pandas 对数据框中列中的数据进行分类
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/38936854/
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
Categorize Data in a column in dataframe
提问by Nathaniel Babalola
I have a column of numbers in my dataframe, i want to categorize these numbers into e.g high , low, excluded. How do i accomplish that. I am clueless , i have tried looking at the cut function and category datatype.
我的数据框中有一列数字，我想将这些数字分类为例如高、低、排除。我如何做到这一点。我一无所知，我曾尝试查看剪切函数和类别数据类型。
回答by ptrj
A short example with pd.cut.
一个简短的例子pd.cut。
Let's start with some data frame:
让我们从一些数据框开始：
df = pd.DataFrame({'A': [0, 8, 2, 5, 9, 15, 1]})
and, say, we want to assign the numbers to the following categories: 'low'if a number is in the interval [0, 2], 'mid'for (2, 8], 'high'for (8, 10], and we exclude numbers above 10 (or below 0).
并且，比如说，我们想要将数字分配到以下类别：'low'如果一个数字在区间[0, 2]、'mid'for (2, 8]、'high'for 中(8, 10]，我们排除大于 10（或小于 0）的数字。
Thus, we have 3 bins with edges: 0, 2, 8, 10. Now, we can use cutas follows:
因此，我们有 3 个带边的 bin：0、2、8、10。现在，我们可以使用cut如下：
pd.cut(df['A'], bins=[0, 2, 8, 10], include_lowest=True)
Out[33]:
0 [0, 2]
1 (2, 8]
2 [0, 2]
3 (2, 8]
4 (8, 10]
5 NaN
6 [0, 2]
Name: A, dtype: category
Categories (3, object): [[0, 2] < (2, 8] < (8, 10]]
The argument include_lowest=Trueincludes the left end of the first interval. (If you want intervals open on the right, then use right=False.)
参数include_lowest=True包括第一个间隔的左端。（如果您希望在右侧打开间隔，请使用right=False。）
The intervals are probably not the best names for the categories. So, let's use names: low/mid/high:
间隔可能不是类别的最佳名称。所以，让我们使用名称low/mid/high：
pd.cut(df['A'], bins=[0, 2, 8, 10], include_lowest=True, labels=['low', 'mid', 'high'])
Out[34]:
0 low
1 mid
2 low
3 mid
4 high
5 NaN
6 low
Name: A, dtype: category
Categories (3, object): [low < mid < high]
The excluded number 15 gets a "category" NaN. If you prefer a more meaningful name, probably the simplest solution (there're other ways to deal with NaN's) is to add another bin and a category name, for example:
排除的数字 15 获得一个“类别” NaN。如果您更喜欢更有意义的名称，可能最简单的解决方案（还有其他方法可以处理 NaN）是添加另一个 bin 和类别名称，例如：
pd.cut(df['A'], bins=[0, 2, 8, 10, 1000], include_lowest=True, labels=['low', 'mid', 'high', 'excluded'])
Out[35]:
0 low
1 mid
2 low
3 mid
4 high
5 excluded
6 low
Name: A, dtype: category
Categories (4, object): [low < mid < high < excluded]
回答by draco_alpine
This question is pretty broad, but a good place to start might be this page in the documentation:
这个问题非常广泛，但一个很好的起点可能是文档中的这个页面：
http://pandas.pydata.org/pandas-docs/stable/indexing.html#boolean-indexing
http://pandas.pydata.org/pandas-docs/stable/indexing.html#boolean-indexing
Or you could look into numpy.where
或者你可以看看 numpy.where
import numpy as np
df['is_high'] = np.where(df.['column_of_interest'] > 5 ,1,0)
