A data domain is the business type of a column: choosing one sets the column's storage type and its suggested default values. You set the domain, the default value and the PII Domain tag in the Columns step of a table; see Define table columns for the grid itself.
The picker groups the domains into six categories: Basic Types, Time Types, Identifiers, Specialized Types, Arrays and Complex Types and Custom. Around 50 domains are available across the core and extended lists, and each domain comes with suggested default values you can start from.
When you pick Custom, the types offered depend on the engine:
| Engine | Data types |
|---|---|
| PostgreSQL | integer, bigint, float, decimal, boolean, text, varchar, date, timestamp |
| ClickHouse | Int, Float32, String, FixedString, Date32, Datetime64 |
| Domain | Category | Stored as |
|---|---|---|
| Text | Basic Types | String |
| ShortText | Basic Types | FixedString(50) |
| SingleCharacter | Basic Types | FixedString(1) |
| Number | Basic Types | Int32 |
| SmallNumber | Basic Types | Int16 |
| LargeNumber | Basic Types | Int64 |
| UnsignedNumber | Basic Types | UInt32 |
| Decimal | Basic Types | Decimal(18,2) |
| Amount | Basic Types | Decimal64(2) |
| LargeAmount | Basic Types | Decimal128(2) |
| HighPrecisionAmount | Basic Types | Decimal(18,6) |
| Percentage | Basic Types | Decimal(5,2) |
| Rate | Basic Types | Decimal(10,6) |
| Probability | Basic Types | Decimal(5,4) |
| Boolean | Basic Types | Bool |
| Date | Time Types | Date32 |
| DateTime | Time Types | DateTime |
| PreciseDateTime | Time Types | DateTime64(3) |
| Time | Time Types | DateTime64(0) |
| Year | Time Types | UInt16 |
| Month | Time Types | UInt8 |
| Quarter | Time Types | UInt8 |
| DateRange | Time Types | Tuple(Date, Date) |
| Duration | Time Types | Int32, seconds |
| ID | Identifiers | UInt32 |
| LargeID | Identifiers | UInt64 |
| UUID | Identifiers | UUID |
| ExternalID | Identifiers | String |
| GeoPoint | Specialized Types | Point |
| IPAddress | Specialized Types | IPv4 |
| Status | Specialized Types | LowCardinality(String) |
| CountryCode | Specialized Types | LowCardinality(FixedString(3)) |
| KeyValuePairs | Arrays and Complex Types | Map(String, String) |
| RelatedList | Arrays and Complex Types | Array(Tuple(String, UInt128)) |
now(), today(), -1 and 0. 00.generateUUIDv4(), generateSnowflakeID(), generateShortID() and generateTimeBasedID().The Default Value cell holds the value a column takes when no value is given. While you edit it, DataMind suggests values already used by other columns of the same table whose storage type falls in the same group as this column's type. Suggestions skip the column you are editing, skip empty values and remove duplicates.
Nullable(String) matches String, and Decimal(10,2) matches Decimal. UInt32 and UInt64 group as Decimal Number. A type the groups do not cover matches only its exact type.String, as in String, FixedString(50), Nullable(String) or LowCardinality(String). For those columns, the value appears in the cell wrapped in single quotes, with inner quotes doubled.The PII Domain cell marks which kind of personal data a column holds. The choices span eight categories:
The catalogue holds 51 options in total, covering names, contact details, card and bank details, health data, government identifiers, credentials and sensitive dates. To clear a tag, choose None. The tag decides what the PII policy treats as personal data for that column; how the platform acts on it is set up on PII detection.
A data domain is the business type you recognise, such as Email or Amount. A data type is the storage form behind it, such as String or Decimal64(2). Choosing a domain fills Data Type with the type behind it.
Yes. Each domain carries a storage type, and picking the domain fills Data Type with that type. Custom is the exception: it lets you enter the type yourself.
Custom means you choose the storage type yourself instead of taking the type behind a domain. Its description reads "You can choose ClickHouse datatype on your own". The types offered depend on the engine: PostgreSQL offers integer, bigint, float, decimal, boolean, text, varchar, date and timestamp; ClickHouse offers Int, Float32, String, FixedString, Date32 and Datetime64.
From other columns of the same table: DataMind suggests defaults already used by columns whose storage type falls in the same group as this column's type. Empty values are skipped and duplicates removed; the column you are editing is not used.
Columns whose types fall in the same group: Whole Number, Decimal Number, Date, Date/Time, Time, Duration, True/False and Text. Wrappers and parameters are stripped before matching, UInt32 and UInt64 count as Decimal Number, and a type outside the groups matches only its exact type.
The suggestions show text values without the surrounding quotes. In the cell, the value appears wrapped in single quotes, with inner quotes doubled, and that is how it is stored.
Currency. Its description reads "Currency codes (USD, EUR, GEL)".
Choose None in the PII Domain cell. It clears the tag.
Eight categories: Personal Identity, Contact & Location, Financial / PCI, Health / PHI, Government & Tax, Digital Identity, Credentials & Secrets and Sensitive Dates.
The tag decides what the PII policy treats as personal data in that column. How the platform acts on personal data is set up on PII detection.