#271640 [dv.edish] Feature requests - User-controlled axes ranges

DESCRIPTION

@@ -1,11 +1,12 @@
 Package: dv.edish
 Type: Package
 Title: eDISH Plot Module for DILI assessment
-Version: 1.1.0
+Version: 1.1.0-1
 Authors@R: 
     c(
       person("Boehringer-Ingelheim Pharma GmbH & Co.KG", role = c("cph", "fnd")),
-      person("Isabel", "Glauss", email = "isabel.glauss@boehringer-ingelheim.com", role = c("aut", "cre"))
+      person("Isabel", "Glauss", email = "isabel.glauss@boehringer-ingelheim.com", role = c("aut", "cre")),
+      person("Mark", "Luff", email = "mark.luff.ext@boehringer-ingelheim.com", role = "aut")
     )
 Description: This DaVinci package contains a module which is meant to support the assessment of drug-induced liver 
     injury by means of the (modified) evaluation of Drug-Induced Serious Hepatotoxicity plot.

NEWS.md

@@ -1,3 +1,7 @@
+# dv.edish 1.1.0-1
+
+- The user can now specify the x- and y-axis range limits, or go with the default `{plotly}` ranges.
+
 # dv.edish 1.1.0
 
 - Remove support for data dispatchers.
@@ -18,7 +22,6 @@
 - When there is no data to display, e.g., due to filter settings, the module will provide a meaningful output.
 - The module has been updated to show commonly used default values for reference lines. For the x-axis, the default value is 3, and for the y-axis, the default value is 2.
 
-
 # dv.edish 1.0.1
 
 - Bugfix: Bookmarking the app's state failed for arm and parameter selection but is fixed now.

R/helper_functions.R

@@ -132,7 +132,7 @@ filter_data <- function(dataset, arm_var, sel_arm, lb_test_var, sel_lb_test) {
 #'
 #' @param dataset `[data.frame]`
 #'
-#' A dataframe containing the variables listed below as columns.
+#' A data frame containing the variables listed below as columns.
 #' @param subjectid_var `[character(1)]`
 #'
 #' Name of the variable containing the unique subject IDs.
@@ -175,7 +175,7 @@ derive_req_vars <- function(
     return(NULL)
   }
 
-  # Get the data-frame in required structure (Pivot wider grouped by certain variables)
+  # Get the data frame in required structure (Pivot wider grouped by certain variables)
   dataset <- dataset %>%
     dplyr::filter(.data[[lb_test_var]] %in% c(sel_x, sel_y)) %>%
     dplyr::mutate(
@@ -185,7 +185,7 @@ derive_req_vars <- function(
     dplyr::select(dplyr::all_of(c(subjectid_var, arm_var, lb_test_var, visit_var, "r_ULN", "r_Baseline"))) %>%
     dplyr::group_by(.data[[subjectid_var]], .data[[arm_var]], .data[[lb_test_var]], .data[[visit_var]]) %>%
     dplyr::mutate(row = dplyr::row_number()) %>%
-    tidyr::pivot_wider(names_from = lb_test_var, values_from = c("r_ULN", "r_Baseline")) %>%
+    tidyr::pivot_wider(names_from = tidyr::all_of(lb_test_var), values_from = c("r_ULN", "r_Baseline")) %>%
     dplyr::select(-dplyr::all_of("row")) %>%
     dplyr::mutate(
       "r_ULN_{{sel_x}}" = as.numeric(.data[[paste0("r_ULN_", sel_x)]]),
@@ -205,7 +205,7 @@ derive_req_vars <- function(
 #'
 #' @param dataset `[data.frame]`
 #'
-#' A dataframe containing the variables listed below as columns.
+#' A data frame containing the variables listed below as columns.
 #' @param subjectid_var `[character(1)]`
 #'
 #' Name of the variable containing the unique subject IDs.
@@ -229,6 +229,24 @@ derive_req_vars <- function(
 #'
 #' Character specifying the plot type for the y-axis. This leads to
 #' using the `dataset`'s column "r_<y_plot_type>_<y_sel>" for the y-values.
+#' @param x_ref_line_num `[numeric(1)]`
+#'
+#' Numeric specifying the reference line for the x-axis.
+#' @param y_ref_line_num `[numeric(1)]`
+#'
+#' Numeric specifying the reference line for the y-axis.
+#' @param x_rng_lower `[numeric(1)]`
+#'
+#' Numeric specifying the lower limit in the x-axis range.
+#' @param x_rng_upper `[numeric(1)]`
+#'
+#' Numeric specifying the upper limit in the x-axis range.
+#' @param y_rng_lower `[numeric(1)]`
+#'
+#' Numeric specifying the lower limit in the y-axis range.
+#' @param y_rng_upper `[numeric(1)]`
+#'
+#' Numeric specifying the upper limit in the y-axis range.
 #'
 #' @return A plotly object specifying the generated eDISH plot.
 #'
@@ -243,11 +261,29 @@ generate_plot <- function(
     x_plot_type,
     y_plot_type,
     x_ref_line_num,
-    y_ref_line_num) {
+    y_ref_line_num,
+    x_rng_lower,
+    x_rng_upper,
+    y_rng_lower,
+    y_rng_upper) {
   if (is.null(dataset)) {
     return(dataset)
   }
 
