From f6ee03a6e2198ea9af17e732ab49f9fe93caed58 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 12:36:40 +0200
Subject: [PATCH 01/14] feat: create (draft) function to include hba1c criteria

---
 R/include-hba1c.R | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)
 create mode 100644 R/include-hba1c.R

diff --git a/R/include-hba1c.R b/R/include-hba1c.R
new file mode 100644
index 0000000..aa578be
--- /dev/null
+++ b/R/include-hba1c.R
@@ -0,0 +1,25 @@
+#' Include only those with HbA1c in the required range.
+#'
+#' In the `lab_forsker` register, NPU27300 is HbA1c in the modern units (IFCC)
+#' while NPU03835 is HbA1c in old units (DCCT).
+#'
+#' @param data The `lab_forsker` register.
+#'
+#' @return An object of the same input type, default as a [tibble::tibble()],
+#'   with two columns: `pnr` and `included_hba1c`.
+#' @export
+#'
+#' @examples
+#' register_data$lab_forsker |> include_hba1c()
+include_hba1c <- function(data) {
+  verify_required_variables(data, "lab_forsker")
+  hba1c_criteria <- get_algorithm_logic("hba1c")
+  data |>
+    column_names_to_lower() |>
+    dplyr::filter({{ hba1c_criteria }}) |>
+    dplyr::transmute(
+      pnr = .data$patient_cpr,
+      included_hba1c = TRUE
+    )
+}
+

From 8c870a5704a3dd2b5a8bde0f35e53aa5bb2afa91 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 12:37:12 +0200
Subject: [PATCH 02/14] feat: internal function to extract logic from algorithm
 data. Untested

---
 R/get-algorithm.R | 7 +++++++
 1 file changed, 7 insertions(+)
 create mode 100644 R/get-algorithm.R

diff --git a/R/get-algorithm.R b/R/get-algorithm.R
new file mode 100644
index 0000000..5aa3c1d
--- /dev/null
+++ b/R/get-algorithm.R
@@ -0,0 +1,7 @@
+get_algorithm_logic <- function(criteria) {
+  algorithm |>
+    dplyr::filter(.data$name == criteria) |>
+    dplyr::pull(.data$logic) |>
+    stringr::str_replace_all("AND", "&") |>
+    stringr::str_replace_all("OR", "|")
+}

From a41fec8d0171cf3b4c6d8924e433060c476976e0 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 16:29:19 +0200
Subject: [PATCH 03/14] docs: add roxygen docs to logic getter

---
 R/get-algorithm.R | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/R/get-algorithm.R b/R/get-algorithm.R
index 5aa3c1d..b3a292e 100644
--- a/R/get-algorithm.R
+++ b/R/get-algorithm.R
@@ -1,3 +1,12 @@
+#' Get the criteria algorithmic logic and convert to an R logic condition.
+#'
+#' @param criteria The name of the inclusion or exclusion criteria to use.
+#'
+#' @return A character string.
+#' @keywords internal
+#'
+#' @examples
+#' get_algorithm_logic("hba1c")
 get_algorithm_logic <- function(criteria) {
   algorithm |>
     dplyr::filter(.data$name == criteria) |>

From 32219d2cdf5569287bcd3ab4a614970f34579da2 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 16:29:38 +0200
Subject: [PATCH 04/14] docs: keep this function as internal only

---
 R/include-hba1c.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/R/include-hba1c.R b/R/include-hba1c.R
index aa578be..873793f 100644
--- a/R/include-hba1c.R
+++ b/R/include-hba1c.R
@@ -7,7 +7,7 @@
 #'
 #' @return An object of the same input type, default as a [tibble::tibble()],
 #'   with two columns: `pnr` and `included_hba1c`.
-#' @export
+#' @keywords internal
 #'
 #' @examples
 #' register_data$lab_forsker |> include_hba1c()

From 3ea542f5ad976efad15d79de41a3b93d91e15c78 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 16:30:42 +0200
Subject: [PATCH 05/14] refactor: keep only earliest two dates, might not work
 with some databases

