- Fixed problems in
is_notandis.emptycaused by tidyverse/ggplot2#6352.
- Moving arrow package to 'recommended' because of its size on macOS (> 100MB)
- Require Shiny 1.8.1. Adjustments related to icon-buttons were made to address a breaking change in Shiny 1.8.1
- Reverting changes that removed
req(input$dataset)in different places
- Require shiny 1.8.0. This fixes a bug in the shiny 1.7 versions that caused issues with all radiant packages.
- Fixed a bug in Radiant by changing knitr options. Usernames with a space should no longer cause issues on Windows
- Added a dependency on the arrow package to allow loading and saving parquet files
- Added option to load and save parquet files and data description files (see Data > Manage)
- Renamed Radiant to "Radiant for R" to destinguish from "Radiant for Python"
- Addressed a bug that can occur when combining line graphs in Data > Visualize
- Address deprecation issues in markdown >= 1.5
- Added features in the UI to facilitate persistent filters for filtered, sorted, and sliced data
- Improvements to screenshot feature:
- Navigation bar is omitted and the image is adjusted to the length of the UI.
- html2canvas.js is now included so users can take screenshot when offline
- Added a convenience function
add_descriptionto add a description attribute to a data.frame in markdown format - Line graphs treated more similarly to bar-graphs:
- Can have a binary factor variable on the y-axis
- Y-variable only line are now also possible
- Removed all references to
aes_stringwhich is being deprecated in ggplot soon - Improved cleanup after Radiant UI is closed
- Code cleanup in different areas
- gsub("[\x80-\xFF]", "", text) is no longer valid in R 4.2.0 and above. Non-asci symbols will now be escaped using stringi::stri_trans_general when needed
- Add scrolling for dropdown menus that might extend past the edge of the screen
- Addressed warning messages about Font Awesome icons not existing
- gsub("[\x80-\xFF]", "", text) is no longer valid in R 4.2.0 and above. Non-asci symbols will now be escaped using stringi when needed
- Added option to create screenshots of settings on a page. Approach is inspired by the snapper package by @yonicd
- Added contact request for users on Radiant startup
- Fix issue with R_ZIPCMD when 7zip is on the path but not being recognized by R
- Use
allforis.nullandis.naif object length can be greater than 1 as required in R 4.2.0
- Setup to allow formatting of the shiny interface with bootstrap 4
- Addressing
is_emptyfunction clash withrlang - Upgrading
shinydependency to 1.6.0 and fixing project text alignment issue (@cpsievert, #28)
- Fixes related to breaking changes in
magrittr - Fixes related to changes in
readrargument names - Fix to launch radiant in a "windows"
- Add Google Drive to the default set of directories to explore if available
- Add back functionality to convert a column to type
tsin Data > Transform now that this is again supported by dplyr 1.0.1
- Fix for using the
datefunction from the lubridate package in a filter - Removed functionality to convert a column to type
tsas this is not supported by dplyr 1.0.0 and vctrs 0.3.1 - Updated documentation using r-lib/roxygen2#1109
- Updated styling for formatting for modals (e.g., help pages) that will also allow improved sizing of the (shinyFiles) file browser
- Fix for
\rline-endings in Report > Rmd on Windows. Issue was most likely to occur when copy-and-pasting text from PDF into Report > Rmd.
- Minor adjustments in anticipation of dplyr 1.0.0
- Function to calculate "mode"
- Fix for "spread" in Data > Transform with column name includes "."
- If radiant is not opened from an Rstudio project, use the working directory at launch as the base directory for the application
- Updated styling of Notebook and HTML reports (cosmo + zenburn)
- Documentation updates to link to new video tutorials
- Use
patchworkfor grouping multiple plots together - Apply
refactorto any type in the Data > Transform UI - Fix for
weighted.sdwhen missing values differ forxand weights - Avoid resetting the "Column header" to its default value in Data > Explore when other settings are changed.
