storms-validation.R

library(pointblank)

agent <-
  create_agent(
    tbl = ~dplyr::storms,
    actions = action_levels(
      warn_at = 0.05,
      stop_at = 0.10
    ),
    tbl_name = "storms",
    label = "Validation plan generated by `draft_validation()`."
  ) %>%
  # Expect that column `name` is of type: character
  col_is_character(
    columns = vars(name)
  ) %>%
  # Expect that column `year` is of type: numeric
  col_is_numeric(
    columns = vars(year)
  ) %>%
  # Expect that values in `year` should be between `1975` and `2020`
  col_vals_between(
    columns = vars(year),
    left = 1975,
    right = 2020
  ) %>%
  # Expect that column `month` is of type: numeric
  col_is_numeric(
    columns = vars(month)
  ) %>%
  # Expect that values in `month` should be between `1` and `12`
  col_vals_between(
    columns = vars(month),
    left = 1,
    right = 12
  ) %>%
  # Expect that column `day` is of type: integer
  col_is_integer(
    columns = vars(day)
  ) %>%
  # Expect that values in `day` should be between `1` and `31`
  col_vals_between(
    columns = vars(day),
    left = 1,
    right = 31
  ) %>%
  # Expect that column `hour` is of type: numeric
  col_is_numeric(
    columns = vars(hour)
  ) %>%
  # Expect that values in `hour` should be between `0` and `23`
  col_vals_between(
    columns = vars(hour),
    left = 0,
    right = 23
  ) %>%
  # Expect that column `lat` is of type: numeric
  col_is_numeric(
    columns = vars(lat)
  ) %>%
  # Expect that values in `lat` should be between `-90` and `90`
  col_vals_between(
    columns = vars(lat),
    left = -90,
    right = 90
  ) %>%
  # Expect that column `long` is of type: numeric
  col_is_numeric(
    columns = vars(long)
  ) %>%
  # Expect that values in `long` should be between `-180` and `180`
  col_vals_between(
    columns = vars(long),
    left = -180,
    right = 180
  ) %>%
  # Expect that column `status` is of type: character
  col_is_character(
    columns = vars(status)
  ) %>%
  # Expect that column `category` is of type: factor
  col_is_factor(
    columns = vars(category)
  ) %>%
  # Expect that column `wind` is of type: integer
  col_is_integer(
    columns = vars(wind)
  ) %>%
  # Expect that values in `wind` should be between `10` and `160`
  col_vals_between(
    columns = vars(wind),
    left = 10,
    right = 160
  ) %>%
  # Expect that column `pressure` is of type: integer
  col_is_integer(
    columns = vars(pressure)
  ) %>%
  # Expect that values in `pressure` should be between `882` and `1022`
  col_vals_between(
    columns = vars(pressure),
    left = 882,
    right = 1022
  ) %>%
  # Expect that column `tropicalstorm_force_diameter` is of type: integer
  col_is_integer(
    columns = vars(tropicalstorm_force_diameter)
  ) %>%
  # Expect that values in `tropicalstorm_force_diameter` should be between `0` and `870`
  col_vals_between(
    columns = vars(tropicalstorm_force_diameter),
    left = 0,
    right = 870,
    na_pass = TRUE
  ) %>%
  # Expect that column `hurricane_force_diameter` is of type: integer
  col_is_integer(
    columns = vars(hurricane_force_diameter)
  ) %>%
  # Expect that values in `hurricane_force_diameter` should be between `0` and `300`
  col_vals_between(
    columns = vars(hurricane_force_diameter),
    left = 0,
    right = 300,
    na_pass = TRUE
  ) %>%
  # Expect entirely distinct rows across all columns
  rows_distinct() %>%
  # Expect that column schemas match
  col_schema_match(
    schema = col_schema(
      name = "character",
      year = "numeric",
      month = "numeric",
      day = "integer",
      hour = "numeric",
      lat = "numeric",
      long = "numeric",
      status = "character",
      category = c("ordered", "factor"),
      wind = "integer",
      pressure = "integer",
      tropicalstorm_force_diameter = "integer",
      hurricane_force_diameter = "integer"
    )
  ) %>%
  interrogate()

agent