Pandas dataframe sort_values set default ascending=False

Question

Could be possible to define a property to always set sort_values(ascending=False)? Using it quite often in descending order then would like to set the default behavior.

Answer 1

You can subclass the standard pd.DataFrame and redefine just the .sort_values method:

class MyDataFrame(pd.DataFrame):
    def sort_values(self,by,axis=0,ascending=False,inplace=False, 
                    kind='quicksort',na_position='last'):
        return super().sort_values(by,axis,ascending,inplace,kind)

foo = MyDataFrame({'z': [1,2,3,4]})
foo.sort_values('z')
#   z
#3  4
#2  3
#1  2
#0  1
foo.sort_values('z',ascending=True)
#   z
#0  1
#1  2
#2  3
#3  4

Answer 2

You could make your own wrapper for this

def sort_values(df, *args, **kwargs):
    if len(args) < 2 and "ascending" not in kwargs:
        kwargs["ascending"] = False
    return df.sort_values(*args, **kwargs)

But you will have to use it as a function. (You can pass other parameters after 1st agument (dataframe).

print(sort_values(my_df, data))

Answer 3

You cannot modify the pandas function. but a good alternative would be to wrap it in your own function.

def sort_values(df, sort_by):
    return df.sort_values(ascending=False, by=sort_by)
sort_values(data, ['col1']

Answer 4

Another option would be to declare the settings you often use in the beginning of your code and pass them as kwargs.

Personally I would, however, write it out every time.

import pandas as pd
p = {"ascending":False, "inplace":True}


df = pd.DataFrame({
    'col1': [1,6,2,5,9,3]
})

df.sort_values(by='col1', **p)

print(df)

Returns:

   col1
4     9
1     6
3     5
5     3
2     2
0     1