如何基于Pandas DataFrame最后5行signal列生成check校验列？

Got it, let's tackle this problem step by step. The goal is to build a check column that follows your specific rules based on the rolling 5-row window of the signal column. Here's how you can do it in pandas:

Step 1: Set up your sample DataFrame #

First, let's recreate the data you provided to test our solution:

import pandas as pd

data = {'signal': [1,1,1,1,1,1,0,0,0,0,0,0,0,1,0,1,1,1,1,1]}
df = pd.DataFrame(data, index=range(20))

Step 2: Calculate rolling window flags #

We need to identify two scenarios: when the last 5 rows (including current) are all 1s, and when they're all 0s. We'll use pandas' rolling function with min_periods=1 to handle the first few rows where the window isn't yet 5 rows long:

window_size = 5

# Flag rows where the rolling window is all 1s
df['all_1'] = df['signal'].rolling(window=window_size, min_periods=1).apply(
    lambda x: (x == 1).all(), raw=True
).astype(int)

# Flag rows where the rolling window is all 0s
df['all_0'] = df['signal'].rolling(window=window_size, min_periods=1).apply(
    lambda x: (x == 0).all(), raw=True
).astype(int)

Step 3: Build the 'check' column #

Now we'll initialize the check column, update it for our flagged scenarios, and use forward fill to carry over values for other cases:

# Initialize check with 1 (matches your sample's first rows)
df['check'] = 1

# Update check to 0 where the rolling window is all 0s
df.loc[df['all_0'] == 1, 'check'] = 0

# Forward fill to carry over the last valid check value for non-flagged rows
df['check'] = df['check'].ffill()

# Finally, update check back to 1 where the rolling window is all 1s (covers later rows like index 19)
df.loc[df['all_1'] == 1, 'check'] = 1

# Clean up helper columns
df = df.drop(['all_1', 'all_0'], axis=1)

Step 4: Verify the result #

If you print df, you'll get exactly the output you expected:

signal  check
0        1      1
1        1      1
2        1      1
3        1      1
4        1      1
5        1      1
6        0      1
7        0      1
8        0      1
9        0      1
10       0      0
11       0      0
12       0      0
13       1      0
14       0      0
15       1      0
16       1      0
17       1      0
18       1      0
19       1      1

How this works: #

• The rolling window flags catch exactly when we need to switch the check value to 1 or 0.
• Forward fill (ffill()) ensures that any row that doesn't meet the all-1 or all-0 condition keeps the last valid check value.
• We update the all-1 flag last because after switching to 0, we might later hit an all-1 window that needs to set check back to 1.

内容的提问来源于stack exchange，提问作者Marcel Mendes Reis