您不能使用变异的DF
row此处添加新列，你要么是指原来的DF或使用
.loc，
.iloc或
.ix，例如：
In [29]:df = pd.Dataframe(columns=list('abc'), data = np.random.randn(5,3))dfOut[29]: a b c0 -1.525011 0.778190 -1.0103911 0.619824 0.790439 -0.6925682 1.272323 1.620728 0.1921693 0.193523 0.070921 1.0675444 0.057110 -1.007442 1.706704In [30]:for index,row in df.iterrows(): df.loc[index,'d'] = np.random.randint(0, 10)dfOut[30]: a b c d0 -1.525011 0.778190 -1.010391 91 0.619824 0.790439 -0.692568 92 1.272323 1.620728 0.192169 13 0.193523 0.070921 1.067544 04 0.057110 -1.007442 1.706704 9您可以修改现有行：
In [31]:# reset the df by slicingdf = df[list('abc')]for index,row in df.iterrows(): row['b'] = np.random.randint(0, 10)dfOut[31]: a b c0 -1.525011 8 -1.0103911 0.619824 2 -0.6925682 1.272323 8 0.1921693 0.193523 2 1.0675444 0.057110 3 1.706704但是，使用row添加新列将行不通：
In [35]:df = df[list('abc')]for index,row in df.iterrows(): row['d'] = np.random.randint(0,10)dfOut[35]: a b c0 -1.525011 8 -1.0103911 0.619824 2 -0.6925682 1.272323 8 0.1921693 0.193523 2 1.0675444 0.057110 3 1.706704
