按列值汇总Pandas DataFrame：二进制指标行分组统计需求

To solve this problem—grouping identical binary rows and adding a count column—we can leverage Pandas' built-in value_counts() method, which efficiently counts occurrences of unique rows. Here's a step-by-step breakdown:

Step 1: Understand the Goal #

We need to take a DataFrame with binary columns, identify all unique row combinations, count how many times each unique row appears, and create a new DataFrame that pairs these unique rows with their respective counts.

Step 2: Use value_counts() with sort=False #

The value_counts() method directly counts unique row combinations out of the box. Setting sort=False ensures we preserve the order of the first occurrence of each unique row (matching the order in your example output).

Step 3: Convert to a DataFrame and Rename the Count Column #

After generating the counts, we use reset_index() to convert the resulting Series back into a structured DataFrame, then rename the count column for clarity.

Full Code Example #

import pandas as pd

# Your input DataFrame
df = pd.DataFrame([
    [0,1,1,0], [0,1,1,0], [0,0,0,1], 
    [0,0,0,1], [1,1,1,0], [1,1,1,1], [1,1,1,0]
])

# Generate the result DataFrame
res = df.value_counts(sort=False).reset_index(name='count')

# Optional: If you prefer the count column to be unnamed (Pandas recommends named columns for clarity)
# res.rename(columns={'count': ''}, inplace=True)

print(res)

Output #

This produces exactly the result you specified:

0  1  2  3  count
0  0  1  1  0      2
1  0  0  0  1      2
2  1  1  1  0      2
3  1  1  1  1      1

Alternative Approach Using groupby #

If you prefer using groupby instead, here's an equivalent solution (note: groupby sorts groups by default, so to match the example's order you'd need extra steps to track first occurrence):

res = df.groupby(df.columns.tolist()).size().reset_index(name='count')

The value_counts() method is more concise and efficient for this use case, especially when preserving the order of first occurrence is important.

内容的提问来源于stack exchange，提问作者Research100