Sorting Your DataFrame on a Single Column

00:00 Sorting Your DataFrame on a Single Column. To sort the DataFrame based on the values in a single column, you’ll use .sort_values(). By default, this will return a new DataFrame sorted in ascending order.

00:14 It doesn’t modify the original DataFrame. To use .sort_values(), you pass a single argument to the method containing the name of the column you want to sort by.

00:25 In this example, you sort the DataFrame by the city08 column, which represents city miles per gallon for fuel-only cars.

00:39 This sorts your DataFrame using the column values from city08, showing the vehicles with the lowest miles per gallon first. By default, .sort_values() sorts your data in ascending order.

00:53 Although you didn’t specify a name for the argument you passed to .sort_values(), you actually used the by parameter, which you’ll see in the next example.

01:03 Another parameter of .sort_values() is ascending, which by default is set to True. If you want the DataFrame sorted in descending order, then you can pass False to this parameter, as seen on-screen.

01:24 By passing False to ascending, you reverse the sort order. Now your DataFrame is sorted in descending order by the average miles per gallon measured in city conditions.

01:34 The vehicles with the highest miles-per-gallon values are in the first rows.

01:40 It’s good to note that pandas allows you to choose different sorting algorithms to use with both .sort_values() and .sort_index().

01:47 The available algorithms are quicksort, mergesort, and heapsort. For more information on these different sorting algorithms, check out this Real Python course.

02:01 The algorithm used by default when sorting on a single column is quicksort. To change this to a stable sorting algorithm, use mergesort.

02:11 You can do that with the kind parameter in .sort_values() or .sort_index(), as seen on-screen.

02:26 Using kind, you set the sorting algorithm to mergesort. The previous output used the default quicksort algorithm.

02:35 Looking at the highlighted indices, you can see the rows are in a different order. This is because quicksort is not a stable sorting algorithm, but mergesort is.

02:45 Note that in pandas, kind is ignored when you sort on more than one column or label. When you’re sorting multiple records that have the same key, a stable sorting algorithm will maintain the original order of those records after sorting.

03:00 For that reason, using a stable sorting algorithm is necessary if you plan to perform multiple sorts. Now that you’re familiar with sorting a DataFrame on a single column, you’re ready to see how to sort one on multiple columns.

03:15 And that’s what will be covered in the next section of the course.