Release Notes Version 9 2: Difference between revisions

From Eigenvector Research Documentation Wiki
Jump to navigation Jump to search
No edit summary
Line 41: Line 41:
|'''[[evricompatibility]]'''
|'''[[evricompatibility]]'''
|
|
* Re-enable checking via website.
* Re-enable version compatibility checking via website.
|----
|----



Revision as of 10:25, 17 January 2023

Changes and Bug Fixes in Version 9.2

Version 9.2 of PLS_Toolbox and Solo is scheduled for release in January, 2023.

(back to Release Notes PLS Toolbox and Solo)

New Features in Solo and PLS_Toolbox

  • BIPLOTS - Include lines drawn from loadings points to origin
  • DATASET2TABLE - Add function to convert DataSets to MATLAB table datatype
  • HMAC - Add object that automatically creates a hierarchical model (PLSDA models) for classification problems
    • Functionality added to the Hierarchical Model Builder to:
      • Load a dataset
      • Select a class set
      • Select preprocessing
      • Automatically build the hierachical model and populate the Hierarchical Model Builder interface
  • PYTHONTOOLS - Add support for Python 3.10 and Mac M1
  • SOFTWAREDEVKIT - Add support for C# and MATLAB to the Solo_Predictor Software Development Kit
  • TESTROBUSTNESS - Enable testrobustness function for non-linear regression methods (lwr, svm, ann, anndl, xgb)

Other Features and Improvements

File Comment
analysis
  • Extend corrmap to yblock.
clutter filter
  • Added capability to use CLS residuals as a clutter source and can cross-validate over alpha parameter for GLSW and over number of PCs for EPO (also know as "Gray CLS").
evricompatibility
  • Re-enable version compatibility checking via website.
experimentreadr
  • Fixed a bug that was introduced in 9.1 to allow experiment file with just file names.
mlr
  • Update internal crossval to give consistent crossvalidation results.
  • Update quadraticInit to be 1 instead of 0.
Parallel Computing
plotgui
  • Can now plot Y-block scores for PLS, PLSDA, and N-PLS models.