Remove values in rows, Pandas DataFrame

Question

im trying to "remove" some rows. This is my code

popcoun = census_df.copy()
popcoun = popcoun[popcoun['SUMLEV'] == 50]
popcoun = popcoun[['STNAME','CTYNAME','CENSUS2010POP']]
popcoun = popcoun.set_index(['STNAME','CTYNAME'])
popcoun = popcoun.sort_values(by = ['STNAME','CENSUS2010POP'],ascending = False)

Visualice table

In the previous image link, you can observe that the information is sorted, so, for example, in the Wioming index, i only want the first three rows of the CENSUS2010POP (which are the highest values in that state), and also for the other states i have. Thank you, i hope somebody help me

Answer 1

Add this -

popcoun = popcoun.groupby(['STNAME']).head(3)

This should work as long as the rows are sorted for each group as you have mentioned above.

Answer 2

If you want to just select the top 3 rows of the table, you can do

df.iloc[:4]

For each state, you can iterate over df["state"].unique() and do df.loc[df.state == state][:4]

Sorry if I misunderstood. Does this help?

Answer 3

Regardless of sorting... this would work with .groupby and .nlargest

popcoun = popcoun.groupby(['STNAME']).apply(lambda x: x.nlargest(3, 'CENSUS2010POP'))[['CITYNAME', 'CENSUS2010POP']]