python - 在数据框中查找不同日期的差异和标准

Question

我有一个包含 2 列“日期”和“温度”的数据框:

              Temp
Date               
2017-08-31  43.8532
2017-08-30  44.1702
2017-08-29  43.7581
2017-08-28  43.8680
2017-08-25  43.5181
2017-08-24  43.9274
2017-08-23  43.6600
2017-08-22  43.5773
2017-08-21  43.7247
2017-08-18  43.7179
2017-08-17  43.7072
2017-08-16  43.3215
2017-08-15  43.3597
2017-08-14  43.3082
2017-08-11  43.5351
2017-08-10  43.6848
...         ...
31/08/2016  37.873
30/08/2016  38.6831
29/08/2016  38.7264
26/08/2016  39.1932
25/08/2016  39.1318
...         ...

我想创建 2 个新列 '1 yr diff' 和 '1 yr std'

'1 yr diff' 是一年中相隔两个日期之间的温差。

'1 yr std' 是过去一年温度的标准偏差。

前三个手工计算:

Date    Temp    1 yr diff   1 yr std
31/08/2017  43.8532 5.9802  1.914
30/08/2017  44.1702 5.4871  1.909
29/08/2017  43.7581 5.0317  1.901

如何对所有行执行此操作？

Answer 1

diff = df.Temp.resample('D').ffill().diff(365).rename('1 yr diff')
std = df.Temp.rolling('365D').std().rename('1 yr std')
pd.concat([df, diff, std], axis=1).dropna()

                Temp  1 yr diff  1 yr std
Date                                     
2017-06-30  0.467765  -0.421425  0.287502
2017-06-29  0.293493  -0.479262  0.288638
2017-06-28  0.089505  -0.080712  0.288987
2017-06-27  0.156842  -0.254886  0.288611
2017-06-26  0.061712  -0.286930  0.287945
2017-06-23  0.691029   0.623226  0.286886
2017-06-22  0.135094  -0.120836  0.287729
2017-06-21  0.597864  -0.286051  0.287270
2017-06-20  0.594055   0.008702  0.288266
2017-06-19  0.823582   0.509880  0.288254
...