Python 2.7下合并同结构DataFrame以减少None值

Since your two DataFrames share identical structure and index, the simplest and most efficient way to achieve your desired result is using pandas' combine_first() method. This method fills missing (None) values in the first DataFrame with non-missing values from the second DataFrame—exactly what you need to minimize Nones.

Step-by-Step Solution: #

• Import pandas: Make sure you have pandas installed (compatible with Python 2.7—versions like pandas 0.25.x work well).
• Define your DataFrames: Recreate the sample DataFrames as provided.
• Apply combine_first(): Use this method on df1 with df2 as the argument to get the merged result.

Code Example: #

import pandas as pd

# Create df1
df1 = pd.DataFrame(
    {'colA': ['A', None, 'B'],
     'colB': [None, None, 'C'],
     'colC': ['X', 'D', None]},
    index=[1, 2, 3]
)

# Create df2
df2 = pd.DataFrame(
    {'colA': [None, 'Y', None],
     'colB': [None, 'Z', None],
     'colC': [None, None, None]},
    index=[1, 2, 3]
)

# Merge to minimize None values
df3 = df1.combine_first(df2)

print(df3)

Output: #

colA colB colC
index               
1        A  NaN    X
2        Y    Z    D
3        B    C  NaN

(Note: In pandas, None is represented as NaN for object columns, but they behave identically in this context.)

How It Works: #

• combine_first() prioritizes values from df1 wherever they are not missing.
• For any position where df1 has a None (or NaN), it pulls the corresponding value from df2 if it's non-missing.
• Since your indexes are identical, the method aligns rows perfectly without needing additional merge keys.

This approach is concise, efficient, and leverages pandas' built-in functionality to handle exactly this use case.

内容的提问来源于stack exchange，提问作者Omar14