Subaru-PFS/spt_etc_webapp

PFS spectral simulator web app using PFS Exposure Time Calculator and Spectrum Simulator

PFS Spectral Simulator Web App

PFS spectral simulator web app using PFS Exposure Time Calculator and Spectrum Simulator.

Requirements

See requirements.txt for the latest dependencies.

For production:

• python (>=3.9)
• panel (>=1.2)
• astropy
• jinja2
• loguru
• matplotlib
• myst-parser
• numpy
• pandas
• setuptools
• synphot
• pfsspecsim @ git+https://github.com/Subaru-PFS/spt_ExposureTimeCalculator.git

For development:

• black
• ipython
• jupyter
• notebook
• pip
• ruff
• pyright
• isort
• specutils

For documentation:

• mkdocs
• mkdocs-material
• mkdocs-macros-plugin
• mkdocs-video
• fontawesome-markdown

Installation

First of all, please clone this repository. There are several ways to install the app.

Create a virtual environment

python3 -m venv .
source .venv/bin/activate

Local installation

# if you use pip
python3 -m pip install -r requirements.txt
python3 -m pip install -e .

# if you use pdm
pdm install

If you wish to build documentation, run the following command in the docs directory.

cd docs

# if you use pip
mkdocs build
cd ../

# if you use pdm
pdm run mkdocs build

The documentation is built under docs/site.

Docker container

If you have Docker installed, you can run the Docker image as follows.

docker run -it -p 8080:8080 --rm monodera/pfs_etc_web

Google Cloud Run

You can deploy the app to Google Cloud Run.

gcloud run deploy pfsetcweb --source .

Usage

Local installation

You can run the app by using the panel serve as follows.

# installed via pip
panel serve ./app.py --static-dirs etc-docs="./docs/site/"

# installed via pdm
pdm run panel serve ./app.py --static-dirs etc-docs="./docs/site/"

Then open http://localhost:5006/app in a web browser.

You can specify the number of threads used by the ETC using OMP_NUM_THREADS environment variable. A larger number of threads will enable to achieve faster running time, but reduce the per-thread efficiency of the computation. My experiment with AMD EPYC 7542 is summarized as follows.

OMP_NUM_THREADS	time (s)
1	274.09
2	158.04
4	82.05
8	45.94
16	23.91
32	13.17
64	7.99
128	4.83

Docker container

Open http://localhost:8080/app in a web browser.

Google Cloud Run

Open your app URL in a web browser.

License

MIT © monodera.