filter_by_validation: Filter feeds or datasets by validation quality

In mobdb: Access the 'Mobility Database' API to Discover Transit Feeds

Filter feeds or datasets by validation quality

Description

Filter feed or dataset results by validation quality thresholds. This is a convenience wrapper around get_validation_report() that returns the original data filtered to only include feeds/datasets meeting your quality criteria.

Note: This function does not support GBFS validation reports at this time as GBFS validation reports are located at a different endpoint and have a different validation criteria.

Usage

filter_by_validation(
  data,
  max_errors = NULL,
  max_warnings = NULL,
  max_info = NULL,
  require_validation = TRUE
)

Arguments

data	A tibble from feeds(), mobdb_datasets(), or mobdb_search().
max_errors	Maximum number of validation errors allowed. Use 0 for error-free feeds. If NULL (default), no error filtering is applied.
max_warnings	Maximum number of validation warnings allowed. If NULL (default), no warning filtering is applied.
max_info	Maximum number of informational notices allowed. If NULL (default), no info filtering is applied.
require_validation	Logical. If TRUE (default), exclude feeds/datasets that have no validation data. If FALSE, include them in results.

Value

A filtered version of the input data frame containing only feeds/datasets that meet the specified quality criteria.

See Also

get_validation_report() to inspect validation metrics, view_validation_report() to view full validation reports

Examples

# Create sample data with validation information (search results structure)
sample_data <- tibble::tibble(
  id = c("mdb-1", "mdb-2", "mdb-3"),
  provider = c("Agency A", "Agency B", "Agency C"),
  latest_dataset = tibble::tibble(
    id = c("mdb-1-202501", "mdb-2-202501", "mdb-3-202501"),
    validation_report = tibble::tibble(
      total_error = c(0L, 5L, 100L),
      total_warning = c(10L, 50L, 500L),
      total_info = c(5L, 10L, 20L)
    )
  )
)

# Filter to feeds with zero errors
filter_by_validation(sample_data, max_errors = 0)

# Filter with multiple criteria
filter_by_validation(sample_data, max_errors = 10, max_warnings = 100)


# With real API data:
ca_feeds <- feeds(
  country_code = "US",
  subdivision_name = "California",
  data_type = "gtfs"
)
clean_feeds <- filter_by_validation(ca_feeds, max_errors = 0)

mobdb documentation built on Aug. 29, 2026, 1:07 a.m.