实盘与回测复盘校对

下载glogg： 高效的日志管理工具

下载链接： http://glogg.bonnefon.org/download.html

回测中打印关键信息，检查策略回测是否按照对的逻辑执行

print & self.writeLog('%s'%())

1. 数据数组（将回测数据画出来检查是否有缺失）
2. 策略信号（打印指标数值）
3. 止损止盈（出场执行类型）
4. 加仓减仓（持仓字典检查是否有多空同时持有与异常下单）
5. 交割单与成交价（检查是否有按照市价成交）

实盘中打印信号与仓位信息，检查实盘是否与回测的下单一致

需要打印关键词方便在日志中进行检索与管理

1. 数据数组的长度
2. 指标数值的计算
3. 信号下单时间点与价位
4. 止损止盈的滑点大小
5. 加仓减仓的执行逻辑

实盘信号检测

将实盘日志字符串转为数据并做成表格核对

with open("vt_20181023.log","r",encoding="utf-8") as f:
    content = f.readlines()
data = []
for logRow in content:
    if logRow.find("CTA_STRATEGY    Mas_Strategy:eos_quarter:OKEX_WUon_1min_bar")>=0:
        result = logRow.split("CTA_STRATEGY Mas_Strategy:eos_quarter:OKEX_WUon_1min_bar")[-1]
        data.append(result)

import pandas as pd
from datetime import datetime

checkDict = {'datetime':[], 'cross':[], 'trend':[], 'wave':[]}
#checkDict = {'datetime':[],'close':[],'cross':[],'trend':[],'wave':[],'n':[]}
for row in data:
    dateTime = row[row.find('time')+5:row.find('date')+26]
    rigthTime = datetime.strptime(dateTime, "%Y-%m-%d %H:%M:%S")
    checkDict['datetime'].append(rigthTime)
    #checkBtDict['close'].append(float(row[row.find('close')+6:].split(',')[0]))
    checkDict['cross'].append((row[row.find('cross')+6:].split(',')[0]))
    checkDict['trend'].append((row[row.find('trend')+6:].split(',')[0]))
    checkDict['wave'].append((row[row.find('wave')+5:].split('\n')[0]))
    #checkBtDict['n'].append(float(row[row.find("n")+1:].split(':')[-1]))

realCheck = pd.DataFrame(checkDict)

realCheckDf = realCheck.set_index('datetime')

realCheckDf

回测信号检查

将回测日志关键的信号输出与实盘合成表格进行检查

checkBtDict = {'datetime':[],'close':[],'cross':[],'trend':[],'wave':[],'n':[]}

import pandas as pd
btLog = pd.read_csv('BacktestLog.csv')
btData = []
for i, value in btLog.iterrows():
    if value.values[0].find("Mas_Strategy:eos_quarter:OKEXon_1min_bar")>=0:
        result = value.values[0].split("Mas_Strategy:eos_quarter:OKEXon_1min_bar")[-1]
        btData.append(result)

from datetime import datetime
for row in btData:
    dateTime = row[row.find('time')+5:25]
    rigthTime = datetime.strptime(dateTime, "%Y-%m-%d %H:%M:%S")
    checkBtDict['datetime'].append(rigthTime)
    checkBtDict['close'].append(float(row[row.find('close')+6:].split(',')[0]))
    checkBtDict['cross'].append(float(row[row.find('cross')+6:].split(',')[0]))
    checkBtDict['trend'].append(float(row[row.find('trend')+6:].split(',')[0]))
    checkBtDict['wave'].append(float(row[row.find("wave")+5:].split(',')[0]))
    checkBtDict['n'].append(float(row[row.find("n")+1:].split(':')[-1]))

checkBtData = pd.DataFrame(checkBtDict)

checkBtDd = checkBtData.set_index('datetime')

FinalDf = pd.concat([checkBtDd, realCheckDf], axis=1).dropna()

FinalDf

FinalDf.to_excel('CheckSignal.xlsx')