熊猫将列表的一列分为多列

问题:熊猫将列表的一列分为多列

我有一列的pandas DataFrame:

import pandas as pd

df = pd.DataFrame(
    data={
        "teams": [
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
        ]
    }
)

print(df)

输出:

       teams
0  [SF, NYG]
1  [SF, NYG]
2  [SF, NYG]
3  [SF, NYG]
4  [SF, NYG]
5  [SF, NYG]
6  [SF, NYG]

如何将列表的这一列分为两列?

I have a pandas DataFrame with one column:

import pandas as pd

df = pd.DataFrame(
    data={
        "teams": [
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
            ["SF", "NYG"],
        ]
    }
)

print(df)

Output:

       teams
0  [SF, NYG]
1  [SF, NYG]
2  [SF, NYG]
3  [SF, NYG]
4  [SF, NYG]
5  [SF, NYG]
6  [SF, NYG]

How can split this column of lists into 2 columns?


回答 0

您可以将DataFrame构造函数与lists创建者to_list

import pandas as pd

d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
                ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
print (df2)
       teams
0  [SF, NYG]
1  [SF, NYG]
2  [SF, NYG]
3  [SF, NYG]
4  [SF, NYG]
5  [SF, NYG]
6  [SF, NYG]

df2[['team1','team2']] = pd.DataFrame(df2.teams.tolist(), index= df2.index)
print (df2)
       teams team1 team2
0  [SF, NYG]    SF   NYG
1  [SF, NYG]    SF   NYG
2  [SF, NYG]    SF   NYG
3  [SF, NYG]    SF   NYG
4  [SF, NYG]    SF   NYG
5  [SF, NYG]    SF   NYG
6  [SF, NYG]    SF   NYG

对于新的DataFrame

df3 = pd.DataFrame(df2['teams'].to_list(), columns=['team1','team2'])
print (df3)
  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG

解决方案apply(pd.Series)非常慢:

#7k rows
df2 = pd.concat([df2]*1000).reset_index(drop=True)

In [121]: %timeit df2['teams'].apply(pd.Series)
1.79 s ± 52.5 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

In [122]: %timeit pd.DataFrame(df2['teams'].to_list(), columns=['team1','team2'])
1.63 ms ± 54.3 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

You can use DataFrame constructor with lists created by to_list:

import pandas as pd

d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
                ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
print (df2)
       teams
0  [SF, NYG]
1  [SF, NYG]
2  [SF, NYG]
3  [SF, NYG]
4  [SF, NYG]
5  [SF, NYG]
6  [SF, NYG]

df2[['team1','team2']] = pd.DataFrame(df2.teams.tolist(), index= df2.index)
print (df2)
       teams team1 team2
0  [SF, NYG]    SF   NYG
1  [SF, NYG]    SF   NYG
2  [SF, NYG]    SF   NYG
3  [SF, NYG]    SF   NYG
4  [SF, NYG]    SF   NYG
5  [SF, NYG]    SF   NYG
6  [SF, NYG]    SF   NYG

And for new DataFrame:

df3 = pd.DataFrame(df2['teams'].to_list(), columns=['team1','team2'])
print (df3)
  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG

Solution with apply(pd.Series) is very slow:

#7k rows
df2 = pd.concat([df2]*1000).reset_index(drop=True)

In [121]: %timeit df2['teams'].apply(pd.Series)
1.79 s ± 52.5 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

In [122]: %timeit pd.DataFrame(df2['teams'].to_list(), columns=['team1','team2'])
1.63 ms ± 54.3 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

回答 1

更简单的解决方案:

pd.DataFrame(df2["teams"].to_list(), columns=['team1', 'team2'])

Yield

  team1 team2
-------------
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG
7    SF   NYG

如果要拆分一列分隔字符串而不是列表,则可以类似地执行以下操作:

pd.DataFrame(df["teams"].str.split('<delim>', expand=True).values,
             columns=['team1', 'team2'])

Much simpler solution:

pd.DataFrame(df2["teams"].to_list(), columns=['team1', 'team2'])

Yields,

  team1 team2
-------------
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG
7    SF   NYG

If you wanted to split a column of delimited strings rather than lists, you could similarly do:

pd.DataFrame(df["teams"].str.split('<delim>', expand=True).values,
             columns=['team1', 'team2'])

回答 2

df2与使用tolist()以下解决方案的解决方案不同,此解决方案保留了DataFrame 的索引:

df3 = df2.teams.apply(pd.Series)
df3.columns = ['team1', 'team2']

结果如下:

  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG

This solution preserves the index of the df2 DataFrame, unlike any solution that uses tolist():

df3 = df2.teams.apply(pd.Series)
df3.columns = ['team1', 'team2']

Here’s the result:

  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG

回答 3

与提议的解决方案相比,似乎在语法上更简单,因此更容易记住。我假设该列在数据帧df中称为“元”:

df2 = pd.DataFrame(df['meta'].str.split().values.tolist())

There seems to be a syntactically simpler way, and therefore easier to remember, as opposed to the proposed solutions. I’m assuming that the column is called ‘meta’ in a dataframe df:

df2 = pd.DataFrame(df['meta'].str.split().values.tolist())

回答 4

根据先前的答案,这是另一个解决方案,它以更快的运行时间返回与df2.teams.apply(pd.Series)相同的结果:

pd.DataFrame([{x: y for x, y in enumerate(item)} for item in df2['teams'].values.tolist()], index=df2.index)

