Given a dataframe (e.g.)

df = pd.DataFrame({"A": [9, 6, 9, 4, 5], "B": [8, 2, 9, 7, 3], "C": [0, 9, 3, 15, 5],})

I’d like to return a dataframe with columns A, B, C and a single row, where the row value is the smallest absolute delta between the cells in A, B, C (e.g.)

df_result = pd.DataFrame(
    {
        "A": [0], # 9-9
        "B": [1], # 2-3, 9-8, 8-7
        "C": [2], # 5-3
    }
)

I’d like to do this generally. That is take in a dataframe then find the delta between all the cells except itself, and find the minimum value. Then that over each column. Then return a dataframe with the same columns and one row which contains the absolute value of the mininmum delta.

>Solution :

You can sort_values then compute the diff and finally get the min value:

>>> df.agg(lambda x: x.sort_values().diff().min())
A    0.0
B    1.0
C    2.0
dtype: float64

Example for A:

>>> df['A'].sort_values()
3    4
4    5
1    6
0    9
2    9
Name: A, dtype: int64

>>> df['A'].sort_values().diff()
3    NaN
4    1.0
1    1.0
0    3.0
2    0.0
Name: A, dtype: float64

>>> df['A'].sort_values().diff().min()
0.0