The methods of Data Cleaning in Data Science


Data cleaning is a critical step in the data science workflow. It ensures that the data used for analysis is accurate, consistent, and reliable. Data cleaning is an essential first step in  any data analysis process and is a basic topic covered in any entry-level Data Scientist Course. Advanced courses might include advanced techniques for data cleaning and preparing data for analysis.  

Data Cleaning  Methods 

Here are some common methods used in data cleaning:

  • Handling Missing Values
  • Identify missing values: Check for missing values in the dataset.
  • Imputation: Fill in missing values using techniques like mean, median, mode, or more advanced methods like interpolation or machine learning-based imputation.
  • Deletion: Delete rows or columns with a high proportion of missing values if they cannot be reliably imputed.
  • Handling Duplicates
  • Identify duplicate records: Check for and remove duplicate rows in the dataset.
  • Duplicate key columns: Ensure uniqueness of key columns if they should be unique, or merge duplicate key columns.
  • Data Transformation
  • Standardisation: Convert data into a standard format, such as converting all text to lowercase or all dates to a consistent format.
  • Normalisation: Scale numeric features to a standard range, such as between 0 and 1.
  • Encoding categorical variables: Convert categorical variables into numerical format, either by one-hot encoding, label encoding, or other encoding techniques.
  • Outlier Detection and Treatment
  • Identify outliers: Use statistical methods or visualisation techniques to detect outliers.
  • Treatment: Decide whether to remove outliers, cap them, or transform them using techniques like winsorisation.
  • Outliers are of particular significance in research studies. The anomalies and aberrations they represent might be a key area for researchers to investigate. Thus, the way outliers are handled and explained in a Data Scientist Course tailored for researchers might be different from the way there are in a course targeting business professionals.  
  • Data Formatting
  • Date parsing: Convert date and time variables into a consistent format.
  • Text cleaning: Remove special characters, punctuation, and unnecessary whitespace from text data.
  • Data type conversion: Ensure that variables are in the correct data type (for example, numeric, categorical, datetime).
  • Handling Inconsistent Data
  • Standardising units: Convert all measurements to a consistent unit of measurement.
  • Resolving inconsistencies: Resolve discrepancies in data values by correcting errors or reconciling differences.
  • Feature Engineering
  • Creating new features: Derive new features from existing ones to improve model performance or capture additional information. This is a key capability for business strategists and developers and a Data Science Course in Mumbai , Pune and such commercialised cities might offer focused learning in this discipline as an option within the general curriculum in response to the demand among learners. 
  • Addressing Skewed Data
  • Transformation: Apply mathematical transformations like logarithmic or square root transformation to reduce skewness in data distributions.
  • Data Quality Assessment
  • Profiling: Generate summary statistics and data quality metrics to assess the quality of the dataset.
  • Visual inspection: Visualise data distributions and relationships to identify anomalies or errors.
  • Documentation
  • Documenting changes: Keep track of all the cleaning operations performed on the dataset for reproducibility and transparency.


Data cleaning is an iterative process, often requiring multiple rounds of cleaning and validation to ensure that the data is prepared adequately for analysis or modelling. Most learning centres in cities like Mumbai or Chennai where professional courses are offered include substantial coverage on data cleaning in view of its importance in any data analysis initiative. Whether you enrol for a Data Science Course in Mumbai, or in Chennai or Bangalore, the methods related in this article will be part of the course curriculum. While there could be advanced methods evolving, mastery of these basic methods will position you better to learn those methods. 

Contact us:

Name: ExcelR- Data Science, Data Analytics, Business Analytics Course Training Mumbai

Address: 304, 3rd Floor, Pratibha Building. Three Petrol pump, Lal Bahadur Shastri Rd, opposite Manas Tower, Pakhdi, Thane West, Thane, Maharashtra 400602

Phone Number: 09108238354

Email ID: