Open In App

Data Wrangling in Python

Improve
Improve
Like Article
Like
Save
Share
Report

Data Wrangling is the process of gathering, collecting, and transforming Raw data into another format for better understanding, decision-making, accessing, and analysis in less time. Data Wrangling is also known as Data Munging.

Python Data Wrangling

Python Data Wrangling

Importance Of Data Wrangling

Data Wrangling is a very important step in a Data science project. The below example will explain its importance: 

Books selling Website want to show top-selling books of different domains, according to user preference. For example, if a new user searches for motivational books, then they want to show those motivational books which sell the most or have a high rating, etc. 

But on their website, there are plenty of raw data from different users. Here the concept of Data Munging or Data Wrangling is used. As we know Data wrangling is not by the System itself. This process is done by Data Scientists. So, the data Scientist will wrangle data in such a way that they will sort the motivational books that are sold more or have high ratings or user buy this book with these package of Books, etc. On the basis of that, the new user will make a choice. This will explain the importance of Data wrangling.

Data Wrangling in Python

Data Wrangling is a crucial topic for Data Science and Data Analysis. Pandas Framework of Python is used for Data Wrangling. Pandas is an open-source library in Python specifically developed for Data Analysis and Data Science. It is used for processes like data sorting or filtration, Data grouping, etc.

Data wrangling in Python deals with the below functionalities:

  1. Data exploration: In this process, the data is studied, analyzed, and understood by visualizing representations of data.
  2. Dealing with missing values: Most of the datasets having a vast amount of data contain missing values of NaN, they are needed to be taken care of by replacing them with mean, mode, the most frequent value of the column, or simply by dropping the row having a NaN value.
  3. Reshaping data: In this process, data is manipulated according to the requirements, where new data can be added or pre-existing data can be modified.
  4. Filtering data: Some times datasets are comprised of unwanted rows or columns which are required to be removed or filtered
  5. Other: After dealing with the raw dataset with the above functionalities we get an efficient dataset as per our requirements and then it can be used for a required purpose like data analyzing, machine learning, data visualization, model training etc.

Below are  examples of Data Wrangling that implements the above functionalities on a raw dataset:

Data exploration in Python

 Here in Data exploration, we load the data into a dataframe, and then we visualize the data in a tabular format.

Python3




# Import pandas package
import pandas as pd
 
# Assign data
data = {'Name': ['Jai', 'Princi', 'Gaurav',
                 'Anuj', 'Ravi', 'Natasha', 'Riya'],
        'Age': [17, 17, 18, 17, 18, 17, 17],
        'Gender': ['M', 'F', 'M', 'M', 'M', 'F', 'F'],
        'Marks': [90, 76, 'NaN', 74, 65, 'NaN', 71]}
 
# Convert into DataFrame
df = pd.DataFrame(data)
 
# Display data
df


Output:

displaying datset into tabular format

defining the dataframe and displaying in tabular format

Dealing with missing values in Python

As we can see from the previous output, there are NaN values present in the MARKS column which is a missing value in the dataframe that is going to be taken care of in data wrangling by replacing them with the column mean.

Python3




# Compute average
c = avg = 0
for ele in df['Marks']:
    if str(ele).isnumeric():
        c += 1
        avg += ele
avg /= c
 
# Replace missing values
df = df.replace(to_replace="NaN",
                value=avg)
 
# Display data
df


Output:

replacing Nan values with average

replacing Nan values with average

Data Replacing in Data Wrangling

in the GENDER column, we can replace the Gender column data by categorizing them into different numbers.

Python3




# Categorize gender
df['Gender'] = df['Gender'].map({'M': 0,
                                 'F': 1, }).astype(float)
 
# Display data
df


Output:

Data encoding for gender variable in data wrangling

Data encoding for gender variable in data wrangling 

Filtering data in Data Wrangling

suppose there is a requirement for the details regarding name, gender, and marks of the top-scoring students. Here we need to remove some using the pandas slicing method in data wrangling from unwanted data.