---
 R/include-hba1c.R | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/R/include-hba1c.R b/R/include-hba1c.R
index 873793f..5d3e98e 100644
--- a/R/include-hba1c.R
+++ b/R/include-hba1c.R
@@ -17,9 +17,16 @@ include_hba1c <- function(data) {
   data |>
     column_names_to_lower() |>
     dplyr::filter({{ hba1c_criteria }}) |>
+    # Keep only the columns we need.
     dplyr::transmute(
       pnr = .data$patient_cpr,
+      date == .data$samplingdate,
       included_hba1c = TRUE
-    )
+    ) |>
+    dplyr::group_by(pnr) |>
+    # This might not work with some databases
+    # Keep earliest two dates.
+    dplyr::slice_min(date, n = 2) |>
+    dplyr::ungroup()
 }
 

From 21c31cee83004785fb58b42254120868ccbeeb27 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 16:31:12 +0200
Subject: [PATCH 06/14] docs: regenerated the roxygen docs

---
 man/get_algorithm_logic.Rd | 21 +++++++++++++++++++++
 man/include_hba1c.Rd       | 23 +++++++++++++++++++++++
 2 files changed, 44 insertions(+)
 create mode 100644 man/get_algorithm_logic.Rd
 create mode 100644 man/include_hba1c.Rd

diff --git a/man/get_algorithm_logic.Rd b/man/get_algorithm_logic.Rd
new file mode 100644
index 0000000..dfcb894
--- /dev/null
+++ b/man/get_algorithm_logic.Rd
@@ -0,0 +1,21 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/get-algorithm.R
+\name{get_algorithm_logic}
+\alias{get_algorithm_logic}
+\title{Get the criteria algorithmic logic and convert to an R logic condition.}
+\usage{
+get_algorithm_logic(criteria)
+}
+\arguments{
+\item{criteria}{The name of the inclusion or exclusion criteria to use.}
+}
+\value{
+A character string.
+}
+\description{
+Get the criteria algorithmic logic and convert to an R logic condition.
+}
+\examples{
+get_algorithm_logic("hba1c")
+}
+\keyword{internal}
diff --git a/man/include_hba1c.Rd b/man/include_hba1c.Rd
new file mode 100644
index 0000000..91e528c
--- /dev/null
+++ b/man/include_hba1c.Rd
@@ -0,0 +1,23 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/include-hba1c.R
+\name{include_hba1c}
+\alias{include_hba1c}
+\title{Include only those with HbA1c in the required range.}
+\usage{
+include_hba1c(data)
+}
+\arguments{
+\item{data}{The \code{lab_forsker} register.}
+}
+\value{
+An object of the same input type, default as a \code{\link[tibble:tibble]{tibble::tibble()}},
+with two columns: \code{pnr} and \code{included_hba1c}.
+}
+\description{
+In the \code{lab_forsker} register, NPU27300 is HbA1c in the modern units (IFCC)
+while NPU03835 is HbA1c in old units (DCCT).
+}
+\examples{
+register_data$lab_forsker |> include_hba1c()
+}
+\keyword{internal}

From c5311ebd34673c77117e402a58012d0f2299226f Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst86@users.noreply.github.com>
Date: Fri, 31 May 2024 15:43:13 +0200
Subject: [PATCH 07/14] docs: apply suggestions from review
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: Signe Kirk Brødbæk <40836345+signekb@users.noreply.github.com>
---
 R/include-hba1c.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/R/include-hba1c.R b/R/include-hba1c.R
index 5d3e98e..94834c9 100644
--- a/R/include-hba1c.R
+++ b/R/include-hba1c.R
@@ -24,7 +24,7 @@ include_hba1c <- function(data) {
       included_hba1c = TRUE
     ) |>
     dplyr::group_by(pnr) |>
-    # This might not work with some databases
+    # FIXME: This might not work with some databases
     # Keep earliest two dates.
     dplyr::slice_min(date, n = 2) |>
     dplyr::ungroup()

