A new version of Rguroo with major new features has been deployed. To ensure proper application start, please use the Sign in/Register link provided at https://rguroo.com/ or use the direct link https://app.rguroo.com/Rguroo/?ver=0.2.525.

In addition to minor bug fixes, the new version includes the following additions/revisions:

  • Multiple object delete: You can select multiple objects and delete them simultaneously. If your selection consists of a dataset with dependencies, those datasets will not be deleted.
  • Regression (prediction): You can obtain predicted values and other statistic for external cases (cases not used in computing the model parameters). External cases are those with their response variable designated as missing (NA).
  • Saving regression diagnostics and predictions: You can save the case-wise diagnostics (e.g., residuals, predictions, Cook’s distance, etc.) as an Rguroo dataset for further plotting, analysis or export.
  • Covariance and correlation matrices: Covariance and correlation matrices both for the variables used in the model and the parameter estimates can be added to the output in the regression function. These choices are listed in the Details under the section Model Estimates and Diagnostic Graphs.
  • NA’s as a factor level: The missing data (NA’s), if they exist in factor (categorical) variables, are consistently considered as a level of a factor in plots and analyses. As before, NA’s can be excluded either in the Variable Type Editor or in the Factor Level Editors.
  • Histogram bin/break selection new options: The default algorithm (Freedman-Diaconis) for selecting the location of bins in a histogram has not changed. However, in this version, the Basics dialog contains a drop-down menu with the following three algorithms for bin locations: Freedman-Diaconis, Scott, and Sturges. Moreover, in the Details menu you can select one of the following four options to determine bin locations:
    • Drop-down menu with algorithms Freedman-Diaconis, Scott, and Sturges
    • Number of bins
    • A center and width for a single bin (the remaining bins will be placed accordingly)
    • Writing an R code that results in numerical values (e.g., seq(2,50,10)) indicating the bin breakpoints.
  • Histogram bin/break selection for factor levels: Bin locations can be specified for each level of a factor in the Factor Level Editor (when histogram by group is being plotted). The default is Freedman-Diaconis and the options of number of bins and R code are also available.
  • Random seed default set to 100: The default seed for every application that uses random number generation is set to 100. If the seed is left blank, every preview will use a different seed, as generated by R.