Skip to content

Create an output specification

Source: R/recipe.R
omop_output.Rd

Defines how to shape the extracted data into a result table. Each output selects a subset of variables from the recipe, targets a population, and specifies a layout type (e.g. wide person-level, long event-level, or feature matrix). Outputs are passed to omop_recipe via its outputs argument and determine the server-side plan structure produced by recipe_to_plan.

Usage

omop_output(
  name = "output_1",
  type = c("wide", "long", "features", "survival", "intervals", "baseline",
    "temporal_covariates", "person_period"),
  variables = NULL,
  population_id = "base",
  options = list(),
  result_symbol = NULL
)

Arguments

name

Character; output table name (used as key in the recipe).

type

Character; output layout type. One of "wide", "long", "features", "survival", "intervals", "baseline", "temporal_covariates", or "person_period". A "long" output that spans multiple source tables always splits into one per-table output (named <name>_<table>); there is no single cross-table joined frame. The former "joined_long" and "covariates_sparse" recipe labels are rejected because they had no faithful executable mapping; use split "long" outputs or the lower-level temporal-covariates plan API.

variables

Character vector or NULL; variable names to include (NULL means all variables in the recipe).

population_id

Character; which population to use (default "base").

options

Named list; type-specific options (e.g. tar for survival outputs).

result_symbol

Character or NULL; R symbol name for the result on the server (auto-generated as D_<name> if NULL).

Value

An omop_output object.

See also

omop_recipe, recipe_to_plan

Examples

if (FALSE) { # \dontrun{
recipe <- omop_recipe(
  variables = omop_variable(table = "condition_occurrence",
                            concept_id = 201820, format = "binary"),
  outputs = omop_output(name = "features_wide", type = "wide")
)
} # }