- Fix for Data > Transform > Spread when no variables are selected
- Set
debounceto 0 for all shinyAce editors
- Use
zenburnfor code highlighting in Notebook and HTML report from Report > Rmd - Clean up "sf_volumes" from the when radiant is stopped
- Update action buttons that initiate a calculation when one or more relevant inputs are changed. For example, when a model should be re-estimated because the set of explanatory variables was changed by the user, a spinning "refresh" icon will be shown
- Changed default
quantilealgorithm used in thextilefunction from number 2 to 7. See the help forstats::quantilefor details - Added
meandmepropfunctions to calculate the margin of error for a mean and a proportion. Functions are accessible from Data > Pivot and Data > Explore
- Improvements for wrapping generated code to Report > Rmd or Report > R
- Data > Transform > Training now uses the
randomizrpackage to allow blocking variables when creating a training variables.
- Guard against using Data > Transform > Reorder/remove levels with too many levels (i.e., > 100)
- Guard against using Data > Transform > Reorder/remove variables with too many variables (i.e., > 100)
- Fix for DT table callbacks when shiny 1.4 hits CRAN (see rstudio/DT#146 (comment))
- Tables from Data > Pivot and Data > Explore now have
nrset toInfby default (i.e., show all rows). The user can change this to the number of desired rows to show (e.g., select 3 rows in a sorted table) - Fix for example numbering in the help file for Data > Transform
- Numerous small code changes to support enhanced auto-completion, tooltips, and annotations in shinyAce 0.4.1
- Fix for
Data > Transform > Change type - Option to
fix_namesto lower case - Keyboard shortcut (Enter) to load remove csv and rds files
- Use a shinyAce input to generate data descriptions
- Allow custom initial dataset list
- Fix for latex formulas in Report > Rmd on Windows
- Updated requirements for markdown and Rmarkdown
- Fix for
radiant.init.datawith shiny-server - Improvements to setup to allow access to server-side files by adding options to .Rprofile:
- Add
options(radiant.report = TRUE)to allow report generation in Report > Rmd and Report > R - Add
options(radiant.shinyFiles = TRUE)to allow server-side access to files - List specific directories you want to use with radiant using, for example,
options(radiant.sf_volumes = c(Git = "/home/jovyan/git"))
- Add
- Support for series of class
ts(e.g., Data > Transform > Change type > Time series) - Require shinyAce 0.4.0
- Vertical jitter set to 0 by default
- Added option to save Report > Rmd as a powerpoint file using
Rmarkdown - Removed dependency on
summarytoolsdue to breaking changes - Fix for interaction (
iterm) and non-linear term (qterm) creation if character strings rather than integers are passed to the function - Remove specific symbols from reports in Report > Rmd to avoid issues when generating HTML or PDF documents
- Keyboard shortcuts, i.e., CTRL-O and CTRL-S (CMD-O and CMD-S on macOS) to open and save data files in the Data > Manage tab
- Various fixes to address breaking changes in dplyr 0.8.0
- Added
radiant_prefix to all attributes, exceptdescription, to avoid conflicts with other packages (e.g.,varsin dplyr)
- Use
stringi::stri_trans_generalto replace special symbols in Rmarkdown that may cause problems - Add empty line before and after code chunks when saving reports to Rmarkdown
- Use
rioto loadsav,dta, orsas7bdatfiles through theread filesbutton in Report > Rmd and Report > R. - Create a
qscatterplot similar to the function of the same name in Stata - New radiant icon
- Fix for setting where both
xlimandylimare set invisualizefunction - Use an expandable
shinyAceinput for the R-code log in Data > Transform
- Added an "autosave" options. Use
options(radiant.autosave = c(10, 180)); radiant::radiant()to auto-save the application state to the~/.radiant.sessionfolder every 10 minutes for the next 180 minutes. This can be useful if radiant is being used during an exam, for example. - Emergency backups are now saved to
~/.radiant.session/r_some_id.state.rda. The files should be automatically loaded when needed but can also be loaded as a regular radiant state file - Replace option to load an
