Machine learning in trading: theory, models, practice and algo-trading - page 972
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
You'll have a hard time docking this JVM with MT. And first on Julia, and then rewrite it on MT - not a royal business.
You don't need to rewrite anything, you just need to find a way to interact
for example, the same xgboost is on juliaYou don't need to rewrite anything, you just need to find a way to interact
For example, the same xgboost is available on juliaA java machine with native code is very difficult.
Is this xgboost surely available for Python as well?
A java machine with native code is very difficult.
Surely this xgboost is also available for Python?
Sure, I use python for it.
If interaction with python is based on a script call, then the script call in julia can be done in the same wayOf course, I'm spinning it on python.
So spit on this Julia.) What's the rush? Ultimately, you need an application problem, not a model.
If you interact with python through a script call, then the script call in julia can be done in exactly the same way
So spit on that Julia.) What's the rush.
Watch the speed. It's almost the most important thing when working with data.
So spit on that Julia.) What's the rush? What you need is an application, not a model.
Python has a clear C API, while Julia's is not. I have no idea what to do with Java.https://docs.julialang.org/en/release-0.4/manual/embedding/
watch the speed. It's almost the most important thing when dealing with data
You have to look at the interface, and the speed itself is nothing to worry about.
In Python, Python itself is nothing (won't affect anything) andxgboost itselfis in C++, native code.
The interface should be looked at, and the speed itself is nothing to worry about.
In Python, there is Python itself - ugh (won't affect anything), andxgboost itselfis in C++, native code.
there's preprocessing\working with arrays/matrices is expensive in python (although convenient), not to mention R
https://docs.julialang.org/en/release-0.4/manual/embedding/
Ooh, really simple, and I was afraid)). Actually, I've seen her, Julia, before, but I saw Jawa and decided that fuck it. I must have a closer look.
Althoughxgboost itself,etc. will work faster in Python than in Julia in Java. Ifxgboostitself ismade in Java,of course.Ooh..., really simple, and I was afraid)). Actually, I've seen her, Julia, before, but I saw Jawa and decided that nah. I'll have to take a closer look.
I should make it not julia.h but julia.mqh :)