Pandas有一个函数drop_na，您可以给予它的一个子集（在本例中为列C和D）https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.dropna.html
我会这样做：

df = pd.DataFrame({"A":["a","aaa","aa","a",None],"B":["b","bbb","b","b", None],"C":[None, "cc", None, None, "c"], "D": [None, None, None, "dd", None]}) # create the test dataframe

df = df.dropna(subset=["C","D"], how="all") # drop rows from the C and D subset if all are null

正如注解所建议的那样，如果您想将其作为一个函数

def drop_unwanted(exclude_columns,df): 
    df = df.dropna(subset=[x for x in df.columns if x not in exclude_columns], how="all") 
    return df

并将您要排除的列指定给函数！

IIUC，这里有一个带有 boolean indexing 的选项：

out = df[~df.loc[:, "C":].isna().all(axis=1)]

输出：

print(out)

     A    B    C    D
1  aaa  bbb   cc  NaN
3    a    b  NaN   dd
4  NaN  NaN    c  NaN

你可以使用loc对 Dataframe 进行切片，并使用布尔掩码来选择正确的行：

filtered_df = df[~df.loc[:, 'C':].isna().all(axis=1)]
print(filtered_df)

# Output
      A     B     C     D
1   aaa   bbb    cc  None
3     a     b  None    dd
4  None  None     c  None

也可以使用drop：

filtered_df = df.drop(df.loc[:, 'C':].isna().all(axis=1).loc[lambda x: x].index)
print(filtered_df)

# Output
      A     B     C     D
1   aaa   bbb    cc  None
3     a     b  None    dd
4  None  None     c  None