.rdafrom from a URL in Data > Manage to load.rdsfiles instead - Ensure variable and dataset names are valid for R (i.e., no spaces or symbols), "fixing" the input as needed
- Fix to visualize now
ggplot::labsno longer accepts a list as input - Add option to generate square and cubed terms for use in linear and logistic regression in
radiant.model - Fix for error when trying to save invalid predictions in
radiant.model. This action now generates a pop-up in the browser interface - Add a specified description to a data.frame immediately on
register - Option to pass additional arguments to
shiny::runAppwhen starting radiant such as the port to use. For example, radiant.data::radiant.data("https://github.com/radiant-rstats/docs/raw/gh-pages/examples/demo-dvd-rnd.state.rda", port = 8080) - Option for automatic cleanup of deprecated code in both Report > Rmd and Report > R
- Avoid attempt to fix deprecated code in Report > Rmd if
pred_data = "" - Fix for download icon linked to downloading of a state file after upgrade to shiny 1.2
- Update documentation for Data > Combine
- Fix for
format_dfwhen the data.frame contains missing values. This fix is relevant for severalsummaryfunctions run in Report > Rmd or Report > R - Fix for directory set when using
Knit reportin Report > Rmd and Report > R without an Rstudio project. Will now correctly default to the working directory used in R(studio) - Added option to change
smoothsetting for histograms with a density plot - Similar to
pminandpmax,pfunet al. calculate summary statistics elementwise across multiple vectors - Add
Desktopas a default directory to show in theshinyFilesfile browser - Load a state file on startup by providing a (relative) file path or a url. For example, radiant.data::radiant.data("https://github.com/radiant-rstats/docs/raw/gh-pages/examples/demo-dvd-rnd.state.rda") or radiant.data::radiant.data("assignment.state.rda")
- Update example report in Report > Rmd
- Add
deregisterfunction to remove data in radiant from memory and thedatasetsdropdown list - Fix for invalid column names if used in
Data > Pivot
- Use
summarytoolsto generate summary information for datasets in Data > Manage - Show modal with warning about non-writable working directory when saving reports in Report > Rmd or Report > R
- Apply
radiant.data::fix_namesto files loaded into radiant to ensure valid R-object names - Use the content of the
Store filtered data asinput to name the csv download in Data > View - Add "txt" as a recognized file type for
Read filesin Report > Rmd and Report > R - Allow multiple
linesorloesscurves based on a selectedcolorvariable for scatter plots in Data > Visualize - Indicate that a plot in Data > Visualize should be updated when plot labels are changed
- Fix for #81 when variables used in Data > Pivot contain dots
- Fix for
radiant.project_dirwhen no Rstudio project is used which could cause incorrect relative paths to be used - Fix code formatting for Report > Rmd when arguments include a list (e.g., ggplot labels)
- On Linux use a modal to show code in Report > Rmd and Report > R when reporting is set to "manual"
- Use
is_doubleto ensure dates are not treated as numeric variables in Data > View - Make sort and filter state of tables in Data > Explore and Data > Pivot available in Report > Rmd
- Fix names for data sets loaded using the
Read filesbutton in Report > Rmd or Report > R - Cleanup environment after closing app
- Fix column names with spaces, etc. when reading csv files
- Additional styling and labeling options for Data > Visualize are now available in the browser interface
- Fix for code generation related to DT filters
- Using
shinyFilesto provide convenient access to data located on a server - Avoid
XQuartzrequirement
- Load
data(...)into the current environment rather than defaulting only to the global environment file.renamefailed using docker on windows when saving a report. Usingfile.copyinstead- Fix for
sf_volumesused to set the root directories to load and save files - Set default locale to "en_US.UTF-8" when using shiny-server unless
