Currently we use SQLAlchemy for interactions with the database for the webapp and the ingress functions, but for the Arima and GES models we use psycopg2 and JinjaSQL.
This necessitates different python environments, possibly leads to equivalent code doing different things, and makes testing a bit more awkward.
Investigate possibility of modifying dataAccess.py in the GES code to use SQLAlchemy. If this is straightforward, and successful, also use for Arima, and look into unifying the python environments.
Currently we use SQLAlchemy for interactions with the database for the webapp and the ingress functions, but for the Arima and GES models we use psycopg2 and JinjaSQL. This necessitates different python environments, possibly leads to equivalent code doing different things, and makes testing a bit more awkward.
Investigate possibility of modifying
dataAccess.pyin the GES code to use SQLAlchemy. If this is straightforward, and successful, also use for Arima, and look into unifying the python environments.