Forest plot module — mod_forest • dv.explorer.parameter

Display a hybrid table/forest plot of arbitrary statistics (correlations, odds ratios, ...) computed on dataset parameters over a single visit.

Usage

forest_UI(
  id,
  numeric_numeric_function_names = character(0),
  numeric_factor_function_names = character(0),
  default_function = NULL
)

forest_server(
  id,
  bm_dataset,
  group_dataset,
  dataset_name = shiny::reactive(character(0)),
  numeric_numeric_functions = list(),
  numeric_factor_functions = list(),
  subjid_var = "SUBJID",
  cat_var = "PARCAT",
  par_var = "PARAM",
  visit_var = "AVISIT",
  value_vars = c("AVAL", "PCHG"),
  default_cat = NULL,
  default_par = NULL,
  default_visit = NULL,
  default_value = NULL,
  default_var = NULL,
  default_group = NULL,
  default_categorical_A = NULL,
  default_categorical_B = NULL
)

mod_forest(
  module_id,
  bm_dataset_name,
  group_dataset_name,
  numeric_numeric_functions = list(`Pearson Correlation` =
    dv.explorer.parameter::pearson_correlation, `Spearman Correlation` =
    dv.explorer.parameter::spearman_correlation),
  numeric_factor_functions = list(`Odds Ratio` = dv.explorer.parameter::odds_ratio),
  subjid_var = "SUBJID",
  cat_var = "PARCAT",
  par_var = "PARAM",
  visit_var = "AVISIT",
  value_vars = c("AVAL", "PCHG"),
  default_cat = NULL,
  default_par = NULL,
  default_visit = NULL,
  default_value = NULL,
  default_var = NULL,
  default_group = NULL,
  default_categorical_A = NULL,
  default_categorical_B = NULL
)

Arguments

id

[character(1)]

Shiny ID

numeric_numeric_function_names, numeric_factor_function_names

[character(1)]

Vectors of names of functions passed as numeric_numeric_functions and numeric_factor_functions to forest_server

default_function

[character(1)]

Default function

bm_dataset

[data.frame()]

An ADBM-like dataset similar in structure to the one in this example, with one record per subject per parameter per analysis visit.

It should have, at least, the columns specified by the parameters subjid_var, cat_var, par_var, visit_var and value_vars. The semantics of these columns are as described in the CDISC standard for variables USUBJID, PARCAT, PARAM, AVISIT and AVAL, respectively.

group_dataset

[data.frame()]

An ADSL-like dataset similar in structure to the one in this example, with one record per subject.

It should contain, at least, the column specified by the parameter subjid_var.

dataset_name

[shiny::reactive(*)]

A reactive that indicates a possible change in the column structure of any of the two datasets

numeric_numeric_functions, numeric_factor_functions

[function(n)]

Named lists of functions. Each function needs to take two parameters and produce a list of four numbers with the following names:

result, CI_lower_limit, CI_upper_limit and p_value

The module will offer the functions as part of its interface and will run each function if selected.

The values returned by the functions are be captured on the output table and are also displayed as part of the forest plot.

numeric_numeric_functions take two numeric parameters (e.g. dv.explorer.parameter::pearson_correlation) and numeric_factor_functions should accept a numeric first parameter and a categorical (factor) second parameter (e.g. dv.explorer.parameter::odds_ratio).

subjid_var

[character(1)]

Column corresponding to the subject ID

cat_var, par_var, visit_var

[character(1)]

Columns from bm_dataset that correspond to the parameter category, parameter and visit

value_vars

[character(n)]

Columns from bm_dataset that correspond to values of the parameters

default_cat, default_par, default_visit, default_value

[character(1)|NULL]

Default values for the selectors

default_var, default_group, default_categorical_A, default_categorical_B

[character(1)|NULL]

Default values for the selectors

module_id

Shiny ID [character(1)]

Module identifier

bm_dataset_name, group_dataset_name

[character(1)]

Dataset names

Functions

forest_UI(): UI
forest_server(): Server