agg() function in Python

Because of the fantastic ecosystem for the data-centric Python packages, Python is considered as one of the great programming languages for doing data analysis. Panda is one of such packages that is provided to us in Python, and it makes it very easy to import and analyze the data.

In this tutorial, we are going to discuss the agg() function given to us in the panda series, and we will use it with series data given to us.

Introduction: Pandas agg() function

We use the agg() function of Pandas to pass a single function or list of functions that are to be applied on a given data series or sometimes even separately to each element of the data series. In the case where we pass a list of functions inside the agg() function, multiple results will be returned by it.

Syntax

In this section, we are going to look at the syntax and parameters that we have to use inside the agg() method and return type of the function.

The seriesGiven is the data series given to us in the program.

Parameters:

We have to provide the following parameters inside the agg() method.

• Function_name: We have to provide a function, list of functions, or string that contains the name of functions to be called on the data series as a parameter inside the agg() method.
• axis: Axis works like defining the index for rows of data series. We can either give axis equals to 0 or provide 'index' for performing a row-wise operation on data series. Also, we can give either 1 or 'column' in the axis parameter to perform a column-wise operation on the data series.

Return type

The return type of agg() method is not fixed, and it always depends upon the return type of the function that we have passed as a parameter inside the agg() method.

Working with agg() function

Till now, we have learned about the introduction and syntax for using agg() function provided to us in Pandas. To learn about and understand the working of the agg() method, we will use this function in below examples.

Passing a single function inside agg() method

In this example, we will create a random array by numpy module, and then we will make it a data series using the Pandas function. After that, we will use the agg() function and pass a lambda function, as an argument inside it, and thus it will add 3 to each value given in the series. As we are applying the function to a series, the return type we get through the agg() function is also series. Now, let's understand this implementation through the following example.

Example 1: Look at the following Python program:

Output:

Data series of elements before operation: 
 0    -0.510111
1    -0.732670
2    -0.451550
3    -0.435085
4     0.082848
5    -1.051242
6     0.203565
7    -1.014079
8    -0.232350
9    -0.325640
10    0.528320
11   -1.472293
12   -0.639487
13   -2.490666
14   -0.242837
15    0.854955
16    1.076247
17    1.491347
18   -1.767788
19   -0.205003
dtype: float64 

 Data series of elements after operation: 
 0     2.489889
1     2.267330
2     2.548450
3     2.564915
4     3.082848
5     1.948758
6     3.203565
7     1.985921
8     2.767650
9     2.674360
10    3.528320
11    1.527707
12    2.360513
13    0.509334
14    2.757163
15    3.854955
16    4.076247
17    4.491347
18    1.232212
19    2.794997
dtype: float64    

Explanation:

First, we have imported pandas and numpy modules inside our program to use its functions.

Then, we have a created an array of 20 elements that we have randomly generated using the randn() function of the numpy module. After that, we made the array into the series form using the series() function of the panda module.

Then, we have used the agg() function on the series and passed the lambda function as an argument in it. We have passed an argument in the agg() method to add 3 to each value of the series. In last, we printed the data series in the output (both before the operation performed on it and after the operation performed on it).

As we can see in the output, 3 is added to each value of the series after we have performed an operation on it.

Passing a list of functions inside agg() method:

In this example, after creating the data series, we will pass a list of functions as the arguments inside the agg() function instead of passing a single function parameter in it. When we pass a list of Python's default functions as arguments in the agg() method, multiple results are returned by it into multiple variables. Let's understand the implementation of this method through the following example.

Example 2: Look at the following Python program:

Output:

 Data Series before operation: 
 0     1.324659
1    -1.632943
2    -0.451046
3    -0.119475
4    -1.476469
5     1.550481
6    -0.345283
7    -0.391220
8     1.183295
9     0.945834
10    0.426908
11   -1.373141
12   -1.360714
13    1.029160
14   -0.305868
15    0.520776
16    0.519891
17    0.581810
18   -0.200537
19    2.175055
dtype: float64
Minimum value in the data series = -1.6329428122607905
Maximum value in the data series = 2.175055294872539,      
Sorted data series after operation:
[-1.6329428122607905, -1.476468968840359, -1.3731412602339488, -1.3607141137838996, -0.45104603430414114, -0.3912204479169106, -0.34528253055365704, -0.3058683242351637, -0.20053665016862435, -0.1194753076622943, 0.4269084920204909, 0.519891496565306, 0.5207757216248261, 0.5818098237803292, 0.9458337130436504, 1.02915996695176, 1.1832945335240084, 1.324659481096391, 1.5504805147479754, 2.175055294872539]

Explanation:

After creating a data series as we did in the previous example, we have created a list where we have multiple functions' names in it. In this example, instead of giving a single function as an argument, we have passed multiple default functions inside the agg() function. After passing these functions as arguments, we have printed the data series before operation and after operation in the output.

As we look at the output, we can see that multiple results have been returned by agg() function. This is because we have passed multiple functions as parameters inside it. The max(), min() and sorted() have been returned into different variables i.e., seriesResult1, seriesResult2 and seriesResult3 respectively.