Run pointblank validation on an eData table

Applies pointblank validation rules to check data quality and schema compliance for eData format tables. Can operate in two modes:

• Agent mode (agent = TRUE): Creates an agent for reporting and interrogation

• Pipeline mode (agent = FALSE): Returns validated data directly for use in pipelines

Usage

pb_validate_edata_table(
  data,
  table_name,
  validation_steps,
  agent = TRUE,
  actions = action_levels()
)

Arguments

data

Data frame to validate

table_name

Name of the table being validated (for reporting)

validation_steps

Function that takes a data frame or agent and returns it with validation steps added. Should be a function like: function(x) { x |> col_vals_not_null(...) |> ... }

agent

Logical. If TRUE (default), returns a pointblank agent object. If FALSE, returns the validated data with validation failures removed.

actions

Action levels for pointblank agent (only used when agent = TRUE)

Value

If agent = TRUE, a pointblank agent object containing validation results. If agent = FALSE, the input data with validation failures removed.

Details

In agent mode, the agent object can be used to:

• Extract validation results with get_sundered_data()

• Generate reports with get_agent_report()

• Check pass/fail status

In pipeline mode, failed rows are automatically removed from the data.