Sorting

pandas includes a few different built-in sorting operations. We can sort by an index for either axis of our DataFrame (i.e. we can sort based on row index labels or by column name).

The default for .sort_values is to sort in ascending order. We can use ascending=False to sort in descending order.

df.sort_values('AGEP') # ascending sort for AGEP column values

df.sort_values('MAR', ascending=False) # descending sort for MAR column vlaues

NOTE: When sorting, we are returning a sorted DataFrame. We ARE NOT updating the DataFrame in place. We have a couple of options to sort in place.

dfSort = df.sort_values('AGEP') # create new dataframe from sort
dfSort # show output

df.sort_values('MAR', inplace=True) # store sort output in-place
df # show output

We can also sort by multiple fields.

df.sort_values(['MAR', 'AGEP'], ascending=False) # descending sort for MAR and PWGTP columns

When sorting by fields with string data, a-z is considered ascending and z-a would be descending.

Additional Resources

For more on sorting operations in pandas, check out the package’s “Sorting” documentation.