Take a look at the DataFrame output from step 7. Did you notice that the months are in alphabetical and not chronological order? Pandas unfortunately, in this case at least, orders the months for us alphabetically. We can solve this issue by changing the data type of Month to a categorical variable. Categorical variables map all the values of each column to an integer. We can choose this mapping to be the normal chronological order for the months. Pandas uses this underlying integer mapping during the pivot method to order the months chronologically:

>>> week4a = week4.copy()
>>> month_chron = week4a['Month'].unique() # or use drop_duplicates
>>> month_chron
array(['Jan', 'Feb', 'Mar', 'Apr'], dtype=object)

>>> week4a['Month'] = pd.Categorical(week4a['Month'],
                                     categories=month_chron,
                                     ordered=True)
>>> week4a.pivot(index='Month', columns='Name',
                 values='Perc Weight Loss')

To convert the Month column, use the Categorical constructor. Pass it the original column as a Series and a unique sequence of all the categories in the desired order to the categories parameter. As the Month column is already in chronological order, we can simply use the unique method, which preserves order to get the array that we desire. In general, to sort columns of object data type by something other than alphabetical, convert them to categorical.