展望未来,interpolate
效果很好:
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,它没有解决起始行(仅前进)。 limit_direction
参数允许{‘forward’, ‘backward’, ‘both’}
。这些都不起作用。有向后插值的正确方法吗?
我们可以假设一系列递增或递减1,这可能不会从0开始,因为在这个例子中恰好。
答案 0 :(得分:1)
它似乎仅适用于参数limit
,请参阅docs [In 47]:
添加 limit_direction 关键字参数,该参数与限制一起使用,以启用插值以向前,向后或同时填充NaN值(GH9218,GH10420,GH11115)
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