Tony Denion

The no data error isn't a big issue, you can load and start the orderbook in a while loop like this : ```python def runOrderBook(self): while self.is_running: self.orderbook.start() try: while...

Hello, You can run TradzQAI with all data frequency you want, i've just push 1M data for exemple. Keep me posted if you have any trouble.

Live session allow only 1M frequency, i'm working to add a support to handle any frequency. i will push it soon.

Your agent is not learning, you can do some change on the network, do some fine tuning, also change preprocessing. Also make sur that the memory size you defined in...

the dataloader is supposed to handle 3,4,6 and 6 row. [here](https://github.com/kkuette/TradzQAI/blob/master/TradzQAI/core/environnement/base/dataloader.py#L104) is the getter function, i should add a setting in config/environnement.json to handle this kind of things. Thanks for...

I haven't removed that feature, i've just commented the last 3 names [here](https://github.com/kkuette/TradzQAI/blob/master/TradzQAI/core/environnement/base/dataloader.py#L137). You can uncomment it if you want, but take care about names if you are using complex...

I'm working on live session actually, there is alot of fix to do

Hi, did you change something from the original code ?

This implementation is very slow... If you're using tensorflow i recommend you to see [tensorforce](https://github.com/reinforceio/tensorforce) for a faster training. I can also suggest you to see my project called [TradzQAI](https://github.com/kkuette/TradzQAI).

Your agent did not learn well, i think it's just overfitting.