在寻找其他一些东西时,我在 StackOverlow家族网站之一上遇到了一个问题:Quantitative Finance aka Quant StackExchange。问题:

它被标记为Python,因此值得看看backtrader是否能胜任这项任务。

分析仪本身

这个问题似乎适合简单的分析器。虽然问题只需要高于移动平均线的那些,但我们会保留额外的信息,例如不符合标准的股票,以确保谷物实际上与谷壳分离。

class Screener_SMA(bt.Analyzer):
    params = dict(period=10)

    def start(self):
        self.smas = {data: bt.indicators.SMA(data, period=self.p.period)
                     for data in self.datas}

    def stop(self):
        self.rets['over'] = list()
        self.rets['under'] = list()

        for data, sma in self.smas.items():
            node = data._name, data.close[0], sma[0]
            if data > sma:  # if data.close[0] > sma[0]
                self.rets['over'].append(node)
            else:
                self.rets['under'].append(node)

Tips:当然也需要import backtrader as bt

这几乎解决了问题。分析仪分析

  • 有一个period作为参数有一个灵活的分析器
  • start方法

    为系统中的每个数据制作一个简单移动平均线SMA)。

  • stop方法

    查看哪些数据close如果未指定其他内容)在其 sma之上,并将其存储在返回()键下的列表中overrets

    该成员rets分析器中的标准成员,并且恰好是 collections.OrderedDict. 由基类创建。

    将不符合条件的保留在密钥下under

现在的问题是:让分析仪启动并运行。

方法一

backtrader几乎从一开始就包含一个名为 的自动脚本btrun,它可以从 python 模块加载策略、指标、分析器、解析参数,当然还有绘图。

让我们运行一下:

$ btrun --format yahoo --data YHOO --data IBM --data NVDA --data TSLA --data ORCL --data AAPL --fromdate 2016-07-15 --todate 2016-08-13 --analyzer st-screener:Screener_SMA --cerebro runonce=0 --writer --nostdstats
===============================================================================
Cerebro:
  -----------------------------------------------------------------------------
  - Datas:
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data0:
      - Name: YHOO
      - Timeframe: Days
      - Compression: 1
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data1:
      - Name: IBM
      - Timeframe: Days
      - Compression: 1
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data2:
      - Name: NVDA
      - Timeframe: Days
      - Compression: 1
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data3:
      - Name: TSLA
      - Timeframe: Days
      - Compression: 1
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data4:
      - Name: ORCL
      - Timeframe: Days
      - Compression: 1
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data5:
      - Name: AAPL
      - Timeframe: Days
      - Compression: 1
  -----------------------------------------------------------------------------
  - Strategies:
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Strategy:
      *************************************************************************
      - Params:
      *************************************************************************
      - Indicators:
        .......................................................................
        - SMA:
          - Lines: sma
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Params:
            - period: 10
      *************************************************************************
      - Observers:
      *************************************************************************
      - Analyzers:
        .......................................................................
        - Value:
          - Begin: 10000.0
          - End: 10000.0
        .......................................................................
        - Screener_SMA:
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Params:
            - period: 10
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Analysis:
            - over: ('ORCL', 41.09, 41.032), ('IBM', 161.95, 161.221), ('YHOO', 42.94, 39.629000000000005), ('AAPL', 108.18, 106.926), ('NVDA', 63.04, 58.327)
            - under: ('TSLA', 224.91, 228.423)

我们使用了一组众所周知的代码:

  • AAPLIBMNVDAORCLTSLA,YHOO

唯一恰好在10简单移动平均线下的一个是TSLA.

让我们尝试50几天的时间。是的,这也可以用 来控制 btrun。运行(输出缩短):

$ btrun --format yahoo --data YHOO --data IBM --data NVDA --data TSLA --data ORCL --data AAPL --fromdate 2016-05-15 --todate 2016-08-13 --analyzer st-screener:Screener_SMA:period=50 --cerebro runonce=0 --writer --nostdstats
===============================================================================
Cerebro:
  -----------------------------------------------------------------------------
  - Datas:
    +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    - Data0:
...
...
...
        - Screener_SMA:
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Params:
            - period: 50
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Analysis:
            - over: ('ORCL', 41.09, 40.339), ('IBM', 161.95, 155.0356), ('YHOO', 42.94, 37.9648), ('TSLA', 224.91, 220.4784), ('AAPL', 108.18, 98.9782), ('NVDA', 63.04, 51.4746)
            - under:

