Skip to content

Add an event-level extraction to the plan

Source: R/plan.R
ds.omop.plan.events.Rd

Extracts rows from a single OMOP clinical data table, optionally filtered by concept set, time window, temporal specification, and custom filters. The output format is controlled by the representation parameter (long, wide, or features).

Usage

ds.omop.plan.events(
  plan,
  name,
  table,
  columns = NULL,
  concept_set = NULL,
  time_window = NULL,
  temporal = NULL,
  date_handling = NULL,
  filters = NULL,
  visit_filter = NULL,
  concept_col = NULL,
  representation = list(format = "long")
)

Arguments

plan

An omop_plan object.

name

Character; output name used as a key in the plan's outputs list.

table

Character; source OMOP table name (e.g. "condition_occurrence", "drug_exposure").

columns

Character vector; columns to include from the table. If NULL, the server selects default columns.

concept_set

Numeric vector or concept set spec; concept IDs used to filter rows via the standard concept ID column of the table.

time_window

Named list with start_date and end_date for calendar-based filtering.

temporal

An omop_temporal_spec object or list; temporal filtering relative to a cohort index date. See omop.temporal.

date_handling

A list; date handling specification controlling how date columns are transformed. See omop.date_handling.

filters

Named list; additional custom filter DSL expressions (nested and/or of leaves, each list(var=, op=, value=)). Validated fail-closed server-side: leaves on identifier or blocked columns, and narrow fingerprinting operators, are rejected. Use this to filter by unit_concept_id or a *_type_concept_id for unit/type scoping.

visit_filter

Named list list(concept_ids = ...); restrict events to visits of those visit_concept_id values via the visit_occurrence_id link.

concept_col

Character; override the concept column the concept_set scopes (default: the table's domain concept), e.g. "unit_concept_id" to extract one unit for harmonization.

representation

Named list with format (one of "long", "wide", "features", or "sparse") and optional format-specific settings. Wide/features/sparse may set grain to "person" (default) or "episode"; an index-relative window requires episode grain.

Value

The modified omop_plan with the event-level output appended.

See also

omop.temporal, omop.date_handling, ds.omop.plan.features

Examples

if (FALSE) { # \dontrun{
plan <- ds.omop.plan()
plan <- ds.omop.plan.events(plan,
  name = "conditions",
  table = "condition_occurrence",
  concept_set = c(201826, 443238),
  temporal = omop.temporal(index_window = list(start = -365, end = 0)),
  date_handling = omop.date_handling(mode = "relative")
)
} # }