Python3




# Filter top scoring students
df = df[df['Marks'] >= 75].copy()
 
# Remove age column from filtered DataFrame
df.drop('Age', axis=1, inplace=True)
 
# Display data
df


Output:

Dropping column and filtering rows

Dropping column and filtering rows

Hence, we have finally obtained an efficient dataset that can be further used for various purposes. 

Now that we have seen the basics of data wrangling using Python and pandas. Below we will discuss various operations using which we can perform data wrangling:

Data Wrangling  Using Merge Operation

Merge operation is used to merge two raw data into the desired format.

Syntax: pd.merge( data_frame1,data_frame2, on=”field “) 

Here the field is the name of the column which is similar in both data-frame.

For example: Suppose that a Teacher has two types of Data, the first type of Data consists of Details of Students and the Second type of Data Consist of Pending Fees Status which is taken from the Account Office. So The Teacher will use the merge operation here in order to merge the data and provide it meaning. So that teacher will analyze it easily and it also reduces the time and effort of the Teacher from Manual Merging.

Creating First Dataframe to Perform Merge Operation using Data Wrangling:

Python3




# import module
import pandas as pd   
 
# creating DataFrame for Student Details
details = pd.DataFrame({
    'ID': [101, 102, 103, 104, 105, 106,
           107, 108, 109, 110],
    'NAME': ['Jagroop', 'Praveen', 'Harjot',
             'Pooja', 'Rahul', 'Nikita',
             'Saurabh', 'Ayush', 'Dolly', "Mohit"],
    'BRANCH': ['CSE', 'CSE', 'CSE', 'CSE', 'CSE',
               'CSE', 'CSE', 'CSE', 'CSE', 'CSE']})
 
# printing details
print(details)


Output:

displaying datset into tabular format

printing dataframe 

Creating Second Dataframe to Perform Merge operation using Data Wrangling:

Python3




# Import module
import pandas as pd
 
# Creating Dataframe for Fees_Status
fees_status = pd.DataFrame(
    {'ID': [101, 102, 103, 104, 105,
            106, 107, 108, 109, 110],
     'PENDING': ['5000', '250', 'NIL',
                 '9000', '15000', 'NIL',
                 '4500', '1800', '250', 'NIL']})
 
# Printing fees_status
print(fees_status)


Output:

cretaing dataframe

Define second dataframe

Data Wrangling Using Merge Operation:

Python3




# Import module
import pandas as pd
 
# Creating Dataframe
details = pd.DataFrame({
    'ID': [101, 102, 103, 104, 105,
           106, 107, 108, 109, 110],
    'NAME': ['Jagroop', 'Praveen', 'Harjot',
             'Pooja', 'Rahul', 'Nikita',
             'Saurabh', 'Ayush', 'Dolly', "Mohit"],
    'BRANCH': ['CSE', 'CSE', 'CSE', 'CSE', 'CSE',
               'CSE', 'CSE', 'CSE', 'CSE', 'CSE']})
 
# Creating Dataframe
fees_status = pd.DataFrame(
    {'ID': [101, 102, 103, 104, 105,
            106, 107, 108, 109, 110],
     'PENDING': ['5000', '250', 'NIL',
                 '9000', '15000', 'NIL',
                 '4500', '1800', '250', 'NIL']})
 
# Merging Dataframe
print(pd.merge(details, fees_status, on='ID'))


Output:

data wrangling using merge opeartions

Merging two dataframes 

Data Wrangling Using Grouping Method 

The grouping method in Data wrangling is used to provide results in terms of various groups taken out from Large Data. This method of pandas is used to group the outset of data from the large data set.

Example: There is a Car Selling company and this company have different Brands of various Car Manufacturing Company like Maruti, Toyota, Mahindra, Ford, etc., and have data on where different cars are sold in different years. So the Company wants to wrangle only that data where cars are sold during the year 2010. For this problem, we use another data Wrangling technique which is a pandas groupby() method.