From 690dd9ea2ff30389ac8111eae2d177607de99987 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Wed, 19 Jun 2024 11:52:49 +0200
Subject: [PATCH 08/14] fix: :bug: keep earliest two dates, not samples

---
 R/include-hba1c.R | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/R/include-hba1c.R b/R/include-hba1c.R
index 94834c9..f044fe0 100644
--- a/R/include-hba1c.R
+++ b/R/include-hba1c.R
@@ -18,15 +18,17 @@ include_hba1c <- function(data) {
     column_names_to_lower() |>
     dplyr::filter({{ hba1c_criteria }}) |>
     # Keep only the columns we need.
-    dplyr::transmute(
+    dplyr::mutate(
       pnr = .data$patient_cpr,
       date == .data$samplingdate,
-      included_hba1c = TRUE
+      included_hba1c = TRUE,
+      .keep = "none"
     ) |>
+    # Remove any duplicates
+    dplyr::distinct() |>
     dplyr::group_by(pnr) |>
     # FIXME: This might not work with some databases
     # Keep earliest two dates.
     dplyr::slice_min(date, n = 2) |>
     dplyr::ungroup()
 }
-

From 65bbd46db163c8e830d5c22ee3648a9e19eca105 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 12:05:15 +0200
Subject: [PATCH 09/14] chore: add csv file that will host the algorithmic
 logic

---
 data-raw/algorithm.csv | 3 +++
 1 file changed, 3 insertions(+)
 create mode 100644 data-raw/algorithm.csv

diff --git a/data-raw/algorithm.csv b/data-raw/algorithm.csv
new file mode 100644
index 0000000..fa6cebc
--- /dev/null
+++ b/data-raw/algorithm.csv
@@ -0,0 +1,3 @@
+name,logic
+hba1c,(analysiscode == 'NPU27300' AND value >= 48) OR (analysiscode == 'NPU03835' AND value >= 6.5)
+

From c023d0e5613e18e075f92b06bbfd6ce2c026a46d Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 12:06:15 +0200
Subject: [PATCH 10/14] test: add unit test for hba1c inclusion.

---
 tests/testthat.R                    |  1 +
 tests/testthat/test-include-hba1c.R | 64 +++++++++++++++++++++++++++++
 2 files changed, 65 insertions(+)
 create mode 100644 tests/testthat/test-include-hba1c.R

diff --git a/tests/testthat.R b/tests/testthat.R
index c11489d..75a7148 100644
--- a/tests/testthat.R
+++ b/tests/testthat.R
@@ -8,5 +8,6 @@
 
 library(testthat)
 library(osdc)
+library(dplyr)
 
 test_check("osdc")
diff --git a/tests/testthat/test-include-hba1c.R b/tests/testthat/test-include-hba1c.R
new file mode 100644
index 0000000..0cc028e
--- /dev/null
+++ b/tests/testthat/test-include-hba1c.R
@@ -0,0 +1,64 @@
+lab_forsker <- tibble::tribble(
+  ~patient_cpr, ~samplingdate, ~analysiscode, ~value,
+  "498718589803", "20220101", "NPU27300", 49,
+  "498718589804", "20220101", "NPU27300", 47,
+  # Duplicate patient_cpr but with the old units.
+  "498718589803", "20220101", "NPU03835", 6.5,
+  "498718589805", "20220101", "NPU03835", 6.5,
+  "498718589806", "20220101", "NPU03835", 6.3,
+  "498718589807", "20220101", "NPU00000", 100,
+  "498718589808", "20220101", "NPU00000", 5
+)
+
+expected <- tibble::tibble(
+  ~pnr, ~include_hba1c,
+  "498718589803", TRUE,
+  "498718589803", TRUE,
+  "498718589805", TRUE
+)
+
+test_that("dataset has expected variables", {
+  actual <- lab_forsker |>
+    select(-patient_cpr)
+  expect_error(include_hba1c(actual))
+})
+
+test_that("only pnr for HbA1c inclusion is kept", {
+  actual <- include_hba1c(lab_forsker)
+  expect_equal(actual, expected)
+})
+
+test_that("casing of input variables doesn't matter", {
+  actual <- lab_forsker |>
+    rename_with(\(columns) toupper(columns)) |>
+    include_hba1c()
+  expect_equal(actual, expected)
+})
+
+test_that("verification works for DuckDB Database", {
+  actual <- arrow::to_duckdb(lab_forsker) |>
+    include_hba1c()
+
+  expect_equal(actual, expected)
+})
+
+test_that("verification works for Arrow Tables (from Parquet)", {
+  actual <- arrow::as_arrow_table(lab_forsker) |>
+    include_hba1c()
+
+  expect_equal(actual, expected)
+})
+
+test_that("verification works for data.frame", {
+  actual <- as.data.frame(lab_forsker) |>
+    include_hba1c()
+
+  expect_equal(actual, expected)
+})
+
+test_that("verification works for data.table", {
+  actual <- data.table::as.data.table(lab_forsker) |>
+    include_hba1c()
+
+  expect_equal(actual, expected)
+})

