This project demonstrates an end-to-end data cleaning workflow using Python. The objective was to convert a raw retail product sales dataset containing over 4,000 records into a clean, structured dataset suitable for business analysis and dashboard development.
The project follows common data preprocessing techniques used by Data Analysts before performing exploratory analysis or creating visualizations.
- Python
- Pandas
- NumPy
- Regular Expressions (Regex)
- Jupyter Notebook
- Loaded the dataset into Pandas DataFrame
- Examined data types
- Checked dataset dimensions
- Reviewed missing values
- Identified null values
- Replaced missing numerical values using Mean and Median
- Filled missing categorical values using Mode where appropriate
- Identified duplicate records
- Removed duplicate rows
- Corrected invalid numeric values
- Standardized categorical entries
- Removed unnecessary whitespace
- Fixed inconsistent text formatting
- Validated customer email addresses using Python Regular Expressions
- Corrected invalid email formats
- Removed unwanted characters
- Converted phone numbers into a consistent format
- Converted object columns into appropriate numeric and datetime formats
- Exported the cleaned dataset for further analysis and visualization
- Data Cleaning
- Data Validation
- Data Wrangling
- Data Transformation
- Pandas
- NumPy
- Regex
- Python
- Exploratory Data Analysis
- Business Data Preparation
- Learning Outcome
Raw Dataset │ ▼ Data Inspection │ ▼ Data Cleaning │ ▼ Data Validation │ ▼ Clean Dataset │ ▼ Business Analysis
Through this project, I strengthened my practical knowledge of preparing real-world datasets for analytics by applying structured data cleaning techniques using Python. The cleaned dataset can now be used for reporting, dashboard creation, and further business analysis.