Data model
Profile
A dataset and its columns.
| Slot | Type | Required | Description |
|---|---|---|---|
title | string | no | Name of the dataset. |
description | string | no | Definition of the dataset, column or value. |
identifier | string | no | Persistent identifier of the dataset. A DOI or URL is preferred. |
license | string | no | IRI of the license, or a statement of the terms of use. |
citation | string | no | Preferred citation. |
access_rights | string | no | Who may access the dataset. |
landing_page | string | no | IRI of a web page about the dataset. |
rows | integer | no | Number of rows of the table. |
columns | list of Column | yes | Columns, in table order. |
Column
One column of the table, described as a variable.
| Slot | Type | Required | Description |
|---|---|---|---|
name | string | yes | Header of the column, exactly as in the table. |
label | string | no | Human-readable name of the column or value. |
description | string | no | Definition of the dataset, column or value. |
datatype | Datatype | no | How values are stored. |
format | string | no | How dates, times or numbers are written in the table, for example %d/%m/%Y, in the notation of the source. |
language | string | no | Language of the text in a free-text column, as a lower-case BCP 47 language tag such as nl or en-gb. Set it only for columns of prose. |
level | Level | no | Level of measurement. |
unit | string | no | Unit as text, for example years. |
unit_iri | string | no | IRI of the unit, for example from QUDT. |
variable | string | no | IRI of the variable concept the column measures. |
source | string | no | IRI of a web page that documents the column or its variable. |
values | list of Value | no | Codes the column may hold, with their labels. Other codes may occur unless closed is true. |
closed | boolean | no | True if values lists every code the column may hold, apart from missing values. A list taken from data is not closed. |
observed | list of Count | no | Codes found in the data, in order of first occurrence, with their counts. They say what occurred, not what is allowed. |
missing_values | list of string | no | Codes that mean no value and have no label. |
missing_range | Range | no | Numbers that mean no value. |
completeness | float | no | Share of the rows that have a value, between 0 and 1. |
applies_when | string | no | Condition under which the column has a value, as the source writes it, for example REDCap branching logic. Rows that do not meet it are missing by design. |
minimum | float | no | Smallest number the column may hold, apart from missing values. A limit of the source, not the smallest number found. |
maximum | float | no | Largest number the column may hold, apart from missing values. |
group | string | no | Name of the group of columns that the column belongs to, for example the form of a REDCap field. |
identifying | boolean | no | True if the values identify a person or a record: names, record numbers, contact details. Such a column needs care when the data are shared or synthesized. |
mappings | list of Mapping | no | Links to concepts. |
Value
A code that the column may hold, with its label.
| Slot | Type | Required | Description |
|---|---|---|---|
code | string | yes | The code as stored in the table, as text. |
label | string | no | Human-readable name of the column or value. |
description | string | no | Definition of the dataset, column or value. |
missing | boolean | no | True if the code means no value. |
mappings | list of Mapping | no | Links to concepts. |
Count
A code found in the data and how often it occurs.
| Slot | Type | Required | Description |
|---|---|---|---|
code | string | yes | The code as stored in the table, as text. |
count | integer | yes | How many rows hold the code. |
Mapping
A link from a column or value to a concept. It is approved when it has a reviewer_id.
| Slot | Type | Required | Description |
|---|---|---|---|
match_type | MatchType | yes | How the concept relates to the column or value. |
object_id | string | yes | IRI of the concept. |
object_label | string | no | Name of the concept. |
confidence | float | no | Confidence between 0 and 1. |
provenance | string | no | Where the link came from, for example the name of a service. |
retrieved_at | string | no | When the link was made, as an ISO 8601 date or time. |
reviewer_id | string | no | Who approved the link. A link without a reviewer is a draft. |
description | string | no | Definition of the dataset, column or value. |
Range
An inclusive range of numbers.
| Slot | Type | Required | Description |
|---|---|---|---|
low | float | yes | Smallest number. |
high | float | yes | Largest number. |
Datatype
Storage type, named as in CSV on the Web.
| Value | Meaning |
|---|---|
string | |
number | |
integer | |
boolean | |
date | |
datetime |
Level
Level of measurement, as defined by DSV.
| Value | Meaning |
|---|---|
categorical | dsvo:CategoricalDataType |
nominal | dsvo:NominalDataType |
ordinal | dsvo:OrdinalDataType |
numerical | dsvo:NumericalDataType |
interval | dsvo:IntervalDataType |
ratio | dsvo:RatioDataType |
MatchType
Relation of a concept to a column or value, as defined by SKOS.
| Value | Meaning |
|---|---|
exact | skos:exactMatch |
close | skos:closeMatch |
broad | skos:broadMatch |
narrow | skos:narrowMatch |
related | skos:relatedMatch |