根据 2 个不同列的最后一个值计算列的 pct_change()

Question

我正在尝试计算数据框中股票价格的价格变化，该数据框中包含单独列中的进入和退出信号，并且只想获取这两者之间范围的盈亏。

Date       Price    Entry Signal    Exit Signal   Price Change
2024-01-18  3595.0  NaN            NaN
2024-01-19  3280.0  Entry          NaN
2024-01-22  3505.0  NaN            NaN
2024-01-23  3945.0  NaN            Exit       (3945 / 3280- 1)
2024-01-24  4075.0  NaN            NaN
2024-03-07  10270.0 Entry          NaN
2024-03-08  8120.0  NaN            nan
2024-03-11  6620.0  NaN            Exit       (6620 / 10270 - 1)

到目前为止，我依赖于 for 循环遍历每一行，并有一个名为position的变量，其中我递增1（一旦触发Entry）并重置为0（一旦触发Exit），但必须有一种更智能/更快的方法鉴于这是一个数据框。

Answer 1

代码

cond = df[['Entry Signal', 'Exit Signal']].notna().any(axis=1)
df['Price Change'] = df.loc[cond, 'Price'].pct_change().where(df['Exit Signal'].notna())

df：

         Date    Price Entry Signal Exit Signal  Price Change
0  2024-01-18   3595.0         None        None           NaN
1  2024-01-19   3280.0        Entry        None           NaN
2  2024-01-22   3505.0         None        None           NaN
3  2024-01-23   3945.0         None        Exit      0.202744
4  2024-01-24   4075.0         None        None           NaN
5  2024-03-07  10270.0        Entry        None           NaN
6  2024-03-08   8120.0         None        None           NaN
7  2024-03-11   6620.0         None        Exit     -0.355404

示例代码

import pandas as pd
data = {"Date": ["2024-01-18", "2024-01-19", "2024-01-22", "2024-01-23", "2024-01-24", "2024-03-07", "2024-03-08", "2024-03-11"], "Price": [3595.0, 3280.0, 3505.0, 3945.0, 4075.0, 10270.0, 8120.0, 6620.0], "Entry Signal": [None, "Entry", None, None, None, "Entry", None, None], "Exit Signal": [None, None, None, "Exit", None, None, None, "Exit"]}
df = pd.DataFrame(data)