在大数据透视表中获取cumprod（）

Question

I want to calculate the geometric return (cumulative return) per annum for a pivot table containing monthly return and add this as a column to the table. The pivot table looks like this:

In [474]: piv.tail(3)
Out[474]: 
month   1    2    3    4    5    6    7    8    9    10   11   12
year                                                             
2015   3.0 -5.6  1.5 -0.1  0.3 -1.7  0.5 -7.0  0.3 -0.1  0.4  0.0
2016   5.2  4.7 -0.3  1.3 -6.1  0.3  3.5 -0.1  1.2 -1.9  0.7  0.9
2017   2.5  2.1  0.1  2.6  4.8  0.1  2.8  0.1 -2.1  0.5  NaN  NaN

tried

piv['annual'] = (piv+1).cumprod(axis=1)-1

Which throws an error. Stuck here. any help appreciated.

Answer 1

1. These returns are way too big to be in decimal space. They must be in percent space. Therefore, divide by 100.
2. Use prod

---

piv.assign(annual=(piv / 100 + 1).prod(1) * 100 - 100)

Month    1    2    3    4    5    6    7    8    9   10   11   12     annual
Year                                                                        
2015   3.0 -5.6  1.5 -0.1  0.3 -1.7  0.5 -7.0  0.3 -0.1  0.4  0.0  -8.600390
2016   5.2  4.7 -0.3  1.3 -6.1  0.3  3.5 -0.1  1.2 -1.9  0.7  0.9   9.272164
2017   2.5  2.1  0.1  2.6  4.8  0.1  2.8  0.1 -2.1  0.5  NaN  NaN  14.156949