validate presence of required cols

seroanalytics · Oct 2, 2024 · cf94cd1 · cf94cd1
1 parent ec279a1
commit cf94cd1
Show file tree

Hide file tree

Showing 3 changed files with 39 additions and 12 deletions.
diff --git a/R/biokinetics.R b/R/biokinetics.R
@@ -237,17 +237,13 @@ biokinetics <- R6::R6Class(
                           covariate_formula = ~0,
                           preds_sd = 0.25,
                           time_type = "relative") {
-      if (!inherits(priors, "biokinetics_priors")) {
-        stop("'priors' must be of type 'biokinetics_priors'")
-      }
+      validate_priors(priors)
       private$priors <- priors
       validate_numeric(preds_sd)
       private$preds_sd <- preds_sd
       validate_time_type(time_type)
       private$time_type <- time_type
-      if (!(class(covariate_formula) == "formula")) {
-        stop("'covariate_formula' must be a formula")
-      }
+      validate_formula(covariate_formula)
       private$covariate_formula <- covariate_formula
       private$all_formula_vars <- all.vars(covariate_formula)
       if (is.null(data) && is.null(file_path)) {
@@ -264,11 +260,8 @@ biokinetics <- R6::R6Class(
         }
         private$data <- data
       }
-      unknown_vars <- private$all_formula_vars[which(!(private$all_formula_vars %in% names(private$data)))]
-      if (length(unknown_vars) > 0) {
-        stop(paste("All variables in 'covariate_formula' must correspond to data columns. Found unknown variables:",
-                   paste(unknown_vars, collapse = ", ")))
-      }
+      validate_required_cols(private$data)
+      validate_formula_vars(private$all_formula_vars, private$data)
       logger::log_info("Preparing data for stan")
       private$data <- convert_log_scale(private$data, "value")
       private$data[, `:=`(titre_type_num = as.numeric(as.factor(titre_type)),

diff --git a/R/validation.R b/R/validation.R
@@ -15,3 +15,32 @@ validate_time_type <- function(time_type) {
     stop("'time_type' must be one of 'relative' or 'absolute'")
   }
 }
+
+validate_formula <- function(covariate_formula) {
+  if (!(class(covariate_formula) == "formula")) {
+    stop("'covariate_formula' must be a formula")
+  }
+}
+
+validate_formula_vars <- function(formula_vars, data) {
+  unknown_vars <- formula_vars[which(!(formula_vars %in% names(data)))]
+  if (length(unknown_vars) > 0) {
+    stop(paste("All variables in 'covariate_formula' must correspond to data columns. Found unknown variables:",
+               paste(unknown_vars, collapse = ", ")))
+  }
+}
+
+validate_required_cols <- function(dat) {
+  required_cols <- c("pid", "date", "last_exp_date", "titre_type", "value", "censored")
+  missing_cols <- required_cols[!(required_cols %in% colnames(dat))]
+  if (length(missing_cols) > 0) {
+    stop(paste("Missing required columns:",
+               paste(missing_cols, collapse = ", ")))
+  }
+}
+
+validate_priors <- function(priors) {
+  if (!inherits(priors, "biokinetics_priors")) {
+    stop("'priors' must be of type 'biokinetics_priors'")
+  }
+}
diff --git a/tests/testthat/test-input-validation.R b/tests/testthat/test-input-validation.R
@@ -29,9 +29,14 @@ test_that("Covariate formula must be a formula", {
 })
 
 test_that("Covariates must be present in data", {
+  expect_error(biokinetics$new(file_path = system.file("delta_full.rds", package = "epikinetics"), covariate_formula = 0~ bad),
+               "All variables in 'covariate_formula' must correspond to data columns. Found unknown variables: bad")
+})
+
+test_that("Required columns must be present in data", {
   dat <- data.table(test = 1)
   expect_error(biokinetics$new(data = dat, covariate_formula = 0~ bad),
-               "All variables in 'covariate_formula' must correspond to data columns. Found unknown variables: bad")
+               "Missing required columns: pid, date, last_exp_date, titre_type, value, censored")
 })
 
 test_that("Data must be a data.table", {