请注意如何50在命令行中指定天数:

  • st-screener:Screener_SMA:period=50

    在之前的运行中,这是st-screener:Screener_SMA并且使用了代码中的默认值 10

我们还需要调整fromdate以确保有足够的柱线来考虑计算简单移动平均线

在这种情况下,所有代码都高于50均线。

方法二

制作一个小脚本(完整代码见下文)以更好地控制我们的工作。但结果是一样的。

核心相当小:

cerebro = bt.Cerebro()
for ticker in args.tickers.split(','):
    data = bt.feeds.YahooFinanceData(dataname=ticker,
                                     fromdate=fromdate, todate=todate)
    cerebro.adddata(data)

cerebro.addanalyzer(Screener_SMA, period=args.period)
cerebro.run(runonce=False, stdstats=False, writer=True)

主要是关于参数解析的其余部分。

10几天(再次缩短输出):

$ ./st-screener.py
===============================================================================
Cerebro:
...
...
...
        - Screener_SMA:
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Params:
            - period: 10
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          - Analysis:
            - over: (u'NVDA', 63.04, 58.327), (u'AAPL', 108.18, 106.926), (u'YHOO', 42.94, 39.629000000000005), (u'IBM', 161.95, 161.221), (u'ORCL', 41.09, 41.032)
            - under: (u'TSLA', 224.91, 228.423)

结果相同。因此,让我们避免50几天重复。

结束语

方法 1方法 2中的小脚本都btrun使用完全相同的分析器,因此提供相同的结果。

反向交易者已经能够承受另一个小挑战

最后的两个注意事项:

  • 这两种方法都使用内置的编写器功能来提供输出。
    • btrun作为参数--writer
    • cerebro.run作为参数writer=True
  • 在这两种情况下runonce都已停用。这是为了确保在线数据保持同步,因为结果可能有不同的长度(其中一只股票的交易量可能较少)

脚本使用

$ ./st-screener.py --help
usage: st-screener.py [-h] [--tickers TICKERS] [--period PERIOD]

SMA Stock Screener

optional arguments:
  -h, --help         show this help message and exit
  --tickers TICKERS  Yahoo Tickers to consider, COMMA separated (default:
                     YHOO,IBM,AAPL,TSLA,ORCL,NVDA)
  --period PERIOD    SMA period (default: 10)

完整的脚本

#!/usr/bin/env python
# -*- coding: utf-8; py-indent-offset:4 -*-
###############################################################################
#
# Copyright (C) 2015, 2016 Daniel Rodriguez
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
###############################################################################
from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import argparse
import datetime

import backtrader as bt


class Screener_SMA(bt.Analyzer):
    params = dict(period=10)

    def start(self):
        self.smas = {data: bt.indicators.SMA(data, period=self.p.period)
                     for data in self.datas}

    def stop(self):
        self.rets['over'] = list()
        self.rets['under'] = list()

        for data, sma in self.smas.items():
            node = data._name, data.close[0], sma[0]
            if data > sma:  # if data.close[0] > sma[0]
                self.rets['over'].append(node)
            else:
                self.rets['under'].append(node)


DEFAULTTICKERS = ['YHOO', 'IBM', 'AAPL', 'TSLA', 'ORCL', 'NVDA']


def run(args=None):
    args = parse_args(args)
    todate = datetime.date.today()
    # Get from date from period +X% for weekeends/bank/holidays: let's double
    fromdate = todate - datetime.timedelta(days=args.period * 2)

    cerebro = bt.Cerebro()
    for ticker in args.tickers.split(','):
        data = bt.feeds.YahooFinanceData(dataname=ticker,
                                         fromdate=fromdate, todate=todate)
        cerebro.adddata(data)

    cerebro.addanalyzer(Screener_SMA, period=args.period)
    cerebro.run(runonce=False, stdstats=False, writer=True)


def parse_args(pargs=None):

    parser = argparse.ArgumentParser(
        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
        description='SMA Stock Screener')

    parser.add_argument('--tickers', required=False, action='store',
                        default=','.join(DEFAULTTICKERS),
                        help='Yahoo Tickers to consider, COMMA separated')

    parser.add_argument('--period', required=False, action='store',
                        type=int, default=10,
                        help=('SMA period'))

    if pargs is not None:
        return parser.parse_args(pargs)

    return parser.parse_args()


if __name__ == '__main__':
    run()

 

评论被关闭。