How to utilise Pandas dataframe and series for data wrangling?

0 0 4 minutes read

In this article, we are going to see how to utilize Pandas DataFrame and series for data wrangling.

The process of cleansing and integrating dirty and complicated data sets for easy access and analysis is known as data wrangling. As the amount of data raises continually and expands, it is becoming more important to organize vast amounts of data for analysis. Data wrangling comprises activities such as data sorting, data filtering, data reduction, data access, and data processing. Data wrangling is one of the most important tasks in data science and data analysis. Let’s see how to utilize Pandas DataFrame and series for data wrangling.

Used CSV: https://drive.google.com/file/d/1QK_0q4TjguwKjRMePk5NoQ29JcSfnGAe/view

Utilize series for data wrangling

Creating Series

pd.Series() method is used to create a pandas Series. In this, a list is given as an argument and we use the index parameter to set the index of the Series. The index helps us to retrieve data based on conditions.

Python3

# importing packages
import pandas as pd
 
# creating a series
population_data = pd.Series([1440297825, 1382345085,
                             331341050,
                             274021604, 212821986],
                            index=['China', 'India', 
                                   'United States',
                                   'Indonesia', 'Brazil'])
 
print(population_data)

Output:

Filtering data- Retrieving insights based on conditions from the data

From the previous data, we retrieve data on two conditions, one is the population of India and another is countries that have a population of more than a billion.

Python3

print('population of india is : \
' + str(population_data['India']))
 
print('population greater than a billion :')
print(population_data[population_data > 1000000000])

Output:

We can also use dictionaries to create Series in python. In this, we have to pass a Dictionary as an argument in the pd.Series() method.

Python3

population_data = pd.Series({'China': 1440297825,
                             'India': 1382345085, 
                             'United States': 331341050, 
                             'Indonesia': 274021604,
                             'Brazil': 212821986})
print(population_data)

Output:

Changing indices by altering the index of series

In pd.Series the index can be manipulated or altered by specifying a new index series.

Python3

population_data.index = ['CHINA','INDIA',
                         'US','INDONESIA',
                         'BRAZIL']
print(population_data)

Output:

Utilize Pandas Dataframe for data wrangling

Creating Dataframe using CSV

In this example, we will use a CSV file to print top n (5 by default) rows of a DataFrame or series using the Pandas.head() method.

Python3

# importing packages
import pandas as pd
 
# loading csv data
population_data = pd.read_csv('employees.csv')
 
# setting the index of the dataframe
population_data=population_data.set_index('First Name')
 
# head of the dataframe
population_data.head()

Output:

Describing DataFrame

pd.Describe() method is used to get the summary statistics of the Dataframe.

Python3

# importing packages
import pandas as pd
 
# loading csv data
population_data = pd.read_csv('employees.csv')
 
population_data.describe()

Output:

Setting and Resetting the index of the Dataframe

pd.set_index is used for setting and resetting the index of the Dataframe. Whereas, pd.reset_index() reverts the Dataframe back to the normal state. Here, the name of the column is given as an argument.

Example 1: Resetting the index of the Dataframe in Start Date columns

Python3

# importing packages
import pandas as pd
 
# creating a pandas Dataframe
population_data = pd.read_csv('employees.csv')
 
# setting the index of the dataframe
population_data = population_data.set_index('Start Date')
pd.DataFrame(population_data)

Output:

Example 2: Resetting the index of the Dataframe in First Name columns

Python3

# importing packages
import pandas as pd
 
# creating a pandas Dataframe
population_data = pd.read_csv('employees.csv')
 
 
# resetting the index of the dataframe
population_data = population_data.reset_index()
 
population_data = population_data.set_index('First Name')
 
pd.DataFrame(population_data)

Output:

Deleting a column from the DataFrames

The column ‘Salary’ is deleted from the DataFrames from our CSV file.

Python3

# importing packages
import pandas as pd
 
# loading csv data
population_data = pd.read_csv('employees.csv')
 
 
# deleting column
del population_data['Salary']
                     
pd.DataFrame(population_data.head())

Output:

Reshaping dataframe

df.Transpose() function is used to find the transpose of the given DataFrame.

Python3

# importing packages
import pandas as pd
 
# loading csv data
population_data = pd.read_csv('employees.csv')
 
# setting the index of the dataframe
population_data=population_data.set_index('First Name')
 
# displaying a transpose of the dataframe
pd.DataFrame(population_data.transpose().head())

Output:

Sorting the Dataframe

df.sort_values() function is used to sort data. In this, the column name is passed as a parameter.

Python3

# importing packages
import pandas as pd
 
# loading csv data
population_data = pd.read_csv('employees.csv')
 
# setting the index of the dataframe
population_data=population_data.set_index('First Name')
 
# sorting the Dataframe based on Density of population per km column
sorted_dataframe = population_data.sort_values('Salary', ascending=False)
pd.DataFrame(sorted_dataframe)

Output:

Dealing with missing values

Missing or null values can be checked with the Pandas df.null() method.

Python3

# importing packages
import pandas as pd
 
# loading csv data
data = pd.read_csv('employees.csv')
 
# checking for null values
data.isnull().sum()

Output:

Dropping Rows

We can filter rows that have null values by using df.dropna() method.

Python3

# importing packages
import pandas as pd
 
# loading csv data
data = pd.read_csv('employees.csv')
 
# dropping NA values
data = data.dropna(axis=0, how='any')
 
# checking for null values
data.isnull().sum()

Output:

Grouping Data

In Data Analysis, the Grouping of data sets is a common requirement when the outcome must be expressed in terms of many groups. Panadas provides us with a built-in mechanism for grouping data into several categories. The pandas‘ df.groupby() technique is used for grouping data.

In the below code, We will create a DataFrames of students and their grades. In this groupby() method is used to group students according to their grades with their names.

Python3

# importing pandas as pd
import pandas as pd
 
# Creating the dataframe
df = pd.read_csv("employees.csv")
 
# First grouping based on "Team"
# Within each team we are grouping based on "Position"
data = df.groupby(['First Name', 'Gender'])
 
# Print the first value in each group
data.first()

Output:

Merging Dataframe

Pandas df.merge() method is used to merge two DataFrames. There are different ways of merging DataFrames like, outer join, inner join, left join, right join, etc.

Python3

import pandas as pd
 
# reading two csv files
data = pd.read_csv('employees.csv')
 
# creating two dataframe 
head_data = data.head()
tail_data = data.tail()
 
# get top 5 rows
print("Head Data :")
display(head_data)
 
# get last 5 rows
print("Tail Data :")
display(tail_data)
 
# merge dataframe
merge_data = pd.merge(head_data, tail_data, how='outer' )
 
print("After merging: ")
display(merge_data)

Output:

Concatenating Data

The Concat function is used to conduct concatenation operations along an axis. Let’s create two DataFrames and concatenate them.

Python3

import pandas as pd
 
# reading two csv files
data1 = pd.read_csv('employees.csv')
data2 = pd.read_csv('borrower.csv')
 
# concatenating the dataframes
pd.DataFrame(pd.concat([data1,data2]))