials/3%20-%20Creating%20your%20own%20inverse%20problem.ipynb)
- [Running parallel Markov chains.ipynb](notebooks/tutorials/4%20-%20Running%20parallel%20Markov%20chains.ipynb)
**Demos:**
- [Sampling linear equations](notebooks/examples/Sampling%20linear%20equations.ipynb)
- [Sampling sparse linear equations](notebooks/examples/Sampling%20sparse%20linear%20equations.ipynb)
- [Locating quakes on Grimsvötn, Iceland](notebooks/examples/Locating%20quakes%20on%20Grimsvötn%2C%20Iceland.ipynb)
- [Elastic 2d FWI](notebooks/examples/Elastic%202d%20FWI.ipynb)
# The long way around: installing the package on your system
For full installation instructions, including creating a proper Python environment, [see the installation instructions](https://python.hmclab.science/setup.html).
Start with making sure that you have HDF5 or h5py installed properly.
Directly to your environment:
```
pip install -e [email protected]:larsgeb/hmclab.git@master#egg=hmclab
```
From the project root directory:
```
pip install -e .
```
### Development dependencies
If you want to develop within this repo, we recommend a few extra packages. They can also be installed using pip.
In Bash:
```
pip install -e [email protected]:larsgeb/hmclab.git@master#egg=hmclab[dev] # from github repo
pip install -e .[dev] # from local clone
```
... or Zsh (which requires escapes for brackets):
```
pip install -e [email protected]:larsgeb/hmclab.git@master#egg=hmclab\[dev\] # from github repo
pip install -e .\[dev\] # from local clone
```