合并列并删除NaNs Pandas [英] Combining columns and removing NaNs Pandas

问题描述

如果我有一个像这样的熊猫数据框:

If I have a pandas dataframe like so:

a1      0.116667          NaN           NaN
a2           NaN     0.516667           NaN
a3           NaN     0.006667           NaN
a4           NaN          NaN      0.426667
a5           NaN     0.506667           NaN
a6      0.583333          NaN           NaN
a7      0.550000          NaN           NaN

我想合并这些列，以便如果任何列中有一个数字，而其他两列中有NaN，则结果是一个列，其预期输出为:

and I want to combine the columns so that if there is a number in any of the columns and NaN in the other two the result is one column, with the expected output:

a1   0.116667
a2   0.516667
a3   0.006667
a4   0.426667
a5   0.506667
a6   0.583333
a7   0.550000

推荐答案

对于某些行，如果有两个或多个不是NaN的值，则您没有指定应该怎么办.

You didn't specify what should happen if, for some row, there are two or more values that are not NaNs.

在这种情况下，根据您的需要，一个简单的按行最大值可能会解决您的问题:

Subject to what you want in this case, a simple row-wise maximum might solve your problem:

df = pd.DataFrame({
    'a': [1, None, None], 
    'b': [None, 3, None],
    'c': [None, None, 4]})
>>> df
    a   b   c
0   1   NaN     NaN
1   NaN     3   NaN
2   NaN     NaN     4

现在，取逐行最大值:

>>> df.max(axis=1)
0    1
1    3
2    4
dtype: float64

这篇关于合并列并删除NaNs Pandas的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！