Skip to content

Latest commit

 

History

History
37 lines (28 loc) · 1.54 KB

README.md

File metadata and controls

37 lines (28 loc) · 1.54 KB
Raw

Data Assimilation

dass is tool for learning about data assimilation / history matching created by the developers of ERT. It is inspired by DAPPER and HistoryMatching.

It includes implementations of Ensemble Smoother (ES) as given in [1], see dass/analysis.py. The implementation of ES can easily be extended to the Ensemble Smoother with Multiple Data Assimilation (ES-MDA) as described in [2].

For notebooks with examples and tutorials see the notebooks/ folder.

NB! notice that there are no .ipynb files in the notebooks/ folder. This is because we use Jupytext to sync .py and .ipynb files, which means that we only need to keep the .py files in source control.

Installation

git clone https://github.com/equinor/dass.git
cd dass
# dass supports Python 3.8 and above.
python3.9 -m venv .venvdass
source .venvdass/bin/activate
# Add -e if you want to make changes.
pip install -e .
# Install additional requirements for developers.
pip install -r dev-requirements.txt
# Start jupyter notebook
jupyter notebook
# To make sure everything works, run on the of the notebooks in the notebooks/ folder.

References

[1] - Data Assimilation The Ensemble Kalman Filter

[2] - Ensemble smoother with multiple data assimilation