Exploratory Data Analysis (EDA) with SQL in MySQL

Before dashboards or models, you explore. EDA in SQL means profiling the data where it lives — sizes, distributions, gaps, and oddities — before a single row is exported.

Key Points

• Size check: SELECT COUNT(*), MIN(date), MAX(date) FROM t;
• NULL profile: SUM(col IS NULL) per column reveals data gaps
• Distributions: GROUP BY category ORDER BY COUNT(*) DESC
• Outlier scan: MIN, MAX, AVG side by side flags absurd values
• Duplicate check: GROUP BY natural key HAVING COUNT(*) > 1
• Sanity-check totals against a known report before trusting the table