如何将一行的值拆分为 pandas 中的不同列？

Question

我在 pandas 中有以下 dataframe：

a = ['[16.01319488  6.1095932  -0.14837995]',
 '[16.10400501  6.23724404 -0.1727245 ]',
 '[16.195107    6.36434895 -0.19695716]',
 '[16.2864465   6.49178233 -0.22124142]',
 '[16.37796913  6.62041857 -0.24574078]',
 '[16.46962054  6.75113206 -0.27061875]',
 '[16.56134636  6.88479719 -0.29603881]',
 '[16.65309334  7.02229002 -0.32216479]',
 '[16.74480491  7.16448166 -0.34915957]',
 '[16.83642781  7.31224812 -0.37718693]',
 '[16.92790769  7.46646379 -0.4064104 ]',
 '[17.0190533   7.62784345 -0.4369622 ]',
 '[17.10912594  7.79646343 -0.46884957]',
 '[17.19725     7.97224045 -0.50204846]']
b = [0.0,
 0.01999999989745438,
 0.03999999979490875,
 0.05999999969236312,
 0.0799999995898175,
 0.09999999948727188,
 0.1199999993847262,
 0.1399999992821806,
 0.159999999179635,
 0.1799999990770894,
 0.1999999989745438,
 0.2199999988719981,
 0.2399999987694525,
 0.2599999986669069]
b
dDictionary = {
  'A':a,
  'B': b
  } 
test = pd.DataFrame(dDictionary)

“A”列中的每个值都包含三个值，我想将它们分成三个单独的列。 有没有一种简单而强大的方法来做到这一点？

Answer 1

将Series.str.strip与Series.str.split一起使用并转换为浮点数：

test[['c','d','e']] = test.A.str.strip('[]').str.split(expand=True).astype(float)
print (test)
                                        A     B          c         d         e
0   [16.01319488  6.1095932  -0.14837995]  0.00  16.013195  6.109593 -0.148380
1   [16.10400501  6.23724404 -0.1727245 ]  0.02  16.104005  6.237244 -0.172725
2   [16.195107    6.36434895 -0.19695716]  0.04  16.195107  6.364349 -0.196957
3   [16.2864465   6.49178233 -0.22124142]  0.06  16.286447  6.491782 -0.221241
4   [16.37796913  6.62041857 -0.24574078]  0.08  16.377969  6.620419 -0.245741
5   [16.46962054  6.75113206 -0.27061875]  0.10  16.469621  6.751132 -0.270619
6   [16.56134636  6.88479719 -0.29603881]  0.12  16.561346  6.884797 -0.296039
7   [16.65309334  7.02229002 -0.32216479]  0.14  16.653093  7.022290 -0.322165
8   [16.74480491  7.16448166 -0.34915957]  0.16  16.744805  7.164482 -0.349160
9   [16.83642781  7.31224812 -0.37718693]  0.18  16.836428  7.312248 -0.377187
10  [16.92790769  7.46646379 -0.4064104 ]  0.20  16.927908  7.466464 -0.406410
11  [17.0190533   7.62784345 -0.4369622 ]  0.22  17.019053  7.627843 -0.436962
12  [17.10912594  7.79646343 -0.46884957]  0.24  17.109126  7.796463 -0.468850
13  [17.19725     7.97224045 -0.50204846]  0.26  17.197250  7.972240 -0.502048

如果需要删除A使用DataFrame.pop ：

test[['c','d','e']] = test.pop('A').str.strip('[]').str.split(expand=True).astype(float)
print (test)
       B          c         d         e
0   0.00  16.013195  6.109593 -0.148380
1   0.02  16.104005  6.237244 -0.172725
2   0.04  16.195107  6.364349 -0.196957
3   0.06  16.286447  6.491782 -0.221241
4   0.08  16.377969  6.620419 -0.245741
5   0.10  16.469621  6.751132 -0.270619
6   0.12  16.561346  6.884797 -0.296039
7   0.14  16.653093  7.022290 -0.322165
8   0.16  16.744805  7.164482 -0.349160
9   0.18  16.836428  7.312248 -0.377187
10  0.20  16.927908  7.466464 -0.406410
11  0.22  17.019053  7.627843 -0.436962
12  0.24  17.109126  7.796463 -0.468850
13  0.26  17.197250  7.972240 -0.502048

Answer 2

这是使用pandas.concat动态扩展'A'列的另一种方法。

expanded_test = (pd.concat([test['A'].str.strip('[]').str.split(expand=True)], 
                    axis=1, keys=test.columns)
                )
expanded_test.columns = expanded_test.columns.map(lambda x: '_'.join((x[0], str(x[1]+1))))

out = test.join(expanded_test)

>>> print(out)

                                        A     B          A_1         A_2          A_3
0   [16.01319488  6.1095932  -0.14837995]  0.00  16.01319488   6.1095932  -0.14837995
1   [16.10400501  6.23724404 -0.1727245 ]  0.02  16.10400501  6.23724404   -0.1727245
2   [16.195107    6.36434895 -0.19695716]  0.04    16.195107  6.36434895  -0.19695716
3   [16.2864465   6.49178233 -0.22124142]  0.06   16.2864465  6.49178233  -0.22124142
4   [16.37796913  6.62041857 -0.24574078]  0.08  16.37796913  6.62041857  -0.24574078
5   [16.46962054  6.75113206 -0.27061875]  0.10  16.46962054  6.75113206  -0.27061875
6   [16.56134636  6.88479719 -0.29603881]  0.12  16.56134636  6.88479719  -0.29603881
7   [16.65309334  7.02229002 -0.32216479]  0.14  16.65309334  7.02229002  -0.32216479
8   [16.74480491  7.16448166 -0.34915957]  0.16  16.74480491  7.16448166  -0.34915957
9   [16.83642781  7.31224812 -0.37718693]  0.18  16.83642781  7.31224812  -0.37718693
10  [16.92790769  7.46646379 -0.4064104 ]  0.20  16.92790769  7.46646379   -0.4064104
11  [17.0190533   7.62784345 -0.4369622 ]  0.22   17.0190533  7.62784345   -0.4369622
12  [17.10912594  7.79646343 -0.46884957]  0.24  17.10912594  7.79646343  -0.46884957
13  [17.19725     7.97224045 -0.50204846]  0.26     17.19725  7.97224045  -0.50204846
[Finished in 1.0s]