无法从technical.indicators导入SSLchannels的Freqtrade策略加载问题

问题重现 #

执行回测命令 freqtrade backtesting --strategy canal 时出现导入错误，无法从 technical.indicators 导入 SSLchannels，导致策略加载失败。注释掉该导入语句后回测可正常运行，切换Python 3.8/3.10版本问题依旧。

错误日志关键信息：

2022-08-26 03:51:39,097 - freqtrade.resolvers.iresolver - WARNING - Could not import /freqtrade/user_data/strategies/canal.py due to 'cannot import name 'SSLchannels' from 'technical.indicators' (/home/ftuser/.local/lib/python3.10/site-packages/technical/indicators/init.py)'
2022-08-26 03:51:39,182 - freqtrade - ERROR - Impossible to load Strategy 'canal'. This class does not exist or contains Python code errors.

策略核心代码片段：

from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter,
                                IStrategy, IntParameter)
import talib.abstract as ta
import freqtrade.vendor.qtpylib.indicators as qtpylib
from technical.indicators import SSLchannels

class canal(IStrategy):
    INTERFACE_VERSION = 3
    # 其余策略配置及方法省略

解决方案 #

• 检查并升级technical包版本
  SSLchannels是technical包中的指标，可能你的包版本过旧未包含该指标。执行以下命令查看当前版本：

  pip show technical
  

  如果版本低于v1.1.0（SSLchannels大致在该版本加入），执行升级：

  pip install --upgrade technical
  

  若新版本移除了该指标，可指定旧版本安装：

  pip install technical==1.1.0
  

• 确认正确的导入路径
  部分版本的technical包调整了模块结构，SSLchannels可能被移到子模块中。尝试修改导入语句为：

  from technical.indicators.ssl import SSLchannels
  

  也可以直接查看包内的导出配置，确认路径：

  cat /home/ftuser/.local/lib/python3.10/site-packages/technical/indicators/__init__.py
  

  如果文件中没有from .ssl import SSLchannels这类导出语句，说明需要调整导入路径或手动添加导出。

• 手动整合SSLchannels实现到策略中
  若不想依赖外部包，可直接把SSLchannels的代码嵌入策略。以下是标准实现，添加到策略文件即可：

  import numpy as np
  def SSLchannels(dataframe, length=10, mode='sma'):
      df = dataframe.copy()
      if mode == 'sma':
          df['smaHigh'] = ta.SMA(df, timeperiod=length, price='high')
          df['smaLow'] = ta.SMA(df, timeperiod=length, price='low')
      else:
          df['smaHigh'] = ta.EMA(df, timeperiod=length, price='high')
          df['smaLow'] = ta.EMA(df, timeperiod=length, price='low')
      df['hlv'] = np.where(df['close'] > df['smaHigh'], 1, np.where(df['close'] < df['smaLow'], -1, np.NAN))
      df['hlv'] = df['hlv'].ffill()
      df['sslDown'] = np.where(df['hlv'] < 0, df['smaHigh'], df['smaLow'])
      df['sslUp'] = np.where(df['hlv'] < 0, df['smaLow'], df['smaHigh'])
      return df['sslUp'], df['sslDown']
  

  之后在策略的populate_indicators方法中调用该函数即可，无需再导入外部的SSLchannels。

内容的提问来源于stack exchange，提问作者Jose Martinez Sobrado