Chapter 5: Strategy Development in Practice¶

This chapter is currently maintained in Chinese first.

Chinese chapter: 第 5 章：策略开发实战 (Strategy Implementation)
Textbook home: Chinese textbook index
Practice links:
Primary example: examples/textbook/ch05_strategy.py
Extended example: examples/23_functional_callbacks_demo.py
Indicator playbook example: examples/45_talib_indicator_playbook_demo.py
Real-data mode: python examples/45_talib_indicator_playbook_demo.py --data-source akshare --symbol sh600000 --start-date 20240101 --end-date 20260301
Guide: Strategy Guide

TA-Lib Backend Usage¶

backend="auto" defaults to rust.
For baseline alignment with legacy strategies, explicitly use backend="python".
To override auto globally, set AKQUANT_TALIB_AUTO_BACKEND=python|rust.

from akquant import talib as ta

close = df["close"].to_numpy()
high = df["high"].to_numpy()
low = df["low"].to_numpy()
volume = df["volume"].to_numpy()

rsi_py = ta.RSI(close, timeperiod=14, backend="python")
rsi_rs = ta.RSI(close, timeperiod=14, backend="rust")
mfi_rs = ta.MFI(high, low, close, volume, timeperiod=14, backend="rust")
tema_rs = ta.TEMA(close, timeperiod=20, backend="rust")

Indicator Selection Template¶

Trend-following: EMA + ADX + NATR
Mean-reversion: BBANDS + RSI
Volume confirmation: OBV + MFI + ROC
Trailing risk control: SAR + ATR

import numpy as np
from akquant import talib as ta

ema_fast = ta.EMA(close, timeperiod=20, backend="rust")
ema_slow = ta.EMA(close, timeperiod=60, backend="rust")
adx = ta.ADX(high, low, close, timeperiod=14, backend="rust")

if np.isnan(ema_fast[-1]) or np.isnan(adx[-1]):
    return
if ema_fast[-1] > ema_slow[-1] and adx[-1] >= 20:
    self.buy(symbol, 100)

For complete Chinese content including warmup handling and migration tips: - 第 5 章：策略开发实战 (Strategy Implementation) - 指标组合实战手册