+  # Prepare x-axis layout based on whether range has been specified
+  layout_xaxis <- list(title = paste0(sel_x, "/", x_plot_type))
+  if (!is.null(x_rng_lower) && !is.null(x_rng_upper)) {
+    layout_xaxis <- c(layout_xaxis,
+                      list(range = c(x_rng_lower, x_rng_upper)))
+  }
+
+  # Prepare y-axis layout based on whether range has been specified
+  layout_yaxis <- list(title = paste0(sel_y, "/", y_plot_type))
+  if (!is.null(y_rng_lower) && !is.null(y_rng_upper)) {
+    layout_yaxis <- c(layout_yaxis,
+                      list(range = c(y_rng_lower, y_rng_upper)))
+  }
+
   plt_obj <- dataset %>%
     plotly::plot_ly(type = "scatter", mode = "markers", color = .[[arm_var]]) %>%
     plotly::add_trace(
@@ -263,8 +299,8 @@ generate_plot <- function(
       hoverinfo = "text"
     ) %>%
     plotly::layout(
-      xaxis = list(title = paste0(sel_x, "/", x_plot_type)),
-      yaxis = list(title = paste0(sel_y, "/", y_plot_type)),
+      xaxis = layout_xaxis,
+      yaxis = layout_yaxis,
       shapes = list(
         list( # vline
           type = "line",

