如何在Pandas GroupBy分组结果中跨组查找重复的C列值

Hey there! Let's walk through how to spot which values in column C repeat across different groups after running that groupby sum. First, let's start with your code to get our baseline:

import pandas as pd

df = pd.DataFrame({'A': ['foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'foo'], 
                   'B': ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 
                   'C': [3,4,5,8,10,12,14,12]})
grouped = df.groupby(['A','B']).sum()

After grouping and summing, here's what our grouped DataFrame looks like:

C
A   B       
bar one    4
    three  8
    two   12
foo one   17
    three 12
    two   15

As you noticed, the value 12 pops up in two different groups. Here are a few straightforward ways to identify these duplicates:

Method 1: Find duplicate C values first, then map back to groups #

First, we'll use value_counts() to spot which C values appear more than once, then filter our grouped results to show only those cases:

# Get all C values that appear in multiple groups
duplicate_c_values = grouped['C'].value_counts()[grouped['C'].value_counts() > 1].index

# Filter the grouped DataFrame to show only these duplicates
grouped[grouped['C'].isin(duplicate_c_values)]

This will return exactly the groups sharing duplicate C values:

C
A   B       
bar two   12
foo three 12

Method 2: Add a flag column to mark duplicates directly #

If you want to keep the full grouped DataFrame but highlight which rows have duplicate C values, use transform() with duplicated(keep=False) (the keep=False ensures all duplicates get marked, not just the ones after the first occurrence):

grouped['is_duplicate'] = grouped['C'].transform(lambda x: x.duplicated(keep=False))

Now your grouped data will have a clear flag column:

C  is_duplicate
A   B                     
bar one    4         False
    three  8         False
    two   12          True
foo one   17         False
    three 12          True
    two   15         False

Method 3: Group by C values to see all related groups at once #

If you want a bird's-eye view of which groups share each duplicate C value, reverse the grouping:

# Reset index first to turn our multi-index into columns
grouped_reset = grouped.reset_index()
# Group by C and collect all (A,B) pairs for each value
c_to_groups = grouped_reset.groupby('C')[['A', 'B']].agg(list)
# Filter to only show C values with multiple groups
c_to_groups[c_to_groups['A'].str.len() > 1]

This gives you a clean summary of duplicate C values and their associated groups:

A          B
C                             
12  [bar, foo]  [two, three]

内容的提问来源于stack exchange，提问作者Prassanth