Dimensions

Dimension nodes include a query that can select from any other node to create a representation of a dimension. They must always have a primary key configured, and can have any number of associated dimensional attributes.

One key feature of dimension nodes is the ability to configure join links. Any DJ node can be linked to a dimension node via two different types of dimension linking. These links are used to build out the dimensional metadata edges of the DJ DAG, enabling DJ to find all accessible dimensions for each metric.

Here are a few example dimension nodes:

%%{init: { "theme": "base", 'themeVariables': { 'primaryColor': '#ffefd0'}}}%% erDiagram "default.country | Country" { id int PK name str country_code str population long } "default.dispatcher | Dispatcher" { dispatcher_id int PK company_name str phone str } "default.contractor | Contractor" { contractor_id int PK company_name str contact_name str contact_title str address str city str state str postal_code str country str }

Creating Dimension Nodes

AttributeDescriptionType
nameUnique name used by other nodes to select from this nodestring
display_nameA human readable name for the nodestring
descriptionA human readable description of the nodestring
modepublished or draft (see Node Mode)string
queryA SQL query that selects from other nodesstring

Assume a default.dispatchers source node was defined as follows.

curl -X 'POST' \
  'http://localhost:8000/nodes/source/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
        "columns": [
            {"name": "dispatcher_id", "type": "int"},
            {"name": "company_name", "type": "string"},
            {"name": "phone", "type": "string"}
        ],
        "description": "Contact list for dispatchers",
        "mode": "published",
        "name": "default.dispatchers",
        "catalog": "warehouse",
        "schema_": "roads",
        "table": "dispatchers"
  }'

from datajunction import DJBuilder, NodeMode
dj = DJBuilder(DJ_URL)

contractor = dj.create_dimension(
    description="Contractor dimension",
    display_name="Default: Contractor",
    name="default.contractor",
    primary_key=['contractor_id'],
    mode=NodeMode.PUBLISHED,  # for draft nodes, use `mode=NodeMode.DRAFT`
    query="""
    SELECT
        contractor_id,
        company_name,
        contact_name,
        contact_title,
        address,
        city,
        state,
        postal_code,
        country
    FROM default.contractors
    """
)

dj.sources.create(
  {
    name: "default.dispatchers",
    mode: "published",
    description: "Contact list for dispatchers",
    catalog: "warehouse",
    schema_: "roads",
    table: "dispatchers",
    columns: [
        {name: "dispatcher_id", type: "int"},
        {name: "company_name", type: "string"},
        {name: "phone", type: "string"}
    ]
  }
).then(data => console.log(data))

You can then define a dimension node that selects from the default.dispatchers source node and includes a primary key. Let’s call it default.dispatcher.

curl -X POST http://localhost:8000/nodes/dimension/ \
-H 'Content-Type: application/json' \
-d '{
    "name": "default.all_dispatchers",
    "description": "All dispatchers",
    "mode": "published",
    "query": "SELECT dispatcher_id, company_name, phone FROM default.dispatchers",
    "primary_key": ["dispatcher_id"]
}'

dj.dimensions.create(
  {
    name: "default.all_dispatchers",
    mode: "published",
    description: "All dispatchers",
    query: `
        SELECT
        dispatcher_id,
        company_name,
        phone
        FROM default.dispatchers
    `,
    primary_key: ["dispatcher_id"]
  }
).then(data => console.log(data))

← Clients
Dimension Links →