R/mod_edish.R

@@ -10,9 +10,13 @@ EDISH <- pack_of_constants(
   X_REF_ID = "x_ref",
   Y_REF_ID = "y_ref",
   REF_LABEL = "Reference line:",
+  X_RNG_ID = "x_rng",
+  Y_RNG_ID = "y_rng",
+  RNG_LABEL = "Range:",
   X_PLOT_TYPE_ID = "x_plot_type",
   Y_PLOT_TYPE_ID = "y_plot_type",
-  PLOT_TYPE_CHOICES = c("x ULN (eDISH)", "x Baseline (mDISH)"),
+  PLOT_TYPE_CHOICES = c("\u00d7 ULN (eDISH)" = "ULN",
+                        "\u00d7 Baseline (mDISH)" = "Baseline"),
   PLOT_ID = "plot",
   NO_PLOT = "noplot"
 )
@@ -59,6 +63,14 @@ edish_UI <- function(module_id) {
           max = 100,
           step = 0.5
         ),
+        shinyWidgets::numericRangeInput(
+          inputId = ns(EDISH$X_RNG_ID),
+          label = EDISH$RNG_LABEL,
+          value = c(NA, NA),
+          min = 0,
+          max = 100,
+          step = 0.1
+        ),
         shiny::radioButtons(
           inputId = ns(EDISH$X_PLOT_TYPE_ID),
           label = NULL,
@@ -79,6 +91,14 @@ edish_UI <- function(module_id) {
           max = 100,
           step = 0.5
         ),
+        shinyWidgets::numericRangeInput(
+          inputId = ns(EDISH$Y_RNG_ID),
+          label = EDISH$RNG_LABEL,
+          value = c(NA, NA),
+          min = 0,
+          max = 100,
+          step = 0.1
+        ),
         shiny::radioButtons(
           inputId = ns(EDISH$Y_PLOT_TYPE_ID),
           label = NULL,
@@ -269,9 +289,11 @@ edish_server <- function(
         dataset = plot_data(),
         subjectid_var = subjectid_var, arm_var = arm_var, visit_var = visit_var,
         sel_x = input[[EDISH$X_AXIS_ID]], sel_y = input[[EDISH$Y_AXIS_ID]],
-        x_plot_type = ifelse(grepl("eDISH", input[[EDISH$X_PLOT_TYPE_ID]]), "ULN", "Baseline"),
-        y_plot_type = ifelse(grepl("eDISH", input[[EDISH$Y_PLOT_TYPE_ID]]), "ULN", "Baseline"),
-        x_ref_line_num = input[[EDISH$X_REF_ID]], y_ref_line_num = input[[EDISH$Y_REF_ID]]
+        x_plot_type = input[[EDISH$X_PLOT_TYPE_ID]],
+        y_plot_type = input[[EDISH$Y_PLOT_TYPE_ID]],
+        x_ref_line_num = input[[EDISH$X_REF_ID]], y_ref_line_num = input[[EDISH$Y_REF_ID]],
+        x_rng_lower = input[[EDISH$X_RNG_ID]][1], x_rng_upper = input[[EDISH$X_RNG_ID]][2],
+        y_rng_lower = input[[EDISH$Y_RNG_ID]][1], y_rng_upper = input[[EDISH$Y_RNG_ID]][2]
       )
     )
 

README.md

@@ -1,7 +1,7 @@
 # eDISH Module
 
-The {dv.edish} package is meant to support the assessment of drug-induced liver injury 
-by means of the (modified) evaluation of Drug-Induced Serious Hepatotoxicity plot.
+The eDISH module supports the assessment of drug-induced liver injury by means of the (modified)
+evaluation of Drug-Induced Serious Hepatotoxicity plot.
 
 ![](man/figures/full_app.png)
 
@@ -14,19 +14,24 @@ remotes::install_github("Boehringer-Ingelheim/dv.edish")
 
 ## Features
 
-The eDISH module shows a scatter plot depicting correlations of two lab parameters on subject-level. 
-The parameter to be displayed on each axis can be selected by the app user. 
-Moreover, the parameters can be displayed in multiples of either their upper limit normals 
-(resulting in the eDISH plot) or the corresponding subject's baseline values (resulting in the mDISH plot). 
-Horizontal and vertical lines are indicating Hy's law and are customizable as well within the input menu.
+The eDISH module shows a scatter plot depicting correlations of two lab parameters on a
+subject-level.
 
+The parameter to be displayed on each axis can be selected by the app user. Moreover, the parameters
+can be displayed in multiples of either their upper limit normals (resulting in the eDISH plot) or
+the corresponding subject's baseline values (resulting in the mDISH plot).
+
+Horizontal and vertical reference lines indicate Hy's law multiples and can be updated by the user.
+
+By default, the x- and y-axis ranges are determined by the data, but the user can specify their own
+ranges. Both lower and upper values of an axis range need to be specified for the range to take
+effect. If either value is missing then the default axis range is used.
 
