🏠 Dashboard · ☑ Progress · 🔎 Finder · 📚 Block · ← Previous · Next →
Sequence 237 — Data Quality deepening - baseline profile and rulebook
Context: Chapter 13 - Data Quality Management | Applied Lab Deepening
Source: CDMP Master Execution Manual
When / purpose
Separate observation, rule definition and measurement before repairing data.
What you should learn
- Profile before repair.
- Rules require business meaning and measurable conditions.
- Dimensions/metrics must be source-grounded to Chapter 13.
- Baseline evidence is required for before/after comparison.
What to read / study
- MUST: Chapter 13 applied lab folder — use once built; do not invent DAMA-specific dimensions from memory.
- MUST: Meridian defect map / synthetic data — use seeded defects and preserve canonical raw data.
Do this in order
- Profile selected datasets with SQL and/or pandas.
- Record row counts, nulls, duplicates, invalid references, stale/impossible values as applicable.
- Define corresponding quality rules using Chapter 13 terminology.
- Record baseline measures before remediation.
Meridian application
Use the documented Meridian defect catalog rather than inventing random bad data.
Create / save
DQ baseline profile + rulebook.
Completion gate
You can explain each rule, how it is measured and why the business cares.
NEXT → Sequence 238