Machine learning in trading: theory, models, practice and algo-trading - page 2580

[mytarmailS]

elibrarius #:

SanSanych's hint was DukasKopi

Limit there, up to 3 years of history, bullshit.

No one has any historical data ?????

[Maxim Dmitrievsky]

mytarmailS #:

Limit there, up to 3 years of history, bullshit.

No one has any historical data ?????

You make such scary things to do on Saturday... download from MT5 :)

[mytarmailS]

I built the last version of the TS with the filter to reduce the risks, you can without the filter then the profit is twice as much but the risks are higher

[Maxim Dmitrievsky]

mytarmailS #:

I built the last version of TS with the filter to reduce the risks, you can without the filter then the profit is twice as much but the risks are higher

Ok, let's trade

Is this a Wellslab with you? How do you transfer models there?

[Roman]

Maxim Dmitrievsky #:
Norm, start to trade

Is that your wellslab? How do you transfer models there?

No, it's a tslab.
Also wondered how?
I wonder how he builds models on dice...
There's also a C# api, maybe through it.

[mytarmailS]

Maxim Dmitrievsky #:
Norm, run in trade

Is that your wellslab? How do you transfer models there?

Tslab, it's just crutches for convenient visualization... So you don't have to write your tester in R

[mytarmailS]

Roman #:

Nah, it's a tslab.
Also wondered how?
Does it build on dice...
There's also a C# api, maybe through it.

Yes, all calculations are done in R, I send to tslab only prices and vector with buy signal, I put some dices for TS in tslab and that's it...

It's dumb, but it's fast and I know for sure it'll work without errors and can be tweaked a bit.

[Roman]

mytarmailS #:
Yes, I do all my calculations in R , I send only prices and vector with buy signal to tslab, I put some cubes for cp and that's it...

Dumb, but fast and I know for sure that the profit will count without errors + can be tweaked on small things

I see. You're a sly one, aren't you?
I thought I built this model with dice.)

[Maxim Dmitrievsky]

Python is convenient now. I wrote my own tester, but it is possible to transfer models or via api to trade. If ONNX is added, it will be a real gun.

I got in touch with the Mac M1, now I'm waiting half a year for the Catbusta for it, they promise a release in 2 weeks. So far I have been using a virtual machine on vin.

[mytarmailS]

Roman #:

I see. You're a sly one, aren't you?
I thought for a moment that I had built a model with dice.)

I started with cubes, it was my first programming, so I still have tslab...

I don't believe it's possible to build something working with cubes.