Def first_ods (df) : 
Df2=df. Groupby (' the consignee phone number/DiHuoRen mobile phone number) [' order creation time]. Apply (min). Reset_index () 
. 
Return df3 
Df1. Apply (first_ods) 

Error has written many, the first line should be prompt function without the problem in this field, if you have any of this field, but the df is direct operation can be done, I don't have to function so that the apply of the incoming is not a whole table, but the process?

That if we can spread to the whole table, that this need to calculated the function of the other columns may not be able to use it is

In addition, this is I write each user first single, there are other easy way can also raise