From 7457528f36bfc380d91ad23adaf4a476ac564b9a Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 12:11:04 +0200
Subject: [PATCH 11/14] test: fix some small mistakes in tests

---
 tests/testthat/test-include-hba1c.R | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/tests/testthat/test-include-hba1c.R b/tests/testthat/test-include-hba1c.R
index 0cc028e..8ece1cb 100644
--- a/tests/testthat/test-include-hba1c.R
+++ b/tests/testthat/test-include-hba1c.R
@@ -10,20 +10,20 @@ lab_forsker <- tibble::tribble(
   "498718589808", "20220101", "NPU00000", 5
 )
 
-expected <- tibble::tibble(
+expected <- tibble::tribble(
   ~pnr, ~include_hba1c,
   "498718589803", TRUE,
   "498718589803", TRUE,
   "498718589805", TRUE
 )
 
-test_that("dataset has expected variables", {
+test_that("dataset needs expected variables", {
   actual <- lab_forsker |>
     select(-patient_cpr)
   expect_error(include_hba1c(actual))
 })
 
-test_that("only pnr for HbA1c inclusion is kept", {
+test_that("those with inclusion are kept", {
   actual <- include_hba1c(lab_forsker)
   expect_equal(actual, expected)
 })

From 3af8dd297856161c9c5b585a0ac2fb4ff95cf715 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst86@users.noreply.github.com>
Date: Thu, 16 May 2024 16:36:38 +0200
Subject: [PATCH 12/14] test: add an earlier date with the same person

---
 tests/testthat/test-include-hba1c.R | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tests/testthat/test-include-hba1c.R b/tests/testthat/test-include-hba1c.R
index 8ece1cb..53103e3 100644
--- a/tests/testthat/test-include-hba1c.R
+++ b/tests/testthat/test-include-hba1c.R
@@ -1,6 +1,7 @@
 lab_forsker <- tibble::tribble(
   ~patient_cpr, ~samplingdate, ~analysiscode, ~value,
-  "498718589803", "20220101", "NPU27300", 49,
+  "498718589803", "20220101", "NPU27300", 47,
+  "498718589803", "20210101", "NPU27300", 49,
   "498718589804", "20220101", "NPU27300", 47,
   # Duplicate patient_cpr but with the old units.
   "498718589803", "20220101", "NPU03835", 6.5,

From 0fed067dcde99e8dd90a37ba4ca8184892c0b1c7 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Thu, 16 May 2024 16:42:47 +0200
Subject: [PATCH 13/14] test: updates based on meeting discussions

