English / Japanese

About Version 3.2.0

Version 3.2.0 contains improved R package support, and additional analysis features including machine learning and graphics.

Details are shown below. You can refer to this document by Help > About R AnalyticFlow at any time.

New Features

Learning Algorithms

New algorithms are added in the Model > Build Predictive Model operation:

• Random Forest ( ranger )
• Gradient Boosting ( xgboost )
• Elastic Net ( glmnet)

Graphics

Chart > Quick Plot operation offers graphics drawing powered by ggplot2 package.

R Package Support

Some features are added to help using R packages smoothly.

R packages manager is now available. You can attach, detach, install or uninstall packages on this tab.

Warnings are shown on any node or operation when required packages are missing.

If you try to run these operations, you will be asked to install missing packages automatically.

Missing packages are also detected from R scripts written by users.

Improvements

• Interaction terms can be added in regression-type models
• Correlation matrices can be shown graphically
• New project bookmark is added on top of the list

Fixes

• Execution is not correctly done when formulas generated for predictive models are too long