Given an agent's validation plan that had undergone interrogation via
interrogate(), did every single validation step result in zero failing
test units? Using the
all_passed() function will let us know whether that's
TRUE or not.
all_passed(agent, i = NULL)
An agent object of class
A vector of validation step numbers. These values are assigned to
each validation step by pointblank in the order of definition. If
NULL (the default), all validation steps will be used for the evaluation
of complete passing.
A logical value.
all_passed() function provides a single logical value based on an
interrogation performed in the agent-based workflow. For very large-scale
validation (where data quality is a known issue, and is perhaps something to
be tamed over time) this function is likely to be less useful since it is
quite stringent (all test units must pass across all validation steps).
Should there be a requirement for logical values produced from validation, a
more flexible alternative is in using the test (
test_*()) variants of the
validation functions. Each of those produce a single logical value and each
and have a
threshold option for failure levels. Another option is to
utilize post-interrogation objects within the agent's x-list (obtained by
get_agent_x_list() function). This allows for many possibilities
in producing a single logical value from an interrogation.
# Create a simple table with # a column of numerical values tbl <- dplyr::tibble(a = c(4, 5, 7, 8)) # Validate that values in column # `a` are always greater than 4 agent <- create_agent(tbl = tbl) %>% col_vals_gt(vars(a), value = 3) %>% col_vals_lte(vars(a), value = 10) %>% col_vals_increasing(vars(a)) %>% interrogate() # Determine if these column # validations have all passed by # using `all_passed()` (they do) all_passed(agent = agent) #>  TRUE