How does pandas series argsort work?

The argsort() is one of the methods of the pandas series constructor, it works similar to the NumPy.ndarray.argsort(). In pandas series, the argmax() method will return a series object with the indices that would sort the original series values.

It returns a series with values replaced by sorted order of indices. And it won’t change the original series position index labels, it only replaces the values by order of sorted values positions. The argsort method tells you where that element comes from the original series object.

Example 1

import pandas as pd

# creating series
s = pd.Series({'A':123,'B':458,"C":556, "D": 238})

print(s)

# apply argsort()
print("Output argsort:",s.argsort())

Explanation

In this following example, we have created a series using a python dictionary and the index will be labeled by keys of the dictionary. Then we applied the argsort() method over the series data.

Output

A 123
B 458
C 556
D 238
dtype: int64

Output argsort:
A 0
B 3
C 1
D 2
dtype: int64

In the above output block, we can see the output series of the argsort method, and you can observe the index and values of that series object if you see the element at label “B” is coming from the “3rd” index of the original series object. In the same way, the element at label “C” is coming from the 1st index of the original series object, which is the second smallest number from the original series element.

Example 2

import pandas as pd

# creating series
series = pd.Series([3,5,2,7])
print(series)

# apply argsort()
print("Output argsort:", series.argsort())

Explanation

Let’s take another example of a pandas series object to apply the argsort method. Initially, we have created a pandas series object with a list of integer values and then applied the argsort method on that data.

Output

0 3
1 5
2 2
3 7
dtype: int64

Output argsort:
0 2
1 0
2 1
3 3
dtype: int64

The element at index position 0 of the output series object is coming from the 2nd index position of the original series and it is the smallest number. In the same way, the element at index position 1 of the output series object is coming from the 0th index position of the original series and it is the second smallest number from the original series values.