在 Pandas 数据框中填充货币缺失数据

Question

我正在从 Quandl 中提取货币数据：

import Quandl
df = Quandl.get("CURRFX/USDEUR", collapse="daily") 
df = df.tail(10).reset_index()

运行上面的代码，可以看到周末当然没有货币数据了

数据框不显示缺失的天数行。

知道如何添加缺失的天数并用前一周的平均值填充它们吗？

更新：

我的代码是：

 exchange = Quandl.get("CURRFX/USDEUR", trim_start="2016-02-01",trim_end="2016-02-28")
 exchange = exchange.drop(['High (est)','Low (est)'], axis = 1) 
 exchange = exchange.reindex(pd.date_range(start=exchange.index[0], end=exchange.index[-1])).ffill().reset_index()
 exchange = exchange.rename(columns={'index':'Day','Rate':'exchange'})

当然，想象一下我运行星期一早上，我仍然没有整个日期的数据，所以星期天和星期六保持空...

如何填写？

Answer 1

IIUC 然后你可以 reindex and pass a new DatetimeIndex generated from date_range 并调用 ffill:

In [115]:
df.reindex(pd.date_range(start=df.index[0], end=df.index[-1])).ffill()

Out[115]:
                Rate  High (est)  Low (est)
1999-09-06  0.941019     0.95269    0.92949
1999-09-07  0.945500     0.95636    0.93476
1999-09-08  0.944376     0.95588    0.93301
1999-09-09  0.943697     0.95412    0.93339
1999-09-10  0.948008     0.00000    0.00000
1999-09-11  0.948008     0.00000    0.00000
1999-09-12  0.948008     0.00000    0.00000
1999-09-13  0.959838     0.97143    0.94838
1999-09-14  0.964873     0.97623    0.95365
1999-09-15  0.965378     0.97708    0.95382
1999-09-16  0.963142     0.97393    0.95248
1999-09-17  0.961971     0.00000    0.00000
1999-09-18  0.961971     0.00000    0.00000
1999-09-19  0.961971     0.00000    0.00000
1999-09-20  0.960917     0.00000    0.00000
1999-09-21  0.959907     0.97166    0.94829
1999-09-22  0.954640     0.96633    0.94309
1999-09-23  0.955403     0.96649    0.94444
1999-09-24  0.955754     0.00000    0.00000
1999-09-25  0.955754     0.00000    0.00000
1999-09-26  0.955754     0.00000    0.00000
1999-09-27  0.956281     0.96693    0.94575
1999-09-28  0.955362     0.00000    0.00000
1999-09-29  0.949949     0.96176    0.93829
1999-09-30  0.942264     0.00000    0.00000
1999-10-01  0.934798     0.00000    0.00000
1999-10-02  0.934798     0.00000    0.00000
1999-10-03  0.934798     0.00000    0.00000
1999-10-04  0.932772     0.00000    0.00000
1999-10-05  0.934479     0.94584    0.92326
...              ...         ...        ...
2016-01-17  0.917878     0.00000    0.00000
2016-01-18  0.917451     0.92788    0.90714
2016-01-19  0.918046     0.92896    0.90726
2016-01-20  0.916549     0.92688    0.90634
2016-01-21  0.918055     0.92901    0.90723
2016-01-22  0.921764     0.00000    0.00000
2016-01-23  0.921764     0.00000    0.00000
2016-01-24  0.921764     0.00000    0.00000
2016-01-25  0.923926     0.00000    0.00000
2016-01-26  0.922914     0.93427    0.91170
2016-01-27  0.920722     0.93114    0.91042
2016-01-28  0.917792     0.92875    0.90696
2016-01-29  0.917405     0.00000    0.00000
2016-01-30  0.917405     0.00000    0.00000
2016-01-31  0.917405     0.00000    0.00000
2016-02-01  0.919592     0.00000    0.00000
2016-02-02  0.918263     0.92859    0.90805
2016-02-03  0.913820     0.92514    0.90264
2016-02-04  0.903148     0.91308    0.89332
2016-02-05  0.895662     0.90574    0.88569
2016-02-06  0.895662     0.90574    0.88569
2016-02-07  0.895662     0.90574    0.88569
2016-02-08  0.896030     0.00000    0.00000
2016-02-09  0.892736     0.90329    0.88231
2016-02-10  0.888489     0.89846    0.87863
2016-02-11  0.885288     0.89508    0.87560
2016-02-12  0.885464     0.00000    0.00000
2016-02-13  0.885464     0.00000    0.00000
2016-02-14  0.885464     0.00000    0.00000
2016-02-15  0.890435     0.00000    0.00000

[6007 rows x 3 columns]

在 Pandas 数据框中填充货币缺失数据

Filling currency missing data in Pandas Dataframe

python

currency

missing-data

pandas

quandl