Question

In [139]: pandas.__version__
Out[139]:  0.7.3

I have two aligned series in a DataFrame which have several unmatched "NaN". I would like to print the intersection between them removing all "NaN s", but without loose alignment. That is, I want to remove the rows from both series whem I find a "NaN" in one of them. It sounds simple, but I not doing any operation between the series to dropna s afterwards and cannot dropna s from the series separately. I couldn t figure out the right df function to do this - several are not documented. Just an example, I want to take this:

10         NaN     -1.200
11         NaN     -1.324
12    0.000585        NaN
13    0.000573     -1.453
14         NaN     -2.006

打印此 :

13    0.000573     -1.453

Answer 1

正如我在上文的评论中写到的,最好的答案是:

newdf = df1.dropna()[[ S ,  JEXP ]]

that dropan s from a slice of the original df, keeping just the series of interest. Karmel has suggested:

newdf = df1.dropna(subset=[ S ,  JEXP ])

并基于子集列表运行和投影, 但是,保存所有其他序列 - 它重复你的数据集。

友情链接