Sys.getlocale(category = "LC_ALL")what set to something other than "C" - Modal shown if and Rmd (R) file is not available when using "To Rstudio (Rmd)" in Report > Rmd or "To Rstudio (R)" in Report > R
- Track progress loading (state) files
- Fix for
radiant.sf_volumesused for theshinyFilesfile browser - Improvements for sending code from Radiant to Rstudio
- Better support for paths when using radiant on a server (i.e., revert to home directory using
radiant.data::find_home()) - Revert from
svgtopngfor plots in_Report > Rmd_ and _Report > R_.svg` scatter plots with many point get to big for practical use on servers that have to transfer images to a local browser - Removed dependency on
methodspackage
- Fix smart comma's in data descriptions
- Search and replace
desc(n)in reports and replace bydesc(n_obs) - Revert to storing the r_data environment as a list on stop to avoid reference problems (@josh1400)
- Fix for plot type in Data > Pivot in older state files (@josh1400)
- Used all declared imports (CRAN)
- Fix for
radiant.data::explorewhen variable names contain an underscore - Fix for
find_gdrivewhen drive is not being synced - Fixes in Report > Rmd and Report > R to accommodate for pandoc > 2
- Don't update a reactive binding for an object if the binding already exists. See issue rstudio/shiny#2065
- Fix to accommodate changes in
deparsein R 3.5 - Fix for saving data in Data > Manage and generating the relevant R-code
- Use
dev = "svg"for plots in Report > Rmd and Report > R
- Add argument to
dtab.data.frameto format specified columns as a percentage
- Round to the specified number of decimal places even if input if not of type integer (e.g., 2.0)
- When using radiant with Rstudio Viewer or in an Rstudio Window, loading and saving data through Data > Manage generates R-code the user can add to Report > Rmd or Report > R. Clicking the
Show R-codecheckbox displays the R-code used to load or save the current dataset - Various changes to the code to accommodate the use of
shiny::makeReactiveBinding. The advantage is that the code generated for Report > Rmd and Report > R will no longer have to use a list (r_data) to store and access data. This means that code generated and used in the Radiant browser interface will be directly usable without the browser interface as well - Removed
loadr,saver,load_csv,loadcsv_url,loadrds_url, andmake_funsfunctions as they are no longer needed - Deprecated
mean_rm,median_rm,min_rm,max_rm,sd_rm,var_rm, andsum_rmfunctions as they are no longer needed
- Added
load_clipandsave_clipto load and save data to the clipboard on Windows and macOS - Improved auto completion in Report > Rmd and Report > R
- Maintain, store, and clean the settings of the interactive table in Data > View
- Address closing Rstudio Window issue (rstudio/shiny#2033)
- Report > Rmd and Report > R will now be evaluated the
r_dataenvironment. This means that the return value fromls()will be much cleaner
- Add option to load files with extension .rdata or .tsv using
loadrwhich add that data to the Datasets dropdown visualizewill default to a scatter plot ifxvarandyvarare specified but no plottypeis provided in the function call- Improvements to
read_filesfunction to interactively generate R-code (or Rmarkdown code-chunks) to read files in various format (e.g., SQLite, rds, csv, xlsx, css, jpg, etc.). Supports relative paths and usesfind_dropbox()andfind_gdrive()when applicable
- Require
shinyAce0.3.0 - Export
read_filesfunction to interactively generate R-code or Rmarkdown code-chunks to read files in various format (e.g., SQLite, rds, csv, xlsx, css, jpg, etc.). Supports relative paths and usesfind_dropbox()andfind_gdrive()when applicable
- Addins option to start app in Rstudio window
- Upload and download data using the Rstudio file browser. Allows using relative paths to files (e.g., data or images inside an Rstudio project)
- Fix for #43 where scatter plot was not shown for a dataset with less than 1,000 rows
- Fix for Report > Rmd and Report > R when R-code or Rmarkdown is being pulled from the Rstudio editor
- Updated equation example in Report > Rmd
- Use thousand separator for
summary.pivotrandsummary.explore - Fix in code-generation for
