跳过 Pandas 数据帧循环中的行

Question

我正在处理以下问题，但似乎在网上找不到任何解决方案。

我有一个针对数据框的for循环。此循环应执行以下操作：如果“reversal”列的内容== 1，则将“action”列填充为1，跳过125行，在下一个“action”列中填充第126行为-1，并继续从下一行重复循环。如果“reversal”列！= 1，则继续进行循环而不填充“action”。

问题1是，当我编写“index = index + 126”时，由于某种原因，Python无法理解它需要跳过126行。

问题2是，当我添加条件以避免使“action”列比“reversal”列更长时，该条件不起作用（请参见代码注释）。

#creating the on/off signal column
df_zinc['action'] = 0

#creating the loop
for index,row in df_zinc.iterrows():
    if row.reversal == 1:
        df_zinc.loc[index,'action'] = 1
        if index<len(df_zinc.index)-126:             #the purpose of this condition is to not have the action column longer than the reversal column. Thuogh, it appears not to be working
            df_zinc.loc[index+126, 'action'] = -1
        index= index + 127

- gianluca

Answer 1

如果可以使用索引，请不要使用itterrows()。

请尝试以下方法：

#creating the on/off signal column
# df_zinc['action'] = 0
#
count = 0
# #creating the loop
for index in df_zinc.index:
    if index < count:
        continue
    if df_zinc.at[index , 'reversal'] == 1:
        df_zinc.at[index , 'action'] = 1
        if index < len(df_zinc)-126:             #the purpose of this condition is to not have the action column longer than the reversal column. Thuogh, it appears not to be working
            df_zinc.at[index+126, 'action'] = -1
        count = index + 127

- Nihal

Answer 2

import numpy as np
import pandas as pd
reversal=np.eye(126,dtype=int)
reversal=reversal.reshape(-1)
data=pd.DataFrame({"reverse":reversal})
data['action']=0
for index in range(len(data)):
    if data.loc[index,"reverse"] == 1:
        data.loc[index,'action'] = 1
        if index<len(data.index)-126:            
            data.loc[index+126, 'action'] = -1
        index= index + 127

你可以试试这个。

- Justice_Lords