This Django web app is designed to "winnow the wheat from the chaff" in the Image Difference analysis of astronomical images.

Image Differencing will leave behind many fake objects due to bad subtraction that will be classified as bogus by a Machine Learning algorithm. This ML algorithm has to be 'trained' with labeled data and this website provides a web interface to generate this training set via human input classification.

Follow the following steps to download, install, and run this Django project.

git clone git@github.com:toros-astro/winnow.git

python manage.py makemigrations
python manage.py migrate

in order to set up the sqlite database.

You can populate the database with mock data for each app by typing:

python populate_winnow.py
python populate_rbmanager.py

python manage.py runserver

http://127.0.0.1:8000/

A live version of this site can be accessed by visiting: http://toros.utrgv.edu/winnow.

For a static preview, please visit: http://toros-astro.github.io/winnow.

The site will look similar to the following screenshot:

Copyright Martin Beroiz, 2015

email: martinberoiz@gmail.com