我有一个数据框，其中包含按产品分类的季度销售额。
edit我正尝试使用上一季度的结果来填补明智的产品的缺失值。我尝试使用方法选项为“ ffill”的fillna。但它会填充值，而与类别无关。
data=pd.DataFrame({"Product":("A","A","A","A","B","B","B","B","C","C","C","C"),
"Quarter":("Q1","Q2","Q3","Q4","Q1","Q2","Q3","Q4","Q1","Q2","Q3","Q4"),
"Sales":('N/A','N/A',10,10,'N/A','N/A','N/A','N/A','N/A',30,'N/A',20)
})
data["Sales"].replace('N/A',np.NaN,inplace=True)
Product Quarter Sales
0 A Q1 NaN
1 A Q2 NaN
2 A Q3 10.0
3 A Q4 10.0
4 B Q1 NaN
5 B Q2 NaN
6 B Q3 NaN
7 B Q4 NaN
8 C Q1 NaN
9 C Q2 30.0
10 C Q3 NaN
11 C Q4 20.0
获得的o / p：
data["Obtained Sales"]=data["Values"].fillna(method='ffill')
但是如上所述预期的输出：
Product Quarter Sales Obtained Sales
0 A Q1 NaN NaN
1 A Q2 NaN NaN
2 A Q3 10.0 10.0
3 A Q4 10.0 10.0
4 B Q1 NaN 10.0
5 B Q2 NaN 10.0
6 B Q3 NaN 10.0
7 B Q4 NaN 10.0
8 C Q1 NaN 10.0
9 C Q2 30.0 30.0
10 C Q3 NaN 30.0
11 C Q4 20.0 20.0
如何使用熊猫通过 Groups 用以前的非零值填充空白？