Creating dataframe to use Grouping methods[Car selling datasets]:

Python3




# Import module
import pandas as pd
 
# Creating Data
car_selling_data = {'Brand': ['Maruti', 'Maruti', 'Maruti',
                              'Maruti', 'Hyundai', 'Hyundai',
                              'Toyota', 'Mahindra', 'Mahindra',
                              'Ford', 'Toyota', 'Ford'],
                    'Year':  [2010, 2011, 2009, 2013,
                              2010, 2011, 2011, 2010,
                              2013, 2010, 2010, 2011],
                    'Sold': [6, 7, 9, 8, 3, 5,
                             2, 8, 7, 2, 4, 2]}
 
# Creating Dataframe of car_selling_data
df = pd.DataFrame(car_selling_data)
 
# printing Dataframe
print(df)


Output:

Initiating dataframe

Creating new dataframe 

Creating Dataframe to use Grouping methods[DATA OF THE YEAR 2010]:

Python3




# Import module
import pandas as pd
 
# Creating Data
car_selling_data = {'Brand': ['Maruti', 'Maruti', 'Maruti',
                              'Maruti', 'Hyundai', 'Hyundai',
                              'Toyota', 'Mahindra', 'Mahindra',
                              'Ford', 'Toyota', 'Ford'],
                    'Year':  [2010, 2011, 2009, 2013,
                              2010, 2011, 2011, 2010,
                              2013, 2010, 2010, 2011],
                    'Sold': [6, 7, 9, 8, 3, 5,
                             2, 8, 7, 2, 4, 2]}
 
# Creating Dataframe for Provided Data
df = pd.DataFrame(car_selling_data)
 
# Group the data when year = 2010
grouped = df.groupby('Year')
print(grouped.get_group(2010))


Output:

 dataframe wrangling using groupby methods

Using groupby method on dataframe 

Data Wrangling  by Removing Duplication

Pandas duplicates() method helps us to remove duplicate values from Large Data. An important part of Data Wrangling is removing Duplicate values from the large data set.

Syntax: DataFrame.duplicated(subset=None, keep=’first’)

Here subset is the column value where we want to remove the Duplicate value.

In keeping, we have 3 options :

  • if keep =’first’ then the first value is marked as the original rest of all values if occur will be removed as it is considered duplicate.
  • if keep=’last’ then the last value is marked as the original rest the above same values will be removed as it is considered duplicate values.
  • if keep =’false’ all the values which occur more than once will be removed as all are considered duplicate values.

For example, A University will organize the event. In order to participate Students have to fill in their details in the online form so that they will contact them. It may be possible that a student will fill out the form multiple times. It may cause difficulty for the event organizer if a single student will fill in multiple entries. The Data that the organizers will get can be Easily Wrangles by removing duplicate values.

Creating a Student Dataset who want to participate in the event:

Python3




# Import module
import pandas as pd
 
# Initializing Data
student_data = {'Name': ['Amit', 'Praveen', 'Jagroop',
                         'Rahul', 'Vishal', 'Suraj',
                         'Rishab', 'Satyapal', 'Amit',
                         'Rahul', 'Praveen', 'Amit'],
 
                'Roll_no': [23, 54, 29, 36, 59, 38,
                            12, 45, 34, 36, 54, 23],
 
                'Email': ['xxxx@gmail.com', 'xxxxxx@gmail.com',
                          'xxxxxx@gmail.com', 'xx@gmail.com',
                          'xxxx@gmail.com', 'xxxxx@gmail.com',
                          'xxxxx@gmail.com', 'xxxxx@gmail.com',
                          'xxxxx@gmail.com', 'xxxxxx@gmail.com',
                          'xxxxxxxxxx@gmail.com', 'xxxxxxxxxx@gmail.com']}
 
