In this article, we will explore the fundamentals and most used methods of the Pandas Series.

What is Pandas Series?

Pandas series is a one-dimensional labeled array that is capable of storing data of various types, such as integers, strings, Python objects, etc. The Series object has a single axis (the index) that associates a name to each series element through which the element is accessed.

Data Structures in Pandas

Pandas have three different data structures:

In this article, we will focus on the Series structures, which are one of the most widely used.

There are several ways to create a Pandas Series:

• Creating an Empty Series
• Creating a Series from list
• Creating a Series from dict
• Creating a Series from array
• Creating a Series from scalar value

Create an Empty Series

Let’s create our first Pandas Series; first, we import the pandas module, and then we can use the pd.Series() function as follows:

import pandas as pd
ser = pd.Series()
print(ser)

Output:

Series([], dtype: float64)

The output shows that the series data type is float by default.

Create Pandas Series From List

First, we create a list, and then we can create a series from that list using pd.Series().

import pandas as pd
list = ['O','r','a','a','s','k']
ser = pd.Series(list)
print(ser)

Output:

0    O

1    r

2    a

3    a

4    s

5    k

dtype: object

Create a Series from dict

Pandas Series can be created from dict indicating index or without providing an index. We will explain both methods.

Create a Series from dict Without Specifying an Index

We can pass a dict as input; Although we don’t specify the index, the dictionary keys are taken in a sorted order to construct the index.

data = {'zero' : 0., 'one' : 1., 'two' : 2.}
s = pd.Series(data)
print(s)

Output:

zero    0.0

one     1.0

two     2.0

dtype: float64

Create a Series from dict Indicating an Index

If we pass an index, the values in data corresponding to the index’s labels will be pulled out.

data = {'zero' : 0., 'one' : 1., 'two' : 2.}
s = pd.Series(data,index=['one','two','n','zero'])
print(s)

Output:

one     1.0

two     2.0

n       NaN

zero    0.0

dtype: float64

Note that the index order is persisted, and the missing element is filled with NaN (Not a Number).

Create a Series from Array

Pandas Series can be created from an array indicating an index or without providing an index. We will explain both methods.

Create a Series from Array without Specifying Index

First, import a numpy module, then use the array() function as follows:

import numpy as np

array1 = np.array(['O','r','a','a','s','k'])
 
ser = pd.Series(array1)
print(ser)

Output:

0    O

1    r

2    a

3    a

4    s

5    k

dtype: object

In this case, if the index is not passed, then by default index is range(n) where n is array length: the indexes ranging from 0 to len(array1)-1 or 5.

Create a Series from Array Indicating Index

Now, let’s see how to create a Pandas Series providing a custom Index. To do so, we will use index param, which takes a list of index values. We must ensure the number of elements in the index list matches the array size.

array1 = np.array(['O','r','a','a','s','k'])
ser = pd.Series(array1, index=[100,101,102,103,104,105])
print(ser)

Output:

100    O

101    r

102    a

103    a

104    s

105    k

dtype: object

As we passed the index values, we can now see the customized indexed values in the output.

Create a Series from Scalar Value

To create a series from scalar value, we must indicate an index. The scalar value will be repeated to match the length of the index.

ser = pd.Series(23, index=[0, 1, 2, 3])
print(ser)

Output:

0    23

1    23

2    23

3    23

dtype: int64

The output shows that the scalar value (23) is repeated 4 times.

Binary Operation on Series

Pandas Series allows us to perform binary operations such as sum, mean, product, etc. To do so, we have to use the following functions: sum(), mean() and prod().

We will use the following series to perform the binary operations mentioned before:

s = pd.Series([1, 2, 3, 4], index=['one', 'two', 'three', 'four'])
print(s)

Output:

one      1

two      2

three    3

four     4

dtype: int64

Sum

This method is used to get the sum of the values for the requested axis by using sum().

s.sum()

Output:

10

Mean

Pandas Series.mean() method is used to get the mean of the values over the requested axis.

s.mean()

Output:

2.5

Prod

The prod() method returns the product of the values for the requested axis.

s.prod()

Output:

24

Conversion Operation on Series

On Pandas Series, we can perform conversion operations like changing the datatype of a series, changing a series to a list, etc. In order to perform conversion operation we must use the following methods .astype(), .tolist(), etc.

We will use the following series to perform the conversion operations mentioned before:

ser = pd.Series([1, 2, 3], dtype='int64')
print(ser)

Output:

0    1

1    2

2    3

dtype: int64

Converting Data type of Series

Pandas astype() is used to change the data type of a Series. In this example, we will convert the original Series data type to Float by using .astype() and passing ‘float’ as an argument.

ser_float = ser.astype('float')
print(ser_float)

Output:

0    1.0

1    2.0

2    3.0

dtype: float64

As we can see in the output, we have created a new Series ‘ser_float’ whose data type is ‘float’.

We can also convert the original Series data type to String by using .astype() and passing ‘str as an argument.

ser_str = ser.astype('str')
print(ser_str)

Output:

0    1

1    2

2    3

dtype: object

As we can see in the output, we have created a new Series ‘ser_str’ whose data type is ‘string’.

Converting Data type

To convert a pandas Series to a list, we can use the tolist() method on the series we want to convert.

ser_list = ser.tolist()
print(ser_list)

Output:

[1, 2, 3]

As we can see in the output, we have created a List ‘ser_list’ with the values [1, 2, 3].

Indexing Series

Pandas support three types of multi-axis indexing:

• Using []
• .loc
• .iloc

Having the following Series:

list = ['O','r','a','a','s','k']
list_index = ['O-1','r-2','a-3','a-4','s-5','k-6']
ser = pd.Series(list, index = list_index)
print(ser)

Output:

O-1    O

r-2    r

a-3    a

a-4    a

s-5    s

k-6    k

dtype: object

Using []

This is the most basic indexing; we indicate the index of the data we want to retrieve

ser[5]

Output:

‘k’

Using .loc

Uses row and column labels.

ser.loc['k-6']

Output:

‘k’

Using .iloc

Uses row and column indexes.

ser.iloc[5]

Output:

‘k’

Conclusion

In this article, we have learned the main features of the Pandas library and how to create Pandas Series from lists, arrays, dicts, and scalar values in Python. We have also covered binary operations on series such as sum, mean, and product, and we explored the conversion operation on series using .tolist() and astype(). Finally, we explored how to handle the data from these data structures by selecting rows and columns.