groupby逗号分隔值在单个DataFrame列python / pandas中(groupby comma

groupby逗号分隔值在单个DataFrame列python / pandas中(groupby comma-separated values in single DataFrame column python/pandas)

作为一个例子，假设我有一个python pandas DataFrame，如下所示：

# PERSON THINGS 0 Joe Candy Corn, Popsicles 1 Jane Popsicles 2 John Candy Corn, Ice Packs 3 Lefty Ice Packs, Hot Dogs

我想使用熊猫groupby功能有以下输出：

THINGS COUNT Candy Corn 2 Popsicles 2 Ice Packs 2 Hot Dogs 1

我通常了解以下groupby命令：

df.groupby(['THINGS']).count()

但是输出不是单个项目，而是整个字符串。 我想我明白为什么会出现这种情况，但我不清楚如何最好地处理问题以获得所需的输出，而不是以下内容：

THINGS PERSON Candy Corn, Ice Packs 1 Candy Corn, Popsicles 1 Ice Packs, Hot Dogs 1 Popsicles 1

大熊猫是否有像SQL中的LIKE这样的函数，或者我在考虑在熊猫中如何做到这一点？

任何援助赞赏。

As an example, let's say I have a python pandas DataFrame that is the following:

# PERSON THINGS 0 Joe Candy Corn, Popsicles 1 Jane Popsicles 2 John Candy Corn, Ice Packs 3 Lefty Ice Packs, Hot Dogs

I would like to use the pandas groupby functionality to have the following output:

THINGS COUNT Candy Corn 2 Popsicles 2 Ice Packs 2 Hot Dogs 1

I generally understand the following groupby command:

df.groupby(['THINGS']).count()

But the output is not by individual item, but by the entire string. I think I understand why this is, but it's not clear to me how to best approach the problem to get the desired output instead of the following:

THINGS PERSON Candy Corn, Ice Packs 1 Candy Corn, Popsicles 1 Ice Packs, Hot Dogs 1 Popsicles 1

Does pandas have a function like the LIKE in SQL, or am I thinking about how to do this wrong in pandas?

Any assistance appreciated.

最满意答案

通过分词来创建一个系列，并使用value_counts

In [292]: pd.Series(df.THINGS.str.cat(sep=', ').split(', ')).value_counts() Out[292]: Popsicles 2 Ice Packs 2 Candy Corn 2 Hot Dogs 1 dtype: int64

Create a series by splitting words, and use value_counts

In [292]: pd.Series(df.THINGS.str.cat(sep=', ').split(', ')).value_counts() Out[292]: Popsicles 2 Ice Packs 2 Candy Corn 2 Hot Dogs 1 dtype: int64