Deep-Learning picker based on the U-net architecture of PhaseNet. Final version to be fully embedded in SeisBench.
VERSION: 0.4.12
DATE: 11.2023
This version is the first fully fledged, stable release to be used
in combination with SeisBench (v0.4.0)
.
Soon it will be loaded to PyPI as well (and an easy-install through pip
provided as well)
In next releases, the framework will be available through PyPI.
At the moment, though, the installation is only possible from this repository.
Additionally, it is recommended to create a separate virtual environment with conda
or venv
(e.g., $ conda create -n dkpn python=3.9
). Though, this is taken care already by an
environment-file.
You can follow these simple commands.
$ git clone https://gitlab.rm.ingv.it/some/dkpn # use the master branch for stable releases
$ cd dkpn
$ conda env create -f dkpn_env.yml
$ conda activate dkpn
If problem installing (or unwanted package listed in the env), you could install the bare minimu as:
conda create -n dkpn python=3.9.12
conda activate dkpn
pip install numpy==1.24.3
pip install scipy==1.8.0
pip install torch==1.11.0 torchvision torchaudio
pip install seisbench==0.4.0
pip install pandas==1.3.2
pip install matplotlib==3.5.1
pip install seaborn
pip install obspy==1.4.0
pip install jupyterlab==3.3.3