table2data
- Changed license for help files and images for radiant.data to CC-BY-SA
- Allow all textarea inputs and multi-select inputs to be resized manually by the user
- Use 200 dpi for plots in Report > Rmd and Report > R
- Data > Visualize now has an option to select a sample of data for scatter plots (e.g., 1K, 5K, 10K, or All)
- Fix for
rounddfto ignore dates
- Apply
fixMSto replace curly quotes, em dash, etc. when using Data > Transform > Create - Option to set number of decimals to show in Data > View
- Improved number formatting in interactive tables in Data > View, Data > Pivot, and Data > Explore
- Option to include an interactive view of a dataset in Report > Rmd. By default, the number of rows is set to 100 as, most likely, the user will not want to embed a large dataset in save HTML report
- Data > Transform will leave variables selected, unless switching to
CreateorSpread - Switch focus to editor in Report > Rmd and Report > R when no other input has focus
- Fix for decimals to show in interactive tables Report > Rmd and saved HTML reports
- Better error messages for
xtileand when binning data with too many groups - Fix for variable type warnings in Data > Pivot when filtering the table
- Fix for \ in equations in Report > Rmd
- Allow response variables with NA values in Model > Logistic regression and other classification models
- Support logicals in code generation from Data > View
- Track window size using
input$get_screen_width - Focus on editor when switching to Report > Rmd or Report > R so generated code is shown immediately and the user can navigate and type in the editor without having to click first
- Add information about the first level when plotting a bar chart with a categorical variable on the Y-axis (e.g., mean(buyer {yes}))
- Cleanup now also occurs when the stop button is used in Rstudio to close the app
- Fix to include
DiagrammeRbased plots in Rmarkdown reports - Fix in
read_filesfor SQLite data names - De-activate spell check auto correction in
selectizeInputin Rstudio Viewer shiny #1916 - Fix to allow selecting and copying text output from Report > Rmd and Report > R
- Remove "fancy" quotes from filters
- Known issue: The Rstudio viewer may not always close the viewer window when trying to stop the application with the
Stoplink in the navbar. As a work-around, use Rstudio's stop buttons instead.
- If Rstudio project is used Report > Rmd and Report > R will use the project directory as base. This allows users to use relative paths and making it easier to share (reproducible) code
- Specify options in .Rprofile for upload memory limit and running Report > Rmd on server
find_projectfunction based onrstudioapi- Report > Rmd Read button to generate code to load various types of data (e.g., rda, rds, xls, yaml, feather)
- Report > Rmd Read button to generate code to load various types of files in report (e.g., jpg, png, md, Rmd, R). If Radiant was started from an Rstudio project, the file paths used will be relative to the project root. Paths to files synced to local Dropbox or Google Drive folder will use the
find_dropboxandfind_gdrivefunctions to enhances reproducibility. - Report > Rmd Load Report button can be used to load Rmarkdown file in the editor. It will also extract the source code from Notebook and HTML files with embedded Rmarkdown
- Report > Rmd will read Rmd directly from Rstudio when "To Rstudio (Rmd)" is selected. This will make it possible to use Rstudio Server Pro's Share project option for realtime collaboration in Radiant
- Long lines of code generated for Report > Rmd will be wrapped to enhance readability
- Report > R is now equivalent to Report > Rmd but in R-code format
- Report > Rmd option to view Editor, Preview, or Both
- Show Rstudio project information in navbar if available
- Overflow
preandcodeblocks in HTML reports generated in Report > Rmd - Read rdata files through Data > Manage
- Enhanced keyboard shortcuts
- Enhanced editing features in Report > Rmd and Report > R based on updates to
shinyAce
- Added preview options to Data > Manage based on radiant-rstats/radiant#30
- Add selected dataset name as default table download name in Data > View, Data > Pivot, and Data > Explore
