Skip to contents

When the agent has all the information on what to do (i.e., a validation plan which is a series of validation steps), the interrogation process can occur according its plan. After that, the agent will have gathered intel, and we can use functions like get_agent_report() and all_passed() to understand how the interrogation went down.


  extract_failed = TRUE,
  get_first_n = NULL,
  sample_n = NULL,
  sample_frac = NULL,
  sample_limit = 5000



An agent object of class ptblank_agent that is created with create_agent().


An option to collect rows that didn't pass a particular validation step. The default is TRUE and further options allow for fine control of how these rows are collected.


If the option to collect non-passing rows is chosen, there is the option here to collect the first n rows here. Supply the number of rows to extract from the top of the non-passing rows table (the ordering of data from the original table is retained).


If the option to collect non-passing rows is chosen, this option allows for the sampling of n rows. Supply the number of rows to sample from the non-passing rows table. If n is greater than the number of non-passing rows, then all the rows will be returned.


If the option to collect non-passing rows is chosen, this option allows for the sampling of a fraction of those rows. Provide a number in the range of 0 and 1. The number of rows to return may be extremely large (and this is especially when querying remote databases), however, the sample_limit option will apply a hard limit to the returned rows.


A value that limits the possible number of rows returned when sampling non-passing rows using the sample_frac option.


A ptblank_agent object.


Create a simple table with two columns of numerical values.

tbl <-
    a = c(5, 7, 6, 5, 8, 7),
    b = c(7, 1, 0, 0, 0, 3)


## # A tibble: 6 × 2
##       a     b
##   <dbl> <dbl>
## 1     5     7
## 2     7     1
## 3     6     0
## 4     5     0
## 5     8     0
## 6     7     3

Validate that values in column a from tbl are always less than 5. Using interrogate() carries out the validation plan and completes the whole process.

agent <-
    tbl = tbl,
    label = "`interrogate()` example"
  ) %>%
  col_vals_gt(columns = vars(a), value = 5) %>%

We can print the resulting object to see the validation report.


Function ID


See also

Other Interrogate and Report: get_agent_report()