questions about queue.rs and implementation of new queue model like others

[MatSSrT]

Hello, I have studied the queue probability models in queue.rs within the framework, and I noticed that adding new queue models requires implementation in Rust. I'd like to ask, if I want to implement a Hawkes process as a queue probability model, do I need to write it within queue.rs? The several queue probability models inside seem to have standard function templates? Could you suggest some general steps that need to be written or modified? I've also searched and found that those implemented queue models' name appears in other files.

[nkaz001]

Not necessarily, but you need to write your own queue model by implementing QueueModel trait. Additionally, exposing this custom queue model to Python is more challenging. I recommend using Rust only, or please see py-hftbactest and hftbacktest-derive.

[MatSSrT]

Not necessarily, but you need to write your own queue model by implementing QueueModel trait. Additionally, exposing this custom queue model to Python is more challenging. I recommend using Rust only, or please see py-hftbactest and hftbacktest-derive.

I noticed you mentioned using Rust for implementation. Does this mean that both the grid trading strategy implementation and the backtesting with a custom queue probability model are done in Rust

[nkaz001]

Yes, you need to write both of them in Rust, otherwise you need to modify py-hftbacktest to expose your custom model to Python.