Cleaning Up Bad Data Print

  • 0

Fixing what accumulated.

WHAT TO ESTABLISH FIRST

How the bad data got there.

WHY BEFORE FIXING

Otherwise it reappears.

WHAT COMMON CAUSES ARE

Missing constraints Imports without validation Application defects Manual changes

WHAT TO FIND

Duplicates Orphaned rows Values outside permitted ranges Inconsistent formats Empty strings where absence was meant

HOW TO FIND ORPHANS

A query for rows whose reference has no match.

WHAT TO DO BEFORE CHANGING ANYTHING

Take a backup, and copy the affected rows.

WHAT ORDER TO WORK IN

Identify Quantify Decide the rule Apply in batches Verify

WHY QUANTIFY

The volume determines the approach, and sometimes the decision.

WHAT TO DECIDE PER PROBLEM

Correct it Remove it Leave it, and exclude it

WHO DECIDES

The business, for anything with meaning.

WHAT TO DO AFTERWARDS

Add the constraint that prevents recurrence.

WHY THAT IS THE ACTUAL FIX

Cleaning without constraining means cleaning again.

WHAT TO RECORD

What was changed, and by what rule.


Was this answer helpful?
Back

Are you happy with your experience? Leave us a review on Trustpilot.


Trustpilot