我需要合并两个数据框。第一个是:
page value
shoes 554
sneakers 226
sandals 114
boots 821
T-shirt 213
mobile-phone 284
laptop 361
第二个数据帧是:
path1 path2 path3 path4
fashion footwear shoes-and-other shoes
fashion footwear shoes-and-other sneakers
fashion footwear sandals NaN
fashion footwear shirts T-shirt
electronic devices mobile-and-tablet mobile-phone
electronic devices laptop NaN
我的预期输出将是:
path1 path2 path3 path4 page value
fashion footwear shoes-and-other shoes shoes 554
fashion footwear shoes-and-other sneakers sneakers 226
fashion footwear sandals NaN sandals 114
fashion footwear shirts T-shirt T-shirt 213
electronic devices mobile-and-tablet mobile-phone mobile-phone 284
electronic devices laptop NaN laptop 361
如果第一个数据帧中的任何page
字符串存在于path1
或path2
或path3
或{第二个数据帧的{1}}列。请注意,第一个数据帧的path4
可能与第二个数据帧的page
匹配,我有很多情况。
有没有简单的pythonic方法?
答案 0 :(得分:3)
让我们尝试where
与ffill
创建合并密钥,然后merge
df1['page'] = df1.where(df1.isin(df.page.tolist())).ffill(1).iloc[:,-1]
df1 = df1.merge(df, how='left')
df1
Out[131]:
path1 path2 path3 path4 page value
0 fashion footwear shoes-and-other shoes shoes 554
1 fashion footwear shoes-and-other sneakers sneakers 226
2 fashion footwear sandals NaN sandals 114
3 fashion footwear shirts T-shirt T-shirt 213
4 electronic devices mobile-and-tablet mobile-phone mobile-phone 284
5 electronic devices laptop NaN laptop 361