Handle timeseries, how to model ?

[Franck DG]

Hello,

I am looking for best practices in managing the registration of timeseries. I cannot find a satisfactory model for Django. The problem is to model a driving simulator made up of a steering wheel, brake pedal, clutch, gear lever, ... Following a session, the machine provides a file with the times when the elements were used. Ex: the brake pedal was used at 30, 35, 50, 300 sec ... First gear increased to 5, 60, 300 sec ... There are several users, several machines and a user can do one or more sessions per day on any machine. All this must be recorded - a posteriori therefore - in a database in order to then use the data. Counting of events between 2 times, frequencies, evolution between sessions, ... I can't find the most efficient way to model this into objects. If anyone can help me that would be great, thanks.  

[Derek]

Hi

I am not sure about the links with Django, but we have found that TimescaleDB extension for Postgresql is a very powerful tool for managing and extracting analytics from  timeseries data.  Some further reading to see if this fits your use case:

https://stackoverflow.com/questions/25212009/django-postgres-large-time-series

https://medium.com/smart-digital-garden/timescaledb-and-django-5f2640b28ef1

https://pypi.org/project/django-timescaledb/

https://docs.timescale.com/latest/using-timescaledb

Derek