Machine learning in trading: theory, models, practice and algo-trading - page 1988

New comment
 
Maxim Dmitrievsky:

What did you do to the signal to make it trade so well?

 
mytarmailS:

What did you do to the signal to make it trade so well?

raised the lot

 
Maxim Dmitrievsky:

I raised the lot.

And right away into such a bummer))

 
Maxim Dmitrievsky:

try just running tester.py from the folder, without vscode

if the same error, then your python doesn't see the folder

Reinstalled the python, ticked off path....

checked the path itself entered


from a folder ( without vscode ) code runs, everything works, but not in the VSC

 
mytarmailS:

And right away into such a bummer ))

well, bad lac.

 
mytarmailS:

reinstalled python, checked path....

Checked the path


the code runs from a folder ( without vscode ), everything works, but not in the VSC

You can google VScode for python. I do not remember, there are a lot of

You need to put something in the config.

but it's a very nice and fast editor

 
Maxim Dmitrievsky:

I raised the lot.

You shouldn't have, two weeks of stats went to hell.


So you know how to make libraries?

 
mytarmailS:

You shouldn't have, two weeks of statistics went to hell.


So you know how to make libraries?

yes

 
Maxim Dmitrievsky:

well, bad varnish

Does it make sense to raise the lot, when you just need to understand if the model is working or not. From what I see, so far no logic.
 
Farkhat Guzairov:
Does it make sense to raise the lot, when you just need to understand if the model works or not. From what I can see, there is no logic.

This is because we should connect the tester with the trader and use a pre-trained model at least for a couple of weeks. On cold runs it does not know anything at first

I wanted to make a save, but it is possible to pre-train.

What does the lot have to do with it. I'm testing the errors

1...198119821983198419851986198719881989199019911992199319941995...3656
New comment