(09-22-2018, 09:31 AM)CryptoCoeus Wrote: (09-18-2018, 10:44 PM)cubit Wrote: From what you've said I want to ask if you've actually imported some data via "Local data" / "Import more data", but probably I just don't get what you are
asking!
Could you maybe elaborate a little?!
Yeah , I do import "local data". But the local data & papertrader data time do not continuous all the time. Fe, I have local data for BTC/XRP from May 15th to Sep 15th. But if I run papertrader today, what will happen to market data from Sep 15th to DateTime(Now) ?
But I was testing the papertrader for a time, and I believe it is doing what I suggested. It downloads the warmup period to run the papertrader on DateTime(Now).
Please let me know, If I am wrong.
To the best of my knowledge (I am a gekko noob!) the papertrader can only run on the data you've already imported.
The tradebot is different in that it sets up a watcher which brings down the data as you need it.
So I guess the answer is to run a new "live gekko" of market watcher type, which will provide you with an ongoing stream.
To answer your specific question "Fe, I have local data for BTC/XRP from May 15th to Sep 15th. But if I run papertrader today, what will happen to market data from Sep 15th to DateTime(Now) ?"
To the best of my knowledge you would have to have been running a market watcher to keep the data up to date, or import the data again. I do when I want to top up my data sets. For example I have this data set -
2018-06-14 13:43
2018-09-16 18:52
3 months, 2 days, 21 hours, 39 minutes
...so I would go to import data and change the start date to
2018-09-16 18:52 so it imports from then until the current time (quickly as it's only importing a few days data), extending my data up to the current time!
Then I could run backtests on the whole dataset, which as I just did the above now reads as follows -
2018-06-14 13:43
2018-09-22 09:34
3 months, 1 week, 1 day, 12 hours
Because when you import data using consecutive dates it binds the data set together.
Hth!