date: "18/12/2021"
author: "Nelson Gonzabato"
title: "Changes to shinymde"
shinymde 0.1.1
-
UI overhaul to ensure a much better user experience. Specifically, inputs were decluttered to make it easier to do analyses with minimal interference.
-
Action buttons are now the same color as the app's skin.
-
The sidebar menu is now collapsed on app start.
-
Changes to the UI to ensure that it is less cluttered. Download buttons now appear above plots/tables.
-
dropdown
was dropped in favour ofdropdownButton
. -
The UI is now red instead of blue.
-
All tabs now have drop downs instead of a more cluttered interface.
-
The dashboard side bar is now collapsed on start.
-
renderUI
was dropped in favour ofupdate*
functions which lead to a significant speed improvement. -
Tooltips were dropped. We now use a simpler interface based on
shinyWidgets
. -
There is now a reset button to allow reset of plot option in the "Visualize Missingness" tab.
-
Following changes in
mde
, users can now filter columns inna_summary
based on regular expression matches. -
bsTooltip
is now imported explicitly fromshinyBS
. This is a workaround to allowR CMD check
passes. -
Non-needed directories for build were removed from the shipped package/app.
shinymde 0.1.0
-
shinymde can now be browsed at https://nelson-gon.shinyapps.io/shinymde.
-
Tool tips are now shown above not below a given widget.
-
For the dropping tab, "Percent NA" is now hidden for
drop_all_na
as expected. -
For visual summaries, users can now choose to turn off the text labels on bar plots.
-
For remote datasets, the file extension is now accepted via a selection from a list of currently supported file types.
-
Users can now choose a sheet number for "xlsx" files.
-
There is now a slider input to choose size of arrows and points for lollipop plots.
-
For plot type "lollipop",
fill_variable
is now hidden. The plot is also now not flipped. -
Users can now choose to either use a bar plot or a lollipop plot.
-
Fixed issues with downloading reports after the move to
shinydashboard
. -
shinyBS
is now in theDepends
field to ensure that tootltips work as expected. This is presumably because some other functions should be loaded first for tooltip adders to work. -
There are now tooltips to aid users with the navigation process. This also adds
shinyBS
to the list of app dependencies. -
For inbuilt datasets, only
data.frame
objects are supported. -
The title at the header now includes the package version.
-
There are now runnable examples (essentially one example to launch the app).
-
There is now a highlight when a user hovers over a table row.
-
Hovers are now more "obvious" with target highlighted.
-
The
Reset
button at the input tab now resets the initial text too. The input tab now also provides a user with system details prior to dataset confirmation. -
There is now a homepage that can direct users to project homepage, project
documentation, and author homepage. -
Massive UI changes to improve user experience. In particular, the layout is now based on
shinydashboard
. -
Users now have a reset button at the "Input Data" stage that may be useful in case of mind changes.
-
run_app
was renamedlaunch_app
to avoid name conflicts for example withgolem::run_app
. -
A simple welcome message was added to ensure that users have something visible as they select a dataset. It also serves as a form of credit to the author(s).
-
Tabs except Input Data are now hidden on app startup.
-
Code refactoring was done to allow for easier showing/hiding of elements, conditionally.
-
Missingness summaries now default to a percent_missing sort that is provided in descending order. The user can always change this.
-
User feedback is now provided via
shinyFeedback
which is a more helpful way for users to view feedback directly within the application. -
A reset button now exists to enable a user to restore default values for plot save preferences.
-
Grouped summaries now support exclusion following a fix in mde.
-
There is now a download button to allow saving of plots.
-
All tabs now use a sidebar style which also makes it more easier to navigate and also makes the download button appear more uniform on the app's page.
-
Users now receive a summary of the input data on uploading/selecting a data set. The Input Data tab also now includes a sidebar layout.
-
Visual summaries are now shown in a sidebar layout.
-
It is now possible to download a summary of missingness, a data set with values recoded as requested, or one where values have been dropped.
-
Initially supported operations
-
Data input supports either user data, an inbuilt data set, or a remote data set provided as a web link. For user data, only ".csv", ".xlsx", and ".tsv" are supported.
-
A tabular missingness summary is provided that can be downloaded by the user. The file format to write to is guessed from the user's input.
-
Recoding values supports
recode_as_na
,recode_na_as
,recode_na_if
,recode_as_na_for
and other recode functions from packagemde
. -
Dropping values using
drop_
frommde
. -
A visual summary of missingness is also available. This is currently "just" a visual alternative to the tabular summary of missingness.
- Added a
NEWS.md
file to track changes to the package.