- Fixes failed first interaction when a participant is selected from another module
- Fixes plotly version, previous version numbers failed with an obscure error
- Fixes legend removal when grading was not present in any of the plots
-
Major API redesign with the following breaking changes
mod_patient_profile()
:- The name of the subject-level dataset is now explicitly provided through the
subject_level_dataset_name
parameter. key
has been renamed tosubjid_var
.pt_info
is now calledsummary
and has been simplified. It only allows to specify a single dataframe from which to select columns. Itsrow_item
submember has been renamed tocolumn_count
.tables
is now calledlistings
and is a named list of pairs ofdataset
anddefault_vars
. This allows control over which columns are shown by default for each dataset.- The legacy
basic_info
,vlines
andplots
are all now collected under a singleplots
parameter. They have been restructured for clarity and are also accompanied by apalette
field that makes it possible for the caller to specify individual colors. See the main vignette for more information. - All parameters referring to Study Days have been reframed to point to absolute days. This ensures consistency of the time axis of plots.
- The module now accepts datetimes as well as dates for easier joint configuration with
dv.clinlines
. The "time" part of datetimes is simply discarded. - Configuration helper functions have been removed in favor of documentation templates.
jumping_vars
is now calledsender_ids
for consistency across davinci modules.
- The name of the subject-level dataset is now explicitly provided through the
-
The module is now easier to configure because it:
- Makes all of its main three sections (
summary
,listings
andplots
) optional, which facilitates having a base working configuration. - Provides clear start-up warnings and messages when the module is misconfigured with respect to input datasets. The app creator can safely assume that if no such message appears, the module will not misbehave when interacted with due to misconfiguration. Any remaining error rests solely on the module developer's shoulders or on those of the platform provider.
- Makes all of its main three sections (
-
Bug fixes:
- Day labels now follow the standard CDISC behavior of skipping day 0 on plots (day 1 refers to the Treatment Start Date; day -1 is the day before).
- Vertical lines now cross all plots.
- bugfix: With the new version of dv.filter 2.1.0 the patient selector did not work. This is now fixed.
- bugfix: Updating patient selector after jumping from another module didn't work. This is now fixed.
-
Primary interface:
mod_patient_profile()
-
Launch mock app for demo and testing via
mock_with_mm_app()
-
Enables bookmarking state of Shiny app via URL
- Plot section added.
- Shinydashboard element removed.
- Added a
NEWS.md
file to track changes to the package.