解决Yahoo Finance数据处理中的TypeError: string indices must be integers问题

构建机器学习模型过程中，从Yahoo Finance获取CSV格式数据后，计算日移动平均时触发TypeError错误。

错误代码 #

ma_day = [10,30,50]
for ma in ma_day:
    for df in df:
        Moving_Average = f"MA for {ma} days"
        df[Moving_Average] = df['Adj Close'].rolling(ma).mean()

报错信息 #

--------------------------------------------------------------------------- TypeError                                 Traceback (most recent call last) /var/folders/m_/7qrjlgz130q_m51m48k7h0w80000gn/T/ipykernel_1389/3705405762.py in <module>
      6     for df in df:
      7         Moving_Average = f"MA for {ma} days"
----> 8         df[Moving_Average] = df['Adj Close'].rolling(ma).mean()

TypeError: string indices must be integers

错误原因 #

内层循环for df in df会将DataFrame遍历为列名字符串，导致循环内的df变量变为字符串类型，无法使用df['Adj Close']这类字符串索引操作（字符串的索引必须是整数）。

修正后的代码 #

ma_day = [10,30,50]
for ma in ma_day:
    Moving_Average = f"MA for {ma} days"
    df[Moving_Average] = df['Adj Close'].rolling(ma).mean()

内容的提问来源于stack exchange，提问作者Prachik