比特币算法无法正确遍历历史数据

Question

我正在比特币上创建一个简单的交易回溯测试程序，但我的代码中的for循环有问题。当前代码基于2个简单的移动平均数q和z（目前出于学习目的，没有真正的策略）info是一个数据帧，保存来自csv文件的比特币历史数据。好像有个盎司外的错误，我想不出来。任何帮助都将不胜感激

import pandas as pd
import numpy as np

cash = 10000

file = 'BTC-USD.csv'
data = pd.read_csv(file)

y = data['Adj Close'][1000:]
x = data['Date'][1000:]
v = data['Volume'][1000:]
h = data['High'][1000:]
l = data['Low'][1000:]

def movAvg(values,time):
    times=np.repeat(1.0,time)/time
    sma = np.convolve(values,times,'valid')
    return sma

z = movAvg(y,12)
q  = movAvg(y,9)
SP = len(x[50-1:])


def AlgoCal(account,info):
    #i = 1050
    bought = False
    test = []
    for x in info.index:
        if q[x]<z[x]:
            if bought == False:
                temp = info[x]
                account = account-info[x]
                test.append(account)
                bought = True
        elif q[x]>z[x]:
            if bought == True:
                temp = info[x]
                account = account + info[x]
                test.append(account)
                bought = False
        else:
            print("Error")
    return(test)

money = AlgoCal(cash,y)
print(money)

雅虎比特币csv中的示例数据

Date,Open,High,Low,Close,Adj Close,Volume
2014-09-17,465.864014,468.174011,452.421997,457.334015,457.334015,21056800
2014-09-18,456.859985,456.859985,413.104004,424.440002,424.440002,34483200
........
........
2020-05-21,9522.740234,9555.242188,8869.930664,9081.761719,9081.761719,39326160532
2020-05-22,9080.334961,9232.936523,9008.638672,9182.577148,9182.577148,29810773699
2020-05-23,9185.062500,9302.501953,9118.108398,9209.287109,9209.287109,27727866812
2020-05-24,9196.930664,9268.914063,9165.896484,9268.914063,9268.914063,27658280960

错误：

Traceback (most recent call last):
  File "main.py", line 47, in <module>
    money = AlgoCal(cash,y)
  File "main.py", line 31, in AlgoCal
    if q[x]<z[x]:
IndexError: index 1066 is out of bounds for axis 0 with size 1066

Answer 1

你的移动平均线有两个不同的长度。一个是12期，另一个是9期。当你试着用AlgoCal来比较它们时，你的短的一个就用完了，给你一个越界错误

如果你想用这种方式比较移动平均线，你需要在开始时添加一个最短的周期，只有在两个平均线都可用时才开始