The Point of Interest Community Broker is a transient alert software (Rubin Observatory (LSST) Alert and Community Broker) that is currently tested with the ZTF alert stream. Incoming alerts will be processed, annotated, classified and forwarded.
For a general overview of LSST Community Broker, see https://www.lsst.org/scientists/alert-brokers
This repository contains the frontend (website) for the Point of Interest broker. The Point of Interest broker can be found at https://poibroker.uantof.cl/
Current and especially upcoming all-sky time-domain surveys, such as LSST, will deliver a vast amount of data each night, requiring for the developent of flexible, straightforward tools for the analysis, selection and forwarding of information regarding astrophysical transients and variable objects.
Our alert broker, called Point of Interest, is tailored towards the needs of astronomers looking for updated observations of variable stars in specific on-sky regions. Developed by a small team at Vanderbilt University, where I'm the main developer responsible for this project, this Point of Interest' alert broker should enable users to get updates on variable star observations from a straightforward, user-friendly web service. Data are processed in real time by big data/ machine learning algorithms and will be immediately available to the user community.
Point of Interest differs from other brokers in the focus on updates on variable stars, thus running a rather specific than the full analysis chain of streamed data. As a consequence, the broker is rather lightweight. Point of Interest users are encouraged to design their own on-sky regions they want receive updates for (such as for planned follow-up campaigns) or select from a list of on-sky regions which are particularly interesting for variable star observers, such as stellar streams, globular clusters and dwarf galaxies.
This repository contais the web frontend, including a small database for testing purposes.
After downloading the package, within the folder _web_frontend install all required packages:
pip3 install -r requirements.txt
Note
requirements2.txt shows the output of pip freeze and lists exact version numbers of the packages. However, this is mostly for reference. You may not be able to install these exact version on your specific system/environment.
Create a .env file in the project root based on the .env.example file and adjust the settings as needed.
Start the web app from the terminal from the project root (the folder that contains the sub-folder poi_boker) with
flask --app "poi_broker:create_app()" run --debug
or
flask --app "poi_broker:create_app()" run --no-debug --no-reload
In the web browser, enter http://127.0.0.1:5000/ to view the front-ent.
In case the website isn't displayed: do a
cat app.log
in your terminal window to see the correct URL
Also, inspect the browser developer console (F12) to see if there are any JavaScript errors, ie. a missmatching bokeh.min.js version.
The app can be profiled like any other python module, e.g.
python -m cProfile -o program.prof -m flask --app "poi_broker:create_app()" run --no-reload
but it's more useful to use the werkzeug ProfilerMiddleware to profile routes. Activate it in __init__.py if needed.