← All nodes
Data Quality

Data Profile

Summarises every column — how many blanks, how many distinct values, the smallest and largest value and the commonest pattern — as a table with one row per column.

When to use it

Use it first on any new or unfamiliar data, and on a schedule to notice when a feed starts arriving with gaps or a changed format.

In a pipeline

Settings

SettingTypeRequiredDefault or hint
Columns to profileselect—None picked: every column is profiled
Rows to samplenumber—10000
Detect value patternscheckbox—true

Example

Profile of a customer list

  • Rows to sample=10000
Customers
IDEmailAmount
C-1ann@example.com120.5
C-2—80
C-3bob@example.com310
One row per column
ColumnRows profiledNull / blankNull %DistinctMinMaxSamplePattern
ID300.0%3C-1C-3C-1text
Email3133.3%2ann@example.combob@example.comann@example.comemail
Amount300.0%380310120.5decimal

Worth knowing

  • The profile is ordinary rows, so you can filter it (Null % above 5%), sort it or export it like any other table.
  • Rows to sample profiles the first N rows (10,000 by default) to keep huge tables quick; 0 profiles every row. The note says when a sample was used.
  • Min and Max order values the way Sort Rows does — numbers as numbers, dates as dates, text alphabetically.
  • Pattern is the commonest shape among the values (integer, decimal, date, email, url, phone, hex32 or text), with its share when not every value has it.

Others in Data Quality