correctness_check: Validate Data Against Correctness Rules

In DQA: Data Quality Assessment Tools

Validate Data Against Correctness Rules

Description

This function validates a data frame against a set of correctness rules specified in another data frame. It allows for complex validation operations, comparison with reference data, and detailed reporting.

Usage

correctness_check(
  S_data,
  M_data,
  Result = FALSE,
  show_column = NULL,
  date_parser_fun = smart_to_gregorian_vec,
  golden_data = NULL,
  key_column = NULL,
  external_data = NULL,
  var_select = "all",
  batch_size = 1000,
  verbose = FALSE
)

Arguments

S_data	A data frame containing the data to be validated.
M_data	A data frame containing the validation rules. Must have at least the following columns: VARIABLE: The name of the variable to validate (must match column names in S_data) Correctness_Rule: The validation rule as an R expression (string) TYPE: The data type of the variable ("date", "numeric", or other) Correctness_Error_Type: (Optional) Classification of the error type
Result	Logical. If TRUE, returns the detailed results for each row in S_data. If FALSE (default), returns a summary of validation results.
show_column	Character vector. When Result=TRUE, specifies additional columns from S_data to include in the output.
date_parser_fun	Function to convert date strings to Date objects. Default is smart_to_gregorian_vec, which should handle various date formats including Jalali dates.
golden_data	Optional data frame or list containing reference data for validation. Accessible within rules via the GOLDEN variable.
key_column	Character string specifying the column name that links rows in S_data to corresponding rows in golden_data. Required when comparing individual rows with golden_data.
external_data	Optional list or data frame containing additional data for validation rules.
var_select	Character vector or numeric indices specifying which variables from M_data to validate. By default, it validates all variables.
batch_size	integer. Number of rows to process in each batch (for efficiency).
verbose	logical. If TRUE, prints progress messages.

Details

The function evaluates each rule specified in M_data against the corresponding data in S_data. Rules are R expressions written as strings, evaluated in an environment where:

• Variables from S_data are available directly by name

• val refers to the current variable being validated

• GOLDEN provides access to reference data (when golden_data is provided)

Type conversion is applied to variables in S_data based on the TYPE column in M_data:

• "date": Values are converted using the date_parser_fun

• "numeric": Values are converted to numeric

• Other types: No conversion is applied

Special handling for date comparisons is provided, including automatic wrapping of GOLDEN references when comparing dates.

Value

If Result=FALSE (default): A data frame with one row per validated variable containing:

• VARIABLE: Variable name

• Condition_Met: Count of rows meeting the condition

• Condition_Not_Met: Count of rows not meeting the condition

• NA_Count: Count of rows where validation produced NA

• Total_Applicable: Count of non-NA validation results

• Total_Rows: Total number of rows in S_data

• Percent_Met: Percentage of applicable rows meeting the condition

• Percent_Not_Met: Percentage of applicable rows not meeting the condition

• Error_Type: Value from Correctness_Error_Type column in M_data

If Result=TRUE: A data frame with one row per row in S_data, containing:

• One column per validated variable with logical values (TRUE/FALSE/NA)

• Any additional columns specified in show_column

Examples

Authorized_drug<-data.frame(
  Drug_ID = 1:10,
  Drug_Name = c("Atorvastatin", "Metformin", "Amlodipine", "Omeprazole", "Aspirin",
                "Levothyroxine", "Sertraline", "Pantoprazole", "Losartan", "ASA"),
  stringsAsFactors = FALSE
)

golde<-data.frame(
  National_code = c("123", "456", "789","545","4454","554","665"),
  LastName = c("Bahman","Johnson","Williams","Brown","Jones","Garcia","Miller"),
  Certificate_Expiry = c("1404-07-01", "2030-01-12", "2025-01-11",
  "1404-06-28","2025-09-19",NA,NA),
  Blood_type = c("A-","B+","AB","A+","O-","O+","AB-"),
  stringsAsFactors = FALSE
)

S_data <- data.frame(
  National_code = c("123", "1456", "789","545","4454","554"),
  LastName = c("Aliyar","Johnson","Williams","Brown","Jones","Garcia"),
  VisitDate = c("2025-09-23", "2021-01-10", "2021-01-03","1404-06-28","1404-07-28",NA),
  Test_date = c("1404-07-01", "2021-01-09", "2021-01-14","1404-06-29","2025-09-19",NA),
  Certificate_validity = c("2025-09-23", "2025-01-12", "2025-02-11","1403-06-28","2025-09-19",NA),
  Systolic_Reading1 = c(110, NA, 145, 125,114,NA),
  Systolic_Reading2 = c(125, 150, NA, 110,100,NA),
  Prescription_drug= c("Atorvastatin", "Metformin", "Amlodipine",
   "Omeprazole", "Aspirin","Metoprolol"),
  Blood_type = c("A-","B+","AB","A+","O-","O+"),
  Height = c(178,195,165,NA,155,1.80),
  stringsAsFactors = FALSE
)

M_data <- data.frame(
  VARIABLE = c("National_code", "Certificate_validity", "VisitDate","Test_date",
               "LastName","Systolic_Reading1","Systolic_Reading2",
               "Prescription_drug","Blood_type","Height"),
  Correctness_Rule = c(
    "National_code %in% GOLDEN$National_code",
    "val <= GOLDEN$Certificate_Expiry",
    "((val >= '1404-06-01' & val <= '1404-06-31') | val == as.Date('2021-01-02'))",
    "val != VisitDate",
    "val %in% GOLDEN$LastName",
    "",
    "",
    "val %in% Authorized_drug$Drug_Name",
    "val %in% GOLDEN$Blood_type",
    ""),
  TYPE=c("numeric","date","date","date","character","numeric",
  "numeric","character","character","numeric"),
  Correctness_Error_Type=c("Error",NA,"Warning","Error",NA,NA,NA,NA,"Error","Warning"),
  stringsAsFactors = FALSE
)

result <- correctness_check(
  S_data = S_data,
  M_data = M_data,
  golden_data = golde,
  key_column = c("National_code"),
  Result =FALSE,
  external_data = Authorized_drug
)

print(result)
#
result <- correctness_check(
  S_data = S_data,
  M_data = M_data,
  golden_data = golde,
  #key_column = c("National_code"),#If you do not select a key, you can use Gold Data as a
  #list and your logical rules will be NA.
  Result =TRUE,
  external_data = Authorized_drug
)
print(result)

DQA documentation built on April 20, 2026, 9:06 a.m.