EDA

EDA

1. Data Granularity - Levels in the data. ie Time, Years, Months, Weeks, Days, Hours

2. Consistency - Dates 01/01/2000 or 1/1/00

3. Corruption and Accuracy - System generated problems / Human errors / Out of date

4. Data Duplication

5. Missing Data

6. SOLUTIONS

   • capitalisation (transform all)
   • Combine or concatenations of variables
   • Careful use of fomats
   • Removals of unwanted characters
     • Exclusion
   • consistent units
   • Add system checks
   • Reduce the variable types

7. Data Types / Model Roles

   • Model Roles

8. Categorise Data

   • Discrete Data
     • Gender
     • Make of car
     • Number of cars
     • Data that can only take certain values.
   • Continuous Data
     • Data that can take any value (within a range)
     • Bank balances
     • Measurements
     • Dates

9. Data Levels

   • Data Source Levels