---
 tests/testthat/test-include-hba1c.R | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/tests/testthat/test-include-hba1c.R b/tests/testthat/test-include-hba1c.R
index 53103e3..42d8e06 100644
--- a/tests/testthat/test-include-hba1c.R
+++ b/tests/testthat/test-include-hba1c.R
@@ -1,6 +1,6 @@
 lab_forsker <- tibble::tribble(
   ~patient_cpr, ~samplingdate, ~analysiscode, ~value,
-  "498718589803", "20220101", "NPU27300", 47,
+  "498718589803", "20230101", "NPU27300", 47,
   "498718589803", "20210101", "NPU27300", 49,
   "498718589804", "20220101", "NPU27300", 47,
   # Duplicate patient_cpr but with the old units.
@@ -12,10 +12,10 @@ lab_forsker <- tibble::tribble(
 )
 
 expected <- tibble::tribble(
-  ~pnr, ~include_hba1c,
-  "498718589803", TRUE,
-  "498718589803", TRUE,
-  "498718589805", TRUE
+  ~pnr, ~date, ~included_hba1c,
+  "498718589803", "20210101", TRUE,
+  "498718589803", "20220101", TRUE,
+  "498718589805", "20220101", TRUE
 )
 
 test_that("dataset needs expected variables", {

From f23609a080df097f73ba6f2dfd164ce1a92ab4a5 Mon Sep 17 00:00:00 2001
From: "Luke W. Johnston" <lwjohnst@gmail.com>
Date: Fri, 14 Jun 2024 12:59:49 +0200
Subject: [PATCH 14/14] test: :white_check_mark: add additional tests for the
 filtering to check duplicate dates

---
 tests/testthat/test-include-hba1c.R | 30 ++++++++++++++++++++---------
 1 file changed, 21 insertions(+), 9 deletions(-)

diff --git a/tests/testthat/test-include-hba1c.R b/tests/testthat/test-include-hba1c.R
index 42d8e06..91545b6 100644
--- a/tests/testthat/test-include-hba1c.R
+++ b/tests/testthat/test-include-hba1c.R
@@ -1,26 +1,38 @@
 lab_forsker <- tibble::tribble(
   ~patient_cpr, ~samplingdate, ~analysiscode, ~value,
+  "498718589800", "20230101", "NPU27300", 49,
+  "498718589801", "20230101", "NPU03835", 6.6,
+  "498718589802", "20230101", "NPU03835", 6.3,
   "498718589803", "20230101", "NPU27300", 47,
-  "498718589803", "20210101", "NPU27300", 49,
-  "498718589804", "20220101", "NPU27300", 47,
   # Duplicate patient_cpr but with the old units.
+  "498718589803", "20210101", "NPU27300", 49,
   "498718589803", "20220101", "NPU03835", 6.5,
-  "498718589805", "20220101", "NPU03835", 6.5,
-  "498718589806", "20220101", "NPU03835", 6.3,
-  "498718589807", "20220101", "NPU00000", 100,
-  "498718589808", "20220101", "NPU00000", 5
+  # Duplicate patient_cpr when old and new units are the same date.
+  "498718589805", "20000101", "NPU03835", 6.5,
+  "498718589805", "20000101", "NPU27300", 49,
+  # Duplicate but with old below threshold and new above it.
+  "498718589806", "20000101", "NPU03835", 6.3,
+  "498718589806", "20000101", "NPU27300", 49,
+  # Duplicate but with new below threshold and old above it.
+  "498718589807", "20200101", "NPU03835", 6.6,
+  "498718589807", "20200101", "NPU27300", 47,
+  "498718589808", "20220101", "NPU00000", 100,
+  "498718589809", "20220101", "NPU00000", 5
 )
 
 expected <- tibble::tribble(
   ~pnr, ~date, ~included_hba1c,
+  "498718589800", "20230101", TRUE,
+  "498718589801", "20230101", TRUE,
   "498718589803", "20210101", TRUE,
   "498718589803", "20220101", TRUE,
-  "498718589805", "20220101", TRUE
+  "498718589805", "20000101", TRUE,
+  "498718589806", "20000101", TRUE,
+  "498718589807", "20200101", TRUE
 )
 
 test_that("dataset needs expected variables", {
-  actual <- lab_forsker |>
-    select(-patient_cpr)
+  actual <- lab_forsker
   expect_error(include_hba1c(actual))
 })