use std::collections::BTreeMap;
#[allow(unused_imports)] use nautilus_core::UnixNanos;
use nautilus_model::enums::OrderSide;
use pyo3::prelude::*;
use crate::{statistic::PortfolioStatistic, statistics::long_ratio::LongRatio};
#[pymethods]
#[pyo3_stub_gen::derive::gen_stub_pymethods]
impl LongRatio {
#[new]
#[pyo3(signature = (precision=None))]
fn py_new(precision: Option<usize>) -> Self {
Self::new(precision)
}
fn __repr__(&self) -> String {
self.to_string()
}
#[getter]
#[pyo3(name = "name")]
fn py_name(&self) -> String {
self.name()
}
#[pyo3(name = "calculate_from_positions")]
#[allow(clippy::needless_pass_by_value)]
fn py_calculate_from_positions(
&mut self,
py: Python,
positions: Vec<Py<PyAny>>,
) -> PyResult<Option<f64>> {
if positions.is_empty() {
return Ok(None);
}
let mut longs = 0;
for position in &positions {
let entry = position.getattr(py, "entry")?;
let entry_value: u8 = entry.extract(py)?;
if entry_value == OrderSide::Buy as u8 {
longs += 1;
}
}
let value = f64::from(longs) / positions.len() as f64;
let scale = 10f64.powi(self.precision as i32);
Ok(Some((value * scale).round() / scale))
}
#[pyo3(name = "calculate_from_realized_pnls")]
fn py_calculate_from_realized_pnls(&mut self, _realized_pnls: Vec<f64>) -> Option<f64> {
None
}
#[pyo3(name = "calculate_from_returns")]
#[allow(unused_variables)] fn py_calculate_from_returns(&mut self, _returns: BTreeMap<u64, f64>) -> Option<f64> {
None
}
}