Creating Custom Blocks

Overview

blockr ships with blocks for common operations, but you can write your own in pure R. A custom block wraps any R function (a data transformation, an API call, a model fit) into a reusable component that anyone can drag onto the canvas.

What the video covers

This 18-minute walkthrough demonstrates the full process:

1. Block anatomy: every block has inputs (UI controls), an expression (the R code that runs), and an output (the result passed downstream)
2. Defining inputs: use blockr's input constructors to create dropdowns, numeric sliders, text fields, and other controls
3. Writing the expression: reference input values in a standard R function that transforms the incoming data
4. Registering the block: make your block discoverable in the block menu so users can find and add it
5. Connecting to the DAG: how blockr's reactive engine automatically re-runs your block when upstream data changes

Key concepts

• Blocks are pure R. No JavaScript, no HTML, no Shiny modules. Just an R function with typed inputs.
• The input/output contract is enforced: your block declares what data types it accepts and produces, so blockr can validate connections at the DAG level.
• Blocks are packaged as regular R functions. Distribute them in an R package, share them as scripts, or register them dynamically at app startup.

Next steps

• Testing blocks: write tests for your custom blocks
• Developer documentation: full blockr.core API reference