如何用Pandas基于列条件将当前值与前值拼接？

Hey there! Since you're new to Python and Pandas, let's walk through exactly how to replicate that Excel task you described. It's a common cumulative concatenation problem with a reset condition, and Pandas has efficient ways to handle this without messy row-by-row loops.

Step 1: Set up your sample data #

First, let's create the DataFrame from your example to test our solution:

import pandas as pd

# Your sample data
data = {
    'A': [False, True, True, False],
    'B': ['bird', 'fish', 'Tiger', 'Elephant']
}
df = pd.DataFrame(data)

Step 2: Create group identifiers #

The key here is to group rows together where we need to concatenate values. Every time column A is False, we start a new group (since that's where we reset the concatenation). We can generate these groups using a cumulative sum:

# Generate group IDs: increments every time A is False
df['group'] = df['A'].eq(False).cumsum()

Step 3: Cumulative concatenation within each group #

Now we'll use groupby to process each group separately. For each group, we'll build the concatenated string incrementally using itertools.accumulate (this is more efficient than looping for larger datasets):

from itertools import accumulate

# Define how to concatenate the previous value with the current B value
def concat_with_comma(prev, current):
    return f"{prev},{current}"

# Apply cumulative concatenation to each group
df['C'] = df.groupby('group')['B'].transform(
    lambda group: list(accumulate(group, concat_with_comma))
)

Step 4: Check the result #

If you print the DataFrame now, you'll see it matches your expected output perfectly:

print(df)

Output:

A        B  group                C
0   False     bird      1              bird
1    True     fish      1        bird,fish
2    True    Tiger      1  bird,fish,Tiger
3   False  Elephant      2         Elephant

How this works #

• Group IDs: df['A'].eq(False).cumsum() creates a unique number for each "block" of rows starting with a False in column A. All subsequent True rows stay in the same group until the next False.
• Cumulative concatenation: accumulate iterates through each group's B values, taking the previous concatenated string and appending the current B value with a comma. This mimics exactly what you'd do manually in Excel when dragging a formula down.

Optional: Remove the temporary group column #

If you don't need the group column in your final output, you can drop it with:

df = df.drop('group', axis=1)

内容的提问来源于stack exchange，提问作者TVD