Version Clustering (1.7) (17/09/2024)

Welcome to this new version of R++, R++ Clustering! This new version includes a new tool... the Clustering! Many users wished for it and now your favorite statistical analysis software grants it! We also added features to answer your most common requests, and significantly improved performance of the software. You will find below all the details of the new features of this version:

Clustering

The Clustering is the major addition of this version of R++. This technique allows you to split your data in groups sharing common characteristics.

Post hoc tests for Anova and Kruskal-Wallis

The Bonferroni correction is added for the Anova and Kruskal-Wallis tests. This technique allows you to find which pairs of groups have significant differences, as the Anova can only tell you that at least one mean in the tested groups is significantly different from the other means but not which ones. Click on<icone>in the header in statistical tests in a column with an Anova or a Kruskal-Wallis test to compute this correction.

DeLong test in ROC

The DeLong test is added in ROC curves to test if two ROC curves are statistically different. To use the DeLong test, create at least two ROC curves, then click on the<icone>button in the toolbar.

Pre-installed graphs styles for Elsevier and Nature

In the Graph Editor, you can create your own styles to apply your parameters to all your graphs with one click. In addition to the default R++ style, we added two new styles for graphs to add in Elsevier journals and Nature.

Test reports for χ2, Kruskal-Wallis et Fisher

We added test reports for your research papers for the tests χ2, Kruskal-Wallis, and Fisher (support for this last test is only partial for now).

Column management

In Data Management and Statistical Tests, a new tool appears called the Column management! To use it, click on<icone>in the toolbar.

This new feature allow you to move columns in your data! Select a variable in Column management, then move it by drag-and-drop, and the corresponding column in the table moves to the correct spot automatically.

You can also hide columns with a right-click, then choose "Hide" in the popup menu.

It is also possible to move or hide several columns at the same time. Keep the Ctrl (Cmd for Mac) key pressed, then click on the variables you want to move or hide to select all of them at the same time, finally apply a drag-and-drop or right-click on one of these variables to apply the action on all selected variables at once.

You can also delete (right-click) or rename (double-click) columns in Column management.

Multiple paired columns

In Statistical Tests, it is now possible to select several paired columns at the same time. To do so, click on the<icone>button in the toolbar, then select one or more paired columns compatible with the reference column. Finally, click again on<icone>to leave the paired columns selection mode.

Custom number of decimals in Table1

You can now change the number of decimals in numbers in Table1. The number of decimals in p-values can be changed independently.

Support for all models in the filmstrip and the sessions

Before this version, ROC, Survival, and PCA models could not be saved in the filmstrip and R++ sessions. It is now possible ! You can also save Clustering operations in the filmstrip and in sessions.

New types Text and Identifier

Two new types have been added: Text, for variables containing information that shouldn't be used for tests, models, graphs, etc. This is useful for instance for a Comment field in a form, which may contain useful data, but wouldn't make sense as a Nominal variable. In big datasets, this type can be used to increase performance of some features. The Identifier type is like the Text type but with the added constraints that values must be unique and non-missing, which is useful to find issues in your data if variables that are supposed to be identifiers do not respect those constraints.

To use these new types, click on<icone>to the left of the Nominal type item in the Type editor.

Create a Binary variable from the line filter

In the Add column tool, in the General tab, you can now choose the "From the line filter" option. This will create a Binary variable containing for each line whether or not the line is still visible with the current line filter.

Version 1.6.1 (13/12/2023)

This version available on Mac and Windows adds some requested features:

Points coordinates of ROC and Survival curves

A new tab named "Points" was added in ROC and Survival curves. It contains the coordinates of the points of the curves.

In Survival, the new tab contains the time values linked to their corresponding survival probability and confidence interval.

In ROC, the new tab contains the specificity values linked to the sensitivity values.

The Points tab table can be copy-pasted from R++ to any other application such as Microsoft Word.

Cite R++

In the Help menu, we added a new option "Cite R++" that gives you the text to include in your research papers to cite your favorite statistical software.

Unit circle for PCA graphs

PCA graphs now include the unit circle. It can be removed in the graph editor.

Upgrade R to 4.3.1

This new upgrade improves performances and solves a few bugs.

Close data frames from Statistical Tests

For convenience, it was already possible to apply some of the operations of Data Management in Statistical Tests. In this new version, we added the possibility to close data frames directly from Statistical Tests.

Bivariate graphs in bivariate reports

Bivariate reports now include the bivariate graphs of the Statistical Tests section.

Version 1.6 (01/07/2023)

Did you enjoy version 1.5? You are going to love version 1.6! In this new version of R++, you will find Principal Component Analysis (PCA); scalable vector graphics which can be resized with no loss of quality and are required by a growing number of journals; new types of graphs; a remodeled graph editor in which you can custom most graphs generated by R++; automatic reports; and more! As always, we strive to answer to your feedback, and hope this update will meet your needs. You will find below all the details of the new features of this version:

General

Improved design for windows, icons, images

Some windows and icons have been remodeled to be more intuitive.
Additionally, some images that represent our mascot Rhinelle, who guides you during your experience of R++, have been improved to make the software even easier to grasp.

Rework of the toolbars

Toolbars on top of the windows have been re-arranged in sub-sections, so you can find the features you need more easily.

