FAIG
FAIG copied to clipboard
tg12
Backtesting trading strategy performance
For back-testing, We would need to get some historical data.... This would be VERY heavy on the API usage. Perhaps we could get this into some kind of database?
Yea, the market data fetched is supposed to be saved for later use.
How about keeping this issue opened, and wait until later when the project structure is better understood then we start making this saving function?
I've been thinking about this for a little while. I don't want to use a central-DB due to open read/write permissions, and while that'd help data sharing, I suspect would break IG's T&Cs. So something local. sqlite3 is near-perfect, but sql is a pain once you've experienced nosql done well, and this project screams nosql. However, it's also burdensome to expect people to setup mongodb or similar, and we really don't need the performance yet, so i think tinydb http://tinydb.readthedocs.io/en/latest/intro.html or similar is the way to go. I was going to implement this for market data until i realised that market dealhandling rules CAN change per-request, and don't count against the history api limit anyway.
@FlickeringIRoam why not consider dockerising the application, and running a database designed for timeseries like InfluxDB? Thus the entire stack could be bought up with docker-compose.
Yep, very valid point. I have no issues at all with distributing it as a container
On 4 Mar 2018 3:02 am, "Callam Delaney" [email protected] wrote:
@FlickeringIRoam https://github.com/flickeringiroam why not consider dockerising the application, and running a database designed for timeseries like InfluxDB?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/tg12/FAIG/issues/10#issuecomment-370198102, or mute the thread https://github.com/notifications/unsubscribe-auth/AbZaZRAhlWa097YMqw48UfbZzLrr04x0ks5ta1lTgaJpZM4SVCmO .