Data Cleaning in Python: From Messy Data to Clean Data

Name: Data Cleaning in Python: From Messy Data to Clean Data
Rating: 4.7 (26 reviews)

Learn how to clean messy real-world data using Python: handle NaNs, outliers, duplicates and inconsistencies

Created byAbdellatif Satir

Last updated 5/2026

English

English [Auto],

What you'll learn

Detect data quality issues using Exploratory Data Analysis (EDA)
Identify and understand missing values (NaNs) in datasets
Handle missing data using practical imputation techniques
Detect and treat outliers using statistical methods and visualization
Fix inconsistent and messy data formats (strings, categories, dates)
Clean real-world datasets using Pandas step by step
Build a structured data cleaning workflow for any project
Prepare clean datasets ready for Machine Learning models

Course content

5 sections • 28 lectures • 53m total length

1- Course Introduction0:36
Learn the definition and importance of data cleaning within the data science and AI lifecycles, and its role in data analysis as an essential starting point.
2- What is Data Cleaning?1:35
Define data cleaning as finding and fixing errors in a dataset, then practice identifying problems and applying fixes before any analysis.
3- Types of Dirty Data: Missing Values, Duplicates, Inconsistencies & Outliers1:51
4- Messy Data vs Clean Data1:09
Compare messy and clean data, highlighting missing values, outliers, and inconsistent formats like city names, then demonstrate cleaning steps to remove duplicates and standardize data.
5- How Humans Learn vs How AI Models Learn from Data1:05
Compare how humans interpret data versus how AI learns, noting AI can take values literally. Emphasize data cleaning to avoid garbage in, garbage out and bad recommendations.

1- Project Introduction: What We Will Build in This Course1:29
2- Dataset Overview: Understanding the Data We Will Clean1:17
Explore an e-commerce orders dataset with seven columns: customer ID, full name, age, city, purchase amount, order date, and premium status; prepare to clean it with Python.
3- Loading the Dataset in Python1:47
Load the ecommerce data in Python by importing Pandas, NumPy, and Matplotlib, read with pd.readcsv into df, and display first five rows to spot missing values and inconsistent city casing.
4- Dataset Structure Overview2:48
Explore dataset structure and quality through exploratory data analysis, inspecting shape, info, missing values, duplicates, and categorical data, and prepare numeric and date columns by converting purchase amount and date.
5- Detecting Missing Values (NaNs)1:34
Identify missing values with df.isInA.sum, tally counts per column, then choose imputation to replace them with a smart number, enabling clean data for analysis.
6- Statistical Summary of Numeric Features1:17
Explore how the describe function computes mean, min, and max for numeric features, highlights outliers, and previews detection and visualization methods to clean data.
7- Categorical Data Analysis1:51
8- Detecting Duplicate Records1:11
Detect duplicate records in Python data cleaning workflows using df.duplicated.sum, reveal how duplicates distort totals and mean, and prepare to remove them as the final step of EDA.

1- Introduction to Handling Missing Values1:05
Explore four methods to handle missing values in data cleaning: replace numerical gaps with mean or median, and categorical gaps with mode or unknown to avoid bias.
2- Handling Missing Values in Numerical Columns2:13
Replace missing values with imputation, using the median for numerical columns like age and purchase amount when outliers skew the data, instead of the mean.
3- Handling Missing Values in Numerical Columns 22:08
4- Handling Missing Values in Categorical Columns0:53
Learn to handle missing values in categorical columns in data cleaning in python by replacing them with the mode or unknown, where the mode is the most frequent value.
5- Handling Missing Values in Categorical Columns 22:48
Handle missing values in the categorical is_premium column by filling with unknown or the mode, apply the unknown option, then drop the remaining row and verify the updated data shape.

1- What is an Outlier?2:07
Detect and handle outliers using z-score, IQR, and visual methods like box and scatter plots, and learn why outliers distort means and affect models.
2- Detecting Outliers Using Data Visualization3:00
3- Detecting Outliers Using Z-Score Method2:43
Compute z-scores for h and purchase amount using their means and standard deviations, then flag values where |z| > 3 as outliers.
4- Detecting Outliers Using IQR Method3:15
Apply the IQR method to detect outliers by computing Q1, Q3, and the IQR, and setting bounds at Q1 minus 1.5 times IQR and Q3 plus 1.5 times IQR.
5- Handling Outliers by Removal1:52
Remove outliers with the IQR-based bounds and a normal filter, then verify with visualization that the data drops from 199 to 196 and the final outlier is removed.
6- Handling Outliers by Removal 20:36
7- Final Step: Exporting the Clean Dataset & Course Conclusion2:48

Requirements

Basic understanding of Python programming
Python installed on your computer (Anaconda recommended)
No prior experience in data cleaning is required

Description

Course Description

Data in the real world is messy.

Missing values, inconsistent formats, duplicate entries, and outliers can completely break your analysis or machine learning models. That's why data cleaning is one of the most important skills in data science.

In this course, you will learn how to clean and prepare real-world datasets step by step, using Python and practical techniques.

By the end of this course, you will be able to confidently clean any dataset and prepare it for Data Science or Machine Learning projects.

What you will learn

How to detect and analyze data quality issues using EDA
How to handle missing values in numerical and categorical data
How to clean inconsistent and messy datasets
How to detect and remove duplicate records
How to detect and handle outliers using multiple methods
How to prepare clean datasets ready for Machine Learning

Why This Course?

Most courses focus only on models... but in reality:
80% of a data scientist's work is data cleaning

This course focuses on the real skills you actually need to work with data.

You will not just learn theory — you will work on practical examples and real datasets.

Tools You'll Use

Python
Pandas
NumPy
Matplotlib

By the End of This Course

You will be able to take any messy dataset and transform it into a clean, structured dataset ready for analysis or machine learning.

Who this course is for:

Beginners in Data Science or Data Analysis
Students who want to learn how to clean real-world datasets
Aspiring Data Analysts and Machine Learning practitioners
Python developers who want to improve their Pandas skills
Anyone who struggles with messy or incomplete data

Data Cleaning in Python: From Messy Data to Clean Data

What you'll learn

Explore related topics

Course content

Introduction5 lectures • 6min

Exploratory Data Analysis (EDA) for Detecting Data Issues8 lectures • 13min

Handling Data Inconsistency3 lectures • 8min

Handling Missing Values (NaNs)5 lectures • 9min

Detecting and Handling Outliers7 lectures • 16min

Requirements

Description

Who this course is for: