Grapher
In the grapher
step the work should be minimal. Here, we create a grapher
view by adapting our Garden dataset to fit the Grapher requirements.
Grapher views are still normal datasets, but they adapt the data to the way it must look when being inserted to MySQL. For each grapher view, there is a corresponding matching grapher://
step automatically generated which does the actual insert to MySQL, if MySQL credentials have been configured.
A typical flow up to the Grapher step could look like:
flowchart LR
upstream1((____)):::node -.->|copy| snapshot1((____)):::node
snapshot1((____)):::node -->|format| meadow1((____)):::node
meadow1((____)):::node -->|harmonize| garden1((____)):::node
garden1((____)):::node -->|format| grapher1((____)):::node
grapher1((____)):::node -->|load| grapher2((____)):::node
subgraph id0 [Upstream]
upstream1
end
subgraph id1 [Snapshot]
snapshot1
end
subgraph id2 [Meadow]
meadow1
end
subgraph id3 [Garden]
garden1
end
subgraph id4 [Grapher]
grapher1
end
subgraph id5 [Grapher]
grapher2
end
subgraph id [ETL]
id1
id2
id3
id4
end
classDef node fill:#002147,color:#002147
classDef node_ss fill:#002147,color:#fff
In principle, a grapher step only loads a single garden step.
Note that the diagram shows a final step outside of the ETL. This is when the grapher://
step is executed, and takes data from the ETL (from the etl garden
step) and imports it to oure database.
TODO: Add an example of code