时间:

In [1]:
import pandas as pd
d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
                ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
df2 = pd.concat([df2]*1000).reset_index(drop=True)

In [2]: %timeit df2['teams'].apply(pd.Series)

8.27 s ± 2.73 s per loop (mean ± std. dev. of 7 runs, 1 loop each)

In [3]: %timeit pd.DataFrame([{x: y for x, y in enumerate(item)} for item in df2['teams'].values.tolist()], index=df2.index)

35.4 ms ± 5.22 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

Based on the previous answers, here is another solution which returns the same result as df2.teams.apply(pd.Series) with a much faster run time:

pd.DataFrame([{x: y for x, y in enumerate(item)} for item in df2['teams'].values.tolist()], index=df2.index)

Timings:

In [1]:
import pandas as pd
d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
                ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
df2 = pd.concat([df2]*1000).reset_index(drop=True)

In [2]: %timeit df2['teams'].apply(pd.Series)

8.27 s ± 2.73 s per loop (mean ± std. dev. of 7 runs, 1 loop each)

In [3]: %timeit pd.DataFrame([{x: y for x, y in enumerate(item)} for item in df2['teams'].values.tolist()], index=df2.index)

35.4 ms ± 5.22 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

回答 5

由于我的nan观察中有上述发现,上述解决方案对我不起作用dataframe。就我而言,df2[['team1','team2']] = pd.DataFrame(df2.teams.values.tolist(), index= df2.index)收益:

object of type 'float' has no len()

我使用列表理解来解决这个问题。这里是可复制的示例:

import pandas as pd
import numpy as np
d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
            ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
df2.loc[2,'teams'] = np.nan
df2.loc[4,'teams'] = np.nan
df2

输出:

        teams
0   [SF, NYG]
1   [SF, NYG]
2   NaN
3   [SF, NYG]
4   NaN
5   [SF, NYG]
6   [SF, NYG]

df2['team1']=np.nan
df2['team2']=np.nan

用列表理解来解决:

for i in [0,1]:
    df2['team{}'.format(str(i+1))]=[k[i] if isinstance(k,list) else k for k in df2['teams']]

df2

Yield:

    teams   team1   team2
0   [SF, NYG]   SF  NYG
1   [SF, NYG]   SF  NYG
2   NaN        NaN  NaN
3   [SF, NYG]   SF  NYG
4   NaN        NaN  NaN
5   [SF, NYG]   SF  NYG
6   [SF, NYG]   SF  NYG

The above solutions didn’t work for me since I have nan observations in my dataframe. In my case df2[['team1','team2']] = pd.DataFrame(df2.teams.values.tolist(), index= df2.index) yields:

object of type 'float' has no len()

I solve this using list comprehension. Here the replicable example:

import pandas as pd
import numpy as np
d1 = {'teams': [['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],
            ['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG'],['SF', 'NYG']]}
df2 = pd.DataFrame(d1)
df2.loc[2,'teams'] = np.nan
df2.loc[4,'teams'] = np.nan
df2

output:

        teams
0   [SF, NYG]
1   [SF, NYG]
2   NaN
3   [SF, NYG]
4   NaN
5   [SF, NYG]
6   [SF, NYG]

df2['team1']=np.nan
df2['team2']=np.nan

solving with list comprehension:

for i in [0,1]:
    df2['team{}'.format(str(i+1))]=[k[i] if isinstance(k,list) else k for k in df2['teams']]

df2

yields:

    teams   team1   team2
0   [SF, NYG]   SF  NYG
1   [SF, NYG]   SF  NYG
2   NaN        NaN  NaN
3   [SF, NYG]   SF  NYG
4   NaN        NaN  NaN
5   [SF, NYG]   SF  NYG
6   [SF, NYG]   SF  NYG

回答 6

清单理解

列表理解的简单实现(我的最爱)

df = pd.DataFrame([pd.Series(x) for x in df.teams])
df.columns = ['team_{}'.format(x+1) for x in df.columns]

输出定时:

CPU times: user 0 ns, sys: 0 ns, total: 0 ns
Wall time: 2.71 ms

输出:

team_1  team_2
0   SF  NYG
1   SF  NYG
2   SF  NYG
3   SF  NYG
4   SF  NYG
5   SF  NYG
6   SF  NYG

list comprehension

simple implementation with list comprehension ( my favorite)

df = pd.DataFrame([pd.Series(x) for x in df.teams])
df.columns = ['team_{}'.format(x+1) for x in df.columns]

timing on output:

CPU times: user 0 ns, sys: 0 ns, total: 0 ns
Wall time: 2.71 ms

output:

team_1  team_2
0   SF  NYG
1   SF  NYG
2   SF  NYG
3   SF  NYG
4   SF  NYG
5   SF  NYG
6   SF  NYG

回答 7

这是另一个使用df.transform和的解决方案df.set_index

>>> (df['teams']
       .transform([lambda x:x[0], lambda x:x[1]])
       .set_axis(['team1','team2'],
                  axis=1,
                  inplace=False)
    )

  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG

Here’s another solution using df.transform and df.set_index:

>>> (df['teams']
       .transform([lambda x:x[0], lambda x:x[1]])
       .set_axis(['team1','team2'],
                  axis=1,
                  inplace=False)
    )

  team1 team2
0    SF   NYG
1    SF   NYG
2    SF   NYG
3    SF   NYG
4    SF   NYG
5    SF   NYG
6    SF   NYG