# Creating Dataframe of Data
df = pd.DataFrame(student_data)
 
# Printing Dataframe
print(df)


Output:

Initiating dataframe

Student Dataset who want to participate in the event

Removing Duplicate data from the Dataset using Data wrangling:

Python3




# import module
import pandas as pd
 
# initializing Data
student_data = {'Name': ['Amit', 'Praveen', 'Jagroop',
                         'Rahul', 'Vishal', 'Suraj',
                         'Rishab', 'Satyapal', 'Amit',
                         'Rahul', 'Praveen', 'Amit'],
 
                'Roll_no': [23, 54, 29, 36, 59, 38,
                            12, 45, 34, 36, 54, 23],
                'Email': ['xxxx@gmail.com', 'xxxxxx@gmail.com',
                          'xxxxxx@gmail.com', 'xx@gmail.com',
                          'xxxx@gmail.com', 'xxxxx@gmail.com',
                          'xxxxx@gmail.com', 'xxxxx@gmail.com',
                          'xxxxx@gmail.com', 'xxxxxx@gmail.com',
                          'xxxxxxxxxx@gmail.com', 'xxxxxxxxxx@gmail.com']}
 
# creating dataframe
df = pd.DataFrame(student_data)
 
# Here df.duplicated() list duplicate  Entries in ROllno.
# So that ~(NOT) is placed in order to get non duplicate values.
non_duplicate = df[~df.duplicated('Roll_no')]
 
# printing non-duplicate values
print(non_duplicate)


Output:D

Removing duplicate entries from dataframe

Remove – Duplicate data from Dataset using Data wrangling

Creating New Datasets Using the Concatenation of Two Datasets In Data Wrangling. 

We can join two dataframe in several ways. For our example in Concanating Two datasets, we use pd.concat() function.  

Creating Two Dataframe For Concatenation.

Python3




# importing pandas module
import pandas as pd
   
# Define a dictionary containing employee data
data1 = {'Name':['Jai', 'Princi', 'Gaurav', 'Anuj'],
        'Age':[27, 24, 22, 32],
        'Address':['Nagpur', 'Kanpur', 'Allahabad', 'Kannuaj'],
        'Qualification':['Msc', 'MA', 'MCA', 'Phd'],
        'Mobile No': [97, 91, 58, 76]}
     
# Define a dictionary containing employee data
data2 = {'Name':['Gaurav', 'Anuj', 'Dhiraj', 'Hitesh'],
        'Age':[22, 32, 12, 52],
        'Address':['Allahabad', 'Kannuaj', 'Allahabad', 'Kannuaj'],
        'Qualification':['MCA', 'Phd', 'Bcom', 'B.hons'],
        'Salary':[1000, 2000, 3000, 4000]}
   
# Convert the dictionary into DataFrame 
df = pd.DataFrame(data1,index=[0, 1, 2, 3])
   
# Convert the dictionary into DataFrame 
df1 = pd.DataFrame(data2, index=[2, 3, 6, 7])


We will join these two dataframe along axis 0.

Python3




res = pd.concat([df, df1])


output:

    Name    Age    Address    Qualification    Mobile No    Salary
0    Jai        27    Nagpur          Msc            97.0        NaN
1    Princi    24    Kanpur        MA            91.0        NaN
2    Gaurav    22    Allahabad    MCA            58.0        NaN
3    Anuj    32    Kannuaj        Phd            76.0        NaN
4    Gaurav    22    Allahabad    MCA            NaN            1000.0
5    Anuj    32    Kannuaj        Phd            NaN            2000.0
6    Dhiraj    12    Allahabad    Bcom        NaN            3000.0
7    Hitesh    52    Kannuaj        B.hons        NaN            4000.0

Note:- We can see that data1 does not have a salary column so all four rows of new dataframe res are Nan values.



Last Updated : 26 Apr, 2023
Like Article
Save Article
Previous
Next
Share your thoughts in the comments
Similar Reads