在pandas数据帧中用NAN替换空格 [英] Replaces spaces with NAN in pandas dataframe

问题描述

我有一个包含三列的pandas数据框：

I have a pandas dataframe with three columns:

Name        Name2           DateTime
                            2016-06-10 05:22
                            2016-06-10 05:23
                            2016-06-10 14:25
Guest       Guest1          2016-06-10 15:32

我必须用NAN替换空格。因此，行1,2,3和4的AccountName和AccountName2应为NAN。我尝试了以下声明：

I have to replace empty spaces with NAN. So AccountName and AccountName2 of rows 1,2,3 and 4 should be NAN. I tried the below statement:

 df3['Name'] = df3['Name'].replace(r'[^\s+]',np.nan, regex=True)

但是因为我之后有空格名称中的Guest，所有5行都替换为NAN。

But since I have white spaces after "Guest " in Name, all 5 rows get replaced with NAN.

编辑：

这是我们的实际数据。

This is our actual data.

Name              Name2                  DateTime
\t\t-\r\n\t      \t\t-\r\n\t            2016-06-10 05:22
\t\t-\r\n\t      \t\t-\r\n\t            2016-06-10 05:23
\t\t-\r\n\t      \t\t-\r\n\t            2016-06-10 14:25
\t\tGuest\r\n\t  \t\tGuest1\r\n\t       2016-06-10 15:32

我用它来删除那些转义字符。

I used this to remove those escape characters.

df['Name'] = df['Name'].str.replace('\r','').str.replace('\t','').str.replace('\n','').str.replace('-','')

删除这些字符后，我不确定现在在那个地方插入了什么。

After removing those characters, I am not sure what gets inserted in that place now.

推荐答案

另一个找到<$ c $的解决方案c>长度数据，然后通过 布尔索引 替换长度 0 或 1 ：

Another solution which found length of data and then by boolean indexing replacing all data with length 0 or 1:

print (df.applymap(len))
   Name  Name2  DateTime
0     0      0        16
1     0      0        16
2     0      0        16
3     5      6        16

df[df.applymap(len) < 2] = np.nan
print (df)
    Name   Name2          DateTime
0    NaN     NaN  2016-06-10 05:22
1    NaN     NaN  2016-06-10 05:23
2    NaN     NaN  2016-06-10 14:25
3  Guest  Guest1  2016-06-10 15:32

这篇关于在pandas数据帧中用NAN替换空格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！