本文介绍了 pandas 在数据帧中向后插值（）的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！
问题描述
展望未来，效果很好：
Going forward, interpolate works great:
name days
0 a NaN
1 a NaN
2 a 2
3 a 3
4 a NaN
5 a NaN
records.loc[:, 'days'].interpolate(method='linear', inplace=True)
name days
0 a NaN
1 a NaN
2 a 2
3 a 3
4 a 4
5 a 5
...但是，它不处理开始的行（仅向前）。 limit_direction 参数允许 {前进，后退，两者} 。这些都不起作用。有没有适当的方法可以向后插值？
...however, it does not address the beginning rows (only goes forward). The limit_direction param allows {‘forward’, ‘backward’, ‘both’}. None of these works. Is there a proper way to interpolate backwards?
我们可以假设序列递增或递减1，但在本例中可能不会从0开始。 / p>
We can assume a series incrementing or decrementing by 1, which may not start at 0 as it happens to in this example.
推荐答案
似乎仅与参数 limit 一起使用，请参见：
It seems it works only with parameter limit see docs [In 47]:
records = pd.DataFrame(
{'name': {0: 'a', 1: 'a', 2: 'a', 3: 'a', 4: 'a', 5: 'a', 6: 'a', 7: 'a', 8: 'a', 9: 'a'},
'days': {0: 0.0, 1: np.nan, 2: np.nan, 3: np.nan, 4: 4.0, 5: 5.0, 6: np.nan, 7: np.nan, 8: np.nan, 9: 9.0}},
columns=['name','days'])
print (records)
name days
0 a 0.0
1 a NaN
2 a NaN
3 a NaN
4 a 4.0
5 a 5.0
6 a NaN
7 a NaN
8 a NaN
9 a 9.0
#by default limit_direction='forward'
records['forw'] = records['days'].interpolate(method='linear',
limit=1)
records['backw'] = records['days'].interpolate(method='linear',
limit_direction='backward',
limit=1)
records['both'] = records['days'].interpolate(method='linear',
limit_direction='both',
limit=1)
print (records)
name days forw backw both
0 a 0.0 0.0 0.0 0.0
1 a NaN 1.0 NaN 1.0
2 a NaN NaN NaN NaN
3 a NaN NaN 3.0 3.0
4 a 4.0 4.0 4.0 4.0
5 a 5.0 5.0 5.0 5.0
6 a NaN 6.0 NaN 6.0
7 a NaN NaN NaN NaN
8 a NaN NaN 8.0 8.0
9 a 9.0 9.0 9.0 9.0
这篇关于 pandas 在数据帧中向后插值（）的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！