dingo-enets

This reposity contains the embedding networks from [1], which are trained for the purpose of gravitational wave parameter estimation. If you find this code useful please cite [1].

Note: This is only a partial release of the code used in [1]. A more comprehensive package will be released in the near future. The present repository will not be maintained once the full package is publicly available.

Usage

Build a virtual environment and install dingo-enets.

$ python3 -m venv venv
$ source venv/bin/activate
$ pip install dingo-enets

Use dingo-enets to download and build a trained embedding network for a particular observing run. The model will be saved in </path/to/model_directory>.

> from dingo_enets import build_enet
> enet = build_enet(run="O1", detectors=["H1", "L1"], model_dir="</path/to/model_directory>")

> import torch
> input = torch.rand(10, 2, 3, 8033)
> output = enet(input)
> print(output.shape)

The function build_enet recognises whether a suitable model is present in the model directory, in which case it is not downloaded but instead loaded directly from disk.

References

[1] M. Dax, S.R. Green, J. Gair, J.H. Macke, A. Buonanno, B. Schölkops, Real-Time Gravitational Wave Science with Neural Posterior Estimation, Phys.Rev.Lett. 127 (2021) 24, 241103. [arXiv] [inspirehep]