- Use "stack" as the default for histograms and frequency charts in Data > Visualize
- Cleanup
Stop & Reportoption in navbar - Upgraded tidyr dependency to 0.7
- Upgraded dplyr dependency to 0.7.1
- Fix for large numbers in Data > Explore that could cause an integer overflow
- Export
ggplotlyfromplotlyfor interactive plots in Report > Rmd - Export
subplotfromplotlyfor grids of interactive plots in Report > Rmd - Set default
res = 96forrenderPlotanddpi = 96forknitr::opts_chunk - Add
fillcol,linecol, andpointcoltovisualizeto set plot colors when nofillorcolorvariable has been selected - Reverse legend ordering in Data > Visualize when axes are flipped using
coor_flip() - Added functions to choose.files and choose.dir. Uses JavaScript on Mac, utils::choose.files and utils::choose.dir on Windows, and reverts to file.choose on Linux
- Added
find_gdriveto determine the path to a user's local Google Drive folder if available fixMsfor encoding in reports on Windows
- Chi-square results were not displayed correctly in Data > Pivot
- Fix for
state_multiple
- Specify the maximum number of rows to load for a csv and csv (url) file through Data > Manage
- Support for loading and saving feather files, including specifying the maximum number of rows to load through Data > Manage
- Added author and year arguments to help modals in inst/app/radiant.R (thanks @kmezhoud)
- Added size argument for scatter plots to create bubble charts (thanks @andrewsali)
- Example and CSS formatting for tables in Report > Rmd
- Added
seedargument tomake_train - Added
prop,sdprop, etc. for working with proportions - Set
yliminvisualizefor multiple plots - Show progress indicator when saving reports from Report > Rmd
copy_attrconvenience functionrefactorfunction to keep only a subset of levels in a factor and recode the remaining (and first) level to, for example, otherregisterfunction to add a (transformed) dataset to the dataset dropdown- Remember name of state files loaded and suggest that name when re-saving the state
- Show dataset name in output if dataframe passed directly to analysis function
- R-notebooks are now the default option for output saved from Report > Rmd and Report > R
- Improved documentation on how to customize plots in Report > Rmd
- Keyboard short-cut to put code into Report > Rmd (ALT-enter)
- When clicking the
renamebutton, without changing the name, the dataset was set to NULL (thanks @kmezhoud, radiant-rstats/radiant#5) - Replace ext with .ext in
mutate_eachfunction call - Variance estimation in Data > Explore would cause an error with unit cell-frequencies (thanks @kmezhoud, radiant-rstats/radiant#6)
- Fix for as_integer when factor levels are characters
- Fix for integer conversion in explore
- Remove \r and special characters from strings in r_data and r_state
- Fix sorting in Report > Rmd for tables created using Data > Pivot and Data > Explore when column headers contain symbols or spaces (thanks @4kammer)
- Set
error = TRUEfor rmarkdown for consistency with knitr as used in Report > Rmd - Correctly handle decimal indicators when loading csv files in Data > Manage
- Don't overwrite a dataset to combine if combine generates an error when user sets the the name of the combined data to that of an already selected dataset
- When multiple variables were selected, data were not correctly summarized in Data > Transform
- Add (function) label to bar plot when x-variable is an integer
- Maintain order of variables in Data > Visualize when using "color", "fill", "comby", or "combx"
- Avoid warning when switching datasets in Data > Transform and variables being summarized do not exists in the new dataset
- which.pmax produced a list but needed to be integer
- To customized predictions in radiant.model indexr must be able to customize the prediction dataframe
- describe now correctly resets the working directory on exit
- removed all calls to summarise_each and mutate_each from dplyr
- varp_rm has been deprecated in favor of varpop
- sdp_rm has been deprecated in favor of sdpop
- mutate_each has been deprecated in favor of mutate_at, mutate_all, and radiant.data::mutate_ext