---
title: "Developers Documentation"
vignette: >
  %\VignetteIndexEntry{dev_documentation}
  %\VignetteEngine{quarto::html}
  %\VignetteEncoding{UTF-8}
knitr:
  opts_chunk:
    collapse: true
    comment: '#>'
---

# Types of nodes

-   **col_copy**: These are absorbed by the domain_init node and do not exist as an independent node in our internal data model. Example:

    ``` yml
    columns:
      - id: USUBJID
    ```

-   **col_mutate**: col_mutates exist as an independent node in our data model and their source is an output of the domain_init node. In other words, a mutate node MUST have the domain_init as a direct parent. Example:

    ``` yml
    columns:
      - id: PLANNED_ARM
        method: ADSL.ARM
    ```

-   **col_echo**: An echo can have a col_compute, col_echo, col_mutate, or a col_copy as parents. The depend_cols of echo's ARE used when constructing the "external" nodes. Example:

    ``` yml
    columns:
      - id: PLANNED_ARM2
        method: ADSL.PLANNED_ARM
    ```

-   **row_compute**: Row operations cannot create a new column in the dataset, but they can operate on any existing column(s). They can have as parents a col_compute, col_mutate, col_echo, row_compute, col_copy, or domain_init

-   **col_compute**: col_computes are always derivations in CDISC terminology and exist as independent nodes in our internal data model. They can have as parents rows, col_echos, col_mutates, col_copies, or domain_inits. Compute MUST reference a component in the YAML domain specification. Example:

    ``` yml
    columns:
      - id: AGE_GRP
        component:
          id: age_group_01
    ```