pandas 用 NaN 替换 Inf 导致 AttributeError
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/23344498/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Replacing Inf by NaN results in AttributeError
提问by Amelio Vazquez-Reina
I have a strange problem in Pandas. I want to replace any entry that has np.Infwith the value np.NaN. However, when I do:
我在 Pandas 中有一个奇怪的问题。我想np.Inf用值替换任何条目np.NaN。但是，当我这样做时：
df[df == np.Inf] = np.NaN
I get:
我得到：
AttributeError: 'float' object has no attribute 'view'
The statement that produces the error specifically is:
具体产生错误的语句是：
df == np.Inf
I wonder if the problem is that I am running the above on a Dataframe with mixed types (see dtypesbelow). But if that is the case, how can I do this replacement still automatically?
我想知道问题是否是我在混合类型的 Dataframe 上运行上述内容（见dtypes下文）。但如果是这种情况，我如何才能仍然自动进行此替换？
In: df.dtypes
Out:
Year int64
Week int64
item_name object
item_uid object
Algorithm object
item Start float64
item 1/4 float64
item 1/2 float64
item 3/4 float64
item Complete float64
Daily Nr Impressions float64
date datetime64[ns]
Weekly rate float64
dtype: object
回答by Ffisegydd
You can use df.replaceto replace your np.infvalues.
您可以使用df.replace来替换您的np.inf值。
In [9]: import pandas as pd
In [10]: df = pd.DataFrame([1, 2, np.inf])
In [11]: df.replace(np.inf, np.nan)
Out[11]:
0
0 1
1 2
2 NaN
[3 rows x 1 columns]
