Cleaning should include

• removal of invalid values
• location of null sets
• normalization of data
• consideration of data types and consistency.