Python Pandas 应用函数将多个值返回到 Pandas 数据帧中的行
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/23690284/
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
pandas apply function that returns multiple values to rows in pandas dataframe
提问by Fra
I have a dataframe with a timeindex and 3 columns containing the coordinates of a 3D vector:
我有一个带有时间索引和包含 3D 矢量坐标的 3 列的数据框：
x y z
ts
2014-05-15 10:38 0.120117 0.987305 0.116211
2014-05-15 10:39 0.117188 0.984375 0.122070
2014-05-15 10:40 0.119141 0.987305 0.119141
2014-05-15 10:41 0.116211 0.984375 0.120117
2014-05-15 10:42 0.119141 0.983398 0.118164
I would like to apply a transformation to each row that also returns a vector
我想对每一行应用一个转换，它也返回一个向量
def myfunc(a, b, c):
do something
return e, f, g
but if I do:
但如果我这样做：
df.apply(myfunc, axis=1)
I end up with a Pandas series whose elements are tuples. This is beacause apply will take the result of myfunc without unpacking it. How can I change myfunc so that I obtain a new df with 3 columns?
我最终得到了一个 Pandas 系列，它的元素是元组。这是因为 apply 将在不解包的情况下获取 myfunc 的结果。如何更改 myfunc 以便获得具有 3 列的新 df？
Edit:
编辑：
All solutions below work. The Series solution does allow for column names, the List solution seem to execute faster.
下面的所有解决方案都有效。Series 解决方案确实允许列名，List 解决方案似乎执行得更快。
def myfunc1(args):
e=args[0] + 2*args[1]
f=args[1]*args[2] +1
g=args[2] + args[0] * args[1]
return pd.Series([e,f,g], index=['a', 'b', 'c'])
def myfunc2(args):
e=args[0] + 2*args[1]
f=args[1]*args[2] +1
g=args[2] + args[0] * args[1]
return [e,f,g]
%timeit df.apply(myfunc1 ,axis=1)
100 loops, best of 3: 4.51 ms per loop
%timeit df.apply(myfunc2 ,axis=1)
100 loops, best of 3: 2.75 ms per loop
采纳答案by Happy001
Just return a list instead of tuple.
只需返回一个列表而不是元组。
In [81]: df
Out[81]:
x y z
ts
2014-05-15 10:38:00 0.120117 0.987305 0.116211
2014-05-15 10:39:00 0.117188 0.984375 0.122070
2014-05-15 10:40:00 0.119141 0.987305 0.119141
2014-05-15 10:41:00 0.116211 0.984375 0.120117
2014-05-15 10:42:00 0.119141 0.983398 0.118164
[5 rows x 3 columns]
In [82]: def myfunc(args):
....: e=args[0] + 2*args[1]
....: f=args[1]*args[2] +1
....: g=args[2] + args[0] * args[1]
....: return [e,f,g]
....:
In [83]: df.apply(myfunc ,axis=1)
Out[83]:
x y z
ts
2014-05-15 10:38:00 2.094727 1.114736 0.234803
2014-05-15 10:39:00 2.085938 1.120163 0.237427
2014-05-15 10:40:00 2.093751 1.117629 0.236770
2014-05-15 10:41:00 2.084961 1.118240 0.234512
2014-05-15 10:42:00 2.085937 1.116202 0.235327
回答by U2EF1
Return Seriesand it will put them in a DataFrame.
返回Series，它将把它们放在一个 DataFrame 中。
def myfunc(a, b, c):
do something
return pd.Series([e, f, g])
This has the bonus that you can give labels to each of the resulting columns. If you return a DataFrame it just inserts multiple rows for the group.
这有一个好处，您可以为每个结果列提供标签。如果您返回一个 DataFrame，它只会为该组插入多行。
回答by Fra
Found a possible solution, by changing myfunc to return an np.array like this:
找到了一个可能的解决方案，通过改变 myfunc 返回一个 np.array 像这样：
import numpy as np
def myfunc(a, b, c):
do something
return np.array((e, f, g))
any better solution?
任何更好的解决方案？
回答by Dennis Golomazov
Based on the excellent answerby @U2EF1, I've created a handy function that applies a specified function that returns tuples to a dataframe field, and expands the result back to the dataframe.
基于@U2EF1的出色回答，我创建了一个方便的函数，该函数应用指定的函数将元组返回到数据帧字段，并将结果扩展回数据帧。
def apply_and_concat(dataframe, field, func, column_names):
return pd.concat((
dataframe,
dataframe[field].apply(
lambda cell: pd.Series(func(cell), index=column_names))), axis=1)
Usage:
用法：
df = pd.DataFrame([1, 2, 3], index=['a', 'b', 'c'], columns=['A'])
print df
A
a 1
b 2
c 3
def func(x):
return x*x, x*x*x
print apply_and_concat(df, 'A', func, ['x^2', 'x^3'])
A x^2 x^3
a 1 1 1
b 2 4 8
c 3 9 27
Hope it helps someone.
希望它可以帮助某人。
回答by Genarito
I've tried returning a tuple (I was using functions like scipy.stats.pearsonrwhich return that kind of structures) but It returned a 1D Series instead of a Dataframe which was I expected. If I created a Series manually the performance was worse, so I fixed It using the result_typeas explained in the official API documentation:
我试过返回一个元组（我正在使用类似scipy.stats.pearsonr返回那种结构的函数），但它返回了一个 1D 系列而不是我期望的数据帧。如果我手动创建一个系列，性能会更差，所以我使用官方 API 文档result_type中的解释来修复它：
Returning a Series inside the function is similar to passing result_type='expand'. The resulting column names will be the Series index.
在函数内部返回一个 Series 类似于传递 result_type='expand'。结果列名将是系列索引。
So you could edit your code this way:
所以你可以这样编辑你的代码：
def myfunc(a, b, c):
# do something
return (e, f, g)
df.apply(myfunc, axis=1, result_type='expand')
