如何重新排序索引行的基础上列表熊猫数据框架

我有一个这样的数据框架:

company  Amazon  Apple  Yahoo
name
A             0    130      0
C           173      0      0
Z             0      0    150

它是用以下代码创建的:

import pandas as pd
df = pd.DataFrame({'name' : ['A', 'Z','C'],
'company' : ['Apple', 'Yahoo','Amazon'],
'height' : [130, 150,173]})


df = df.pivot(index="name", columns="company", values="height").fillna(0)

我想要做的是根据一个预定义的列表对行(使用 name索引)进行排序:

["Z", "C", "A"]`

结果是:

company  Amazon  Apple  Yahoo
name
Z             0      0    150
C           173      0      0
A             0    130      0

我怎么才能做到呢？

154159 次浏览
小开
最佳答案

You could set index on predefined order using reindex like

In [14]: df.reindex(["Z", "C", "A"])
Out[14]:
company  Amazon  Apple  Yahoo
Z             0      0    150
C           173      0      0
A             0    130      0

However, if it's alphabetical order, you could use sort_index(ascending=False)

In [12]: df.sort_index(ascending=False)
Out[12]:
company  Amazon  Apple  Yahoo
name
Z             0      0    150
C           173      0      0
A             0    130      0

Like pointed below, you need to assign it to some variable

In [13]: df = df.sort_index(ascending=False)
小开

We could also use loc:

lst = ["Z", "C", "A"]
df = df.loc[lst]

Output:

company  Amazon  Apple  Yahoo
name
Z             0      0    150
C           173      0      0
A             0    130      0

Note that if there are values in lst that does not exist in df.index (e.g. if lst=['Z','C','A','D']), then loc throws a KeyError (whereas reindex creates a new row 'D' full of NaNs).

MultiIndex

If df is MultiIndex, such as:

           C3
C1 C2
2  evelen   0
ten      1
twelve   2
1  evelen   3
ten      4
twelve   5

and if you want to sort the second level by ten,eleven andtwelve, then using loc:

out = df.loc[:, ['ten','evelen', 'twelve'],:]

Output:

           C3
C1 C2
2  evelen   0
ten      1
twelve   2
1  evelen   3
ten      4
twelve   5

and for both levels:

out = df.loc[[1,2], ['ten','evelen','twelve'], :]

Output:

           C3
C1 C2
1  ten      4
evelen   3
twelve   5
2  ten      1
evelen   0
twelve   2
小开

IMHO, specially if you want to sort by multiples values, the best solution is:

df = df.set_index("C1")
df = df.sort_values(["C1", "C2"])
df.reset_index(inplace=True)

提交答案