Dataset Properties#

Properties define a dataset’s content. Each property has a type, is required or optional, and may or may not allow null. Here’s an example dataset schema:

sample-aapl-dataset-edit-schema.png

Important

The _system prefix (case-insensitive) is reserved for Apperate system tables and columns. You are forbidden to prefix dataset IDs or dataset property names with _system (case-insensitive).

Types#

Here are the various property types:

Type

Description

date

A date, which can be specified using one of several different formats. (The schema editor lists the formats and includes examples.)

number

Floating point number.

integer

Any whole number, positive or negative, including zero.

string

A series of characters.

object

A JSON Object representation.

array

An indexed sequence of values.

boolean

TRUE or FALSE; 1 or 0.

any

This JSON schema option supports using mixed types. It is useful in prototype situations when you are not sure what type you need or in situations where the downstream consumer does not care about type, and you want to move data quickly. Another case for using is when the upstream data is out of your control and is presented as a mixed type or a type that can change, and you want to avoid data ingestion failure.

Constraints#

Required#

A property can be marked as Required or left as optional. Data ingestion fails for data that is missing any required properties.

Allow null#

Allows null values for the property when checked, unselect the Allows null option for properties that must never be null.

Note

For CSV files, an empty field is interpreted as an empty string; it is never interpreted as null.