 ## Creating an eDISH application
 
 The following example shows how to set up a simple DaVinci app by means of {dv.manager}. 
 The app contains the eDISH module to display dummy data provided by the {pharmaverseadam} package: 
 
-
 ``` r
 dm <- pharmaverseadam::adsl
 lb <- pharmaverseadam::adlb
@@ -50,13 +55,15 @@ dv.manager::run_app(
 
 Note that the module expects two datasets: 
 
-- A Demographics dataset (e.g. `dm` or `adsl`) including, at least, variables containing the following information:
+- A Demographics dataset (e.g. `dm` or `adsl`) including, at least, variables containing the
+  following information:
 
   - Unique subject identifier (e.g. `USUBJID`)
 
   - Arm (e.g. `ACTARM`)
 
-- A Laboratory Test Results dataset (e.g. `lb` or `adlb`) including, at least, variables containing the following information:
+- A Laboratory Test Results dataset (e.g. `lb` or `adlb`) including, at least, variables containing
+  the following information:
 
   - Unique subject identifier (e.g. `USUBJID`)
 
@@ -67,5 +74,3 @@ Note that the module expects two datasets:
   - Numeric lab test result (e.g. `LBSTRESN`)
 
   - Lab test reference range upper limit (e.g. `LBSTNRHI`)
-
-

inst/validation/specs.R

@@ -17,6 +17,7 @@ plot_specs <- specs_list(
   normalization = "The plot displays data normalized according to the selected method.",
   axis_labels = "The plot's axis are labelled according to the parameter selection and normalization method.",
   ref_lines = "The plot displays the specified reference lines.",
+  axis_ranges = "The plot shows the x- and y-axis ranges according to specified lower and upper limits.",
   hovering = "The plot shows details when hovering over a data point.",
   multiple_vals = "In case of multiple values per subject, visit, and lab test, only the maximum value should be shown."
 )

tests/testthat/_snaps/05-mod_edish.md

@@ -39,20 +39,26 @@
       [1] "test 1"
 
       $input$`edish-x_plot_type`
-      [1] "x ULN (eDISH)"
+      [1] "ULN"
 
       $input$`edish-x_ref`
       [1] 3
 
+      $input$`edish-x_rng`
+      NULL
+
       $input$`edish-y_axis`
       [1] "test 2"
 
       $input$`edish-y_plot_type`
-      [1] "x ULN (eDISH)"
+      [1] "ULN"
 
       $input$`edish-y_ref`
       [1] 2
 
+      $input$`edish-y_rng`
+      NULL
+
       $input$`plotly_afterplot-A`
       [1] "\"edish-plot\""
 

tests/testthat/test-04-generate_plot.R

@@ -48,6 +48,10 @@ x_plot_type <- "ULN"
 y_plot_type <- "Baseline"
 x_ref_line_num <- 3
 y_ref_line_num <- 2
+x_rng_lower <- 0
+x_rng_upper <- 10
+y_rng_lower <- 0
+y_rng_upper <- 10
 
 # Invoke the function
 plt_obj <- generate_plot(
@@ -60,7 +64,11 @@ plt_obj <- generate_plot(
   x_plot_type = x_plot_type,
   y_plot_type = y_plot_type,
   x_ref_line_num = x_ref_line_num,
-  y_ref_line_num = y_ref_line_num
+  y_ref_line_num = y_ref_line_num,
+  x_rng_lower = x_rng_lower,
+  x_rng_upper = x_rng_upper,
+  y_rng_lower = y_rng_lower,
+  y_rng_upper = y_rng_upper
 )
 
 # Tests
@@ -110,6 +118,18 @@ test_that("the resulting plot object includes the correct reference lines" %>%
   expect_identical(actual_y1, y_ref_line_num)
 })
 
+test_that("the resulting plot object includes the correct axis range" %>%
+  vdoc[["add_spec"]](specs$plot_specs$axis_ranges), {
+  actual_x <- plt_obj$x$layoutAttrs[[1]]$xaxis$range
+  actual_y <- plt_obj$x$layoutAttrs[[1]]$yaxis$range
+
+  expected_x <- c(x_rng_lower, x_rng_upper)
+  expected_y <- c(y_rng_lower, y_rng_upper)
+
+  expect_identical(actual_x, expected_x)
+  expect_identical(actual_y, expected_y)
+})
+
 test_that("the resulting plot object includes the correct coloring" %>%
   vdoc[["add_spec"]](specs$plot_specs$arm_coloring), {
   actual <- plt_obj$x$attrs[[1]]$color