A web-app built using flask following factory guidelines for Piedimonte et al. 2021. The app is served using gunicorn with nginx acting as a reverse proxy and is containerized using docker-compose.
The core of the app are two objective scoring models built to predict the probability of optimal (OPT; opt.py) and no growth residual (NGR; ngr.py) cytoreduction. Variables, or input metadata, in the prediction scripts have been renamed such that they are python compatible variables (eg. no spaces, separated by variables) and are identical to the variable names in forms.py.
These two files, along with jmp_score.py (obtained by being a licensed JMP 15.2+) need to be placed inside the flask/project/ folder for the app to work correctly.
The site can be accessed outside of KidNet and can be visited at either https://ipm-ml.ccm.sickkids.ca or https://ipm.ccm.sickkids.ca.