Upgrade R to 4.2.2

This upgrade of the R software embedded in R++ allows for numerous improvements, including performance. Some of you use our software on 4 million lines!

rpp and rppSession files linked to R++

rpp files were introduced in version 1.5 to save your data including additional information such as column types and modality order.
rppSession files save an entire R++ session, including your data and their additional information, graphs and models of the filmstrip, as well as your R scripts.
Since version 1.6, rpp and rppSession files are linked to R++ on your PC (the same feature on Mac is coming soon!). You can now open these files directly without going to the Import section. Also, opening an rpp or rppSession file when R++ is not opened results in R++ opening and importing the file automatically.

Support for dates

Support for dates in R++ now depends on the language of R++. In particular, if your files contain columns with dates in US format, R++ used with the US language will detect those columns as Date columns.

Support for system proxies

R++ now handles system proxies for updates and bug reports.

PCA

Principal Components Analysis is now available! You can find PCA in the Modeling section of the railroad, with the regressions.

Graphs version 2.0

New types of graphs

R++ supports new types of graphs:
  • The Q-Q plot, used in statistical tests and linear and logistic regressions
  • The violin graph, univariate or bivariate
  • The multi-mean graph

Graph editor version 2.0

The graph editor has been entirely reworked to remove unused features and add all the tools you need for your research papers, such as density and normal distribution curves in the histogram, statistical tests between two boxplots in the bivariate boxplots, or the censor points in the survival curves.

Most of the graphs in R++ can now be saved in the filmstrip and modified in the graph editor. This includes graphs in the sections Data Management, Statistical Tests, Linear and Logical Regression, Survival, ROC, and PCA. The only exceptions are the density graphs and the Q-Q plots in the statistical tests, and the diagnostic graphs in the regressions. Graphs saved in the filmstrip can all be saved in an R++ session.

Scalable vector graphics

R++ graphs can now be saved in scalable vector graphics (svg) format. The graph editor displays graphs in scalable vector graphics format, which make them smoother visually, with no loss of quality when resizing the image.

Graphs for Date variables

R++ now generates graphs for Date columns, including bivariate graphs.

Style manager

R++ graphs have a default style, but also allows you to define your own to apply them on all your graphs. This makes it easy to apply a unique look on all the graphs you wish to add to your research papers.

Data management

Reporting version 2.0

New interface more user-friendly for reporting.

Column lock

You can now lock columns to keep them on screen while you explore your data. For instance, lock the name column to always know which name goes with the data on a specific line.

Numeric and date formats

In the typer, you can now apply Dollar or Euro visual formats on Numeric columns, and change the number of decimals. You can also apply Date, Hour, or Date and Hour formats on Date columns.

Auto-completion in add column custom formulas

R++ now auto-completes column names in "Other formula" in the add column window. No more mistakes in your formulas on complicated or long column names!

Fixed order of modalities in graphs

In version 1.6, column graphs always follow the order of modalities.

Empty string is a missing value by default

In version 1.6, replacing a value by an empty string in a column with modalities that do not include the empty string replaces the value with missing value (NA).

Add 1 to the difference between two Date columns

When creating a column that is the difference between two Date columns, it is now possible to add 1 to the entire result. This feature is a request to allow the inclusion of the start and end dates in a result. For example, if a patient arrives at the hospital at noon on Saturday and leaves at noon the next day, this allows you to have the result say they stayed two days, even though mathematically they stayed 24 hours, so one day.

Limits of the Integer column

The range of Integer columns can no longer exceed 20000 for reasons of coherence and performance. Please use the Numeric type for columns that have a range greater than 20000.

Statistical tests

Test reports

R++ now automatically generates test reports that can be directly included in your research papers. For instance:
The ANOVA suggests that:
  - The main effect of ChocSep is statistically significant and large (F(4, 73) = 33.39, p < .001; Eta2 = 0.65, 95% CI [0.53, 1.00])
Effect sizes were labelled following Field's (2013) recommendations.

Q-Q plots for test graphs

For every test, R++ now generates a Q-Q plot along with the density graph.

Bivariate reports

Bivariate reporting is now available! It works similarly to the univariate report.

Column search

You can now search a column in the Statistical Tests section, just as in Data Management!

Export outside of R++

Drag-and-drop from the graph editor

R++ allows the export of a graph (to Word for instance) by drag-and-dropping from the filmstrip. It is now possible to do the same from the graph editor.

Title in the exported graphs

Exported graphs are now exported with their title.

Import

Automatic detection of CSV parameters

Our artificial intelligence now infers optimal parameters when importing a CSV file. Most of the time, you will no longer need to change import parameters for your CSV files!

Import several sheets at the same time in a spreadsheet

In this new version, several sheets can be imported at the same time in a spreadsheet. A data frame is created for each sheet, as if you imported them one by one.

Improved interface when importing from a database

The interface for the import from a database now allows the import of several tables at once, and includes a preview.

Improved detection of Binary columns

Numeric columns that only contains 0s and 1s are now detected as Binary columns.

The order of modalities in a Binary column is now case-insensitive.

Sort by frequency of modalities

For Ordinal or Nominal columns, modalities are now sorted by frequencies by default. To sort them alphabetically, right-click on the header of the column in the typer.

Empty line deletion during import

The option to delete empty lines during import now uses the list of missing values in parameters.

Regressions, ROC, Survival

Intercept selected by default in logistic regression

In version 1.6, the intercept is selected by default in the logistic regression.

Integer column with two values as State variable in ROC

In ROC, an Integer variable with only two values can be used as State variable.

Table1

Deletion of missing values in results

Missing values in Group and Sub-group variables are now removed from results even when values are not missing in the variables to compare.

Homogeneous precision for p-values

The precision of the p-value is now the same as in Statistical Tests.

Licenses

More explicit error messages

Error messages for license problems are clearer in this version, to solve them more easily.