python - Multiple conditions in selecting pandas dataframe -

i have 2 similar dataframes. df1 , df2 same structure. , have same data in 1 column(policyid) , other columns may have different data , 1 column contains timestamp based on need segregate them. df1 -

policyid,statecode,county,timestamp 114455,fl,clay county,2015-12-23 

df2-

policyid,statecode,county,timestamp 114455,fl,clay county,2016-12-34 

i need create new dataframe df3 contains data df1 if timestamp of df1 higher , df2 if timestamp of df2 higher.

i think need concat sort_values, last need groupby first or first:

print (df1)             0    1             2           3 0  114453   fl   clay county  2016-12-23 1  114455  fl1  clay county1  2016-12-11 2  114457  fl1  clay county1  2016-12-11  print (df2)         0    1             2           3 0  114453  fl1  clay county1  2016-12-15 1  114455  fl1  clay county1  2016-12-15 2  114457  fl1  clay county1  2016-12-15  df1.iloc[:,3] = pd.to_datetime(df1.iloc[:,3]) df2.iloc[:,3] = pd.to_datetime(df2.iloc[:,3])  df3 = pd.concat([df1,df2]) df3 = df3.sort_values(df3.columns[3]).groupby(df3.columns[0]).first() print (df3)           1             2          3 0                                    114453  fl1  clay county1 2016-12-15 114455  fl1  clay county1 2016-12-11 114457  fl1  clay county1 2016-12-11 

---

df3 = pd.concat([df1,df2]) df4 = df3.sort_values(df3.columns[3]).groupby(df3.columns[0]).last() print (df4)           1             2          3 0                                    114453   fl   clay county 2016-12-23 114455  fl1  clay county1 2016-12-15 114457  fl1  clay county1 2016-12-15