Primary Key Finder

What is Primary Key Finder?

Primary Key Finder examines a CSV and reports which columns could actually serve as a primary key: single columns that are unique and complete, composite combinations that become unique together, and — most usefully — the near-misses, columns that would be a key except for a handful of blanks or a couple of duplicate values. Those almost-keys are usually the interesting finding, because they point at a data quality problem rather than a modelling one.

Used by

• Data analysts
• Researchers
• Business analysts
• Students
• Data migration teams
• Integration engineers

Recommended workflow

1. 1

   Check the join side

   Analyse how well two datasets' keys actually overlap.

   Open tool
2. 2

   Find related columns

   Detect functional dependencies between columns.

   Open tool
3. 3

   Investigate the duplicates

   See exactly which values repeat.

   Open tool

Frequently asked questions

What is an almost-key?

A column that would be a valid key except for a small number of blanks or duplicates — often fewer than five rows in a large file. It usually indicates a data quality issue in those rows rather than a genuinely unsuitable column.

Why can't a primary key contain nulls?

A key has to identify a row unambiguously, and a null means "unknown" — two unknown values cannot be shown to be different, so uniqueness cannot be guaranteed. Databases enforce this by rejecting the constraint.

How many columns does it try for composite keys?

It tests combinations up to a practical limit; beyond a few columns the number of combinations grows too fast to be useful, and a key that wide is usually a sign the table needs restructuring.

Does uniqueness in my sample guarantee a valid key?

No. It tells you the constraint holds for this data. Whether it will keep holding depends on what the column means — an email address might